Virgin import of gperf v2.7.2.

39 files changed, 8894 insertions, 1110 deletions

diff --git a/contrib/gperf/AUTHORS b/contrib/gperf/AUTHORS
new file mode 100644
index 000000000000..3429c03166d0
--- /dev/null
+++ b/contrib/gperf/AUTHORS
@@ -0,0 +1,2 @@
+Douglas C. Schmidt  <schmidt@ics.uci.edu>
+Bruno Haible        <haible@clisp.cons.org>
diff --git a/contrib/gperf/COPYING b/contrib/gperf/COPYING
index 9a1703758111..bab08afad347 100644
--- a/contrib/gperf/COPYING
+++ b/contrib/gperf/COPYING
@@ -1,38 +1,41 @@
+                    GNU GENERAL PUBLIC LICENSE
+                       Version 2, June 1991
 
-		    GNU GENERAL PUBLIC LICENSE
-		     Version 1, February 1989
-
- Copyright (C) 1989 Free Software Foundation, Inc.
-                    675 Mass Ave, Cambridge, MA 02139, USA
+ Copyright (C) 1989, 1991 Free Software Foundation, Inc.
+                          59 Temple Place, Suite 330, Boston, MA 02111-1307,
+                          USA.
  Everyone is permitted to copy and distribute verbatim copies
  of this license document, but changing it is not allowed.
 
-			    Preamble
+                            Preamble
 
-  The license agreements of most software companies try to keep users
-at the mercy of those companies.  By contrast, our General Public
+  The licenses for most software are designed to take away your
+freedom to share and change it.  By contrast, the GNU General Public
 License is intended to guarantee your freedom to share and change free
-software--to make sure the software is free for all its users.  The
-General Public License applies to the Free Software Foundation's
-software and to any other program whose authors commit to using it.
-You can use it for your programs, too.
+software--to make sure the software is free for all its users.  This
+General Public License applies to most of the Free Software
+Foundation's software and to any other program whose authors commit to
+using it.  (Some other Free Software Foundation software is covered by
+the GNU Library General Public License instead.)  You can apply it to
+your programs, too.
 
   When we speak of free software, we are referring to freedom, not
-price.  Specifically, the General Public License is designed to make
-sure that you have the freedom to give away or sell copies of free
-software, that you receive source code or can get it if you want it,
-that you can change the software or use pieces of it in new free
-programs; and that you know you can do these things.
+price.  Our General Public Licenses are designed to make sure that you
+have the freedom to distribute copies of free software (and charge for
+this service if you wish), that you receive source code or can get it
+if you want it, that you can change the software or use pieces of it
+in new free programs; and that you know you can do these things.
 
   To protect your rights, we need to make restrictions that forbid
 anyone to deny you these rights or to ask you to surrender the rights.
 These restrictions translate to certain responsibilities for you if you
 distribute copies of the software, or if you modify it.
 
-  For example, if you distribute copies of a such a program, whether
+  For example, if you distribute copies of such a program, whether
 gratis or for a fee, you must give the recipients all the rights that
 you have.  You must make sure that they, too, receive or can get the
-source code.  And you must tell them their rights.
+source code.  And you must show them these terms so they know their
+rights.
 
   We protect your rights with two steps: (1) copyright the software, and
 (2) offer you this license which gives you legal permission to copy,
@@ -45,120 +48,207 @@ want its recipients to know that what they have is not the original, so
 that any problems introduced by others will not reflect on the original
 authors' reputations.
 
+  Finally, any free program is threatened constantly by software
+patents.  We wish to avoid the danger that redistributors of a free
+program will individually obtain patent licenses, in effect making the
+program proprietary.  To prevent this, we have made it clear that any
+patent must be licensed for everyone's free use or not licensed at all.
+
   The precise terms and conditions for copying, distribution and
 modification follow.
 
-		    GNU GENERAL PUBLIC LICENSE
+                    GNU GENERAL PUBLIC LICENSE
    TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
 
-  0. This License Agreement applies to any program or other work which
-contains a notice placed by the copyright holder saying it may be
-distributed under the terms of this General Public License.  The
-"Program", below, refers to any such program or work, and a "work based
-on the Program" means either the Program or any work containing the
-Program or a portion of it, either verbatim or with modifications.  Each
-licensee is addressed as "you".
-
-  1. You may copy and distribute verbatim copies of the Program's source
-code as you receive it, in any medium, provided that you conspicuously and
-appropriately publish on each copy an appropriate copyright notice and
-disclaimer of warranty; keep intact all the notices that refer to this
-General Public License and to the absence of any warranty; and give any
-other recipients of the Program a copy of this General Public License
-along with the Program.  You may charge a fee for the physical act of
-transferring a copy.
-
-  2. You may modify your copy or copies of the Program or any portion of
-it, and copy and distribute such modifications under the terms of Paragraph
-1 above, provided that you also do the following:
-
-    a) cause the modified files to carry prominent notices stating that
-    you changed the files and the date of any change; and
-
-    b) cause the whole of any work that you distribute or publish, that
-    in whole or in part contains the Program or any part thereof, either
-    with or without modifications, to be licensed at no charge to all
-    third parties under the terms of this General Public License (except
-    that you may choose to grant warranty protection to some or all
-    third parties, at your option).
-
-    c) If the modified program normally reads commands interactively when
-    run, you must cause it, when started running for such interactive use
-    in the simplest and most usual way, to print or display an
-    announcement including an appropriate copyright notice and a notice
-    that there is no warranty (or else, saying that you provide a
-    warranty) and that users may redistribute the program under these
-    conditions, and telling the user how to view a copy of this General
-    Public License.
-
-    d) You may charge a fee for the physical act of transferring a
-    copy, and you may at your option offer warranty protection in
-    exchange for a fee.
-
-Mere aggregation of another independent work with the Program (or its
-derivative) on a volume of a storage or distribution medium does not bring
-the other work under the scope of these terms.
+  0. This License applies to any program or other work which contains
+a notice placed by the copyright holder saying it may be distributed
+under the terms of this General Public License.  The "Program", below,
+refers to any such program or work, and a "work based on the Program"
+means either the Program or any derivative work under copyright law:
+that is to say, a work containing the Program or a portion of it,
+either verbatim or with modifications and/or translated into another
+language.  (Hereinafter, translation is included without limitation in
+the term "modification".)  Each licensee is addressed as "you".
+
+Activities other than copying, distribution and modification are not
+covered by this License; they are outside its scope.  The act of
+running the Program is not restricted, and the output from the Program
+is covered only if its contents constitute a work based on the
+Program (independent of having been made by running the Program).
+Whether that is true depends on what the Program does.
+
+  1. You may copy and distribute verbatim copies of the Program's
+source code as you receive it, in any medium, provided that you
+conspicuously and appropriately publish on each copy an appropriate
+copyright notice and disclaimer of warranty; keep intact all the
+notices that refer to this License and to the absence of any warranty;
+and give any other recipients of the Program a copy of this License
+along with the Program.
+
+You may charge a fee for the physical act of transferring a copy, and
+you may at your option offer warranty protection in exchange for a fee.
+
+  2. You may modify your copy or copies of the Program or any portion
+of it, thus forming a work based on the Program, and copy and
+distribute such modifications or work under the terms of Section 1
+above, provided that you also meet all of these conditions:
+
+    a) You must cause the modified files to carry prominent notices
+    stating that you changed the files and the date of any change.
+
+    b) You must cause any work that you distribute or publish, that in
+    whole or in part contains or is derived from the Program or any
+    part thereof, to be licensed as a whole at no charge to all third
+    parties under the terms of this License.
+
+    c) If the modified program normally reads commands interactively
+    when run, you must cause it, when started running for such
+    interactive use in the most ordinary way, to print or display an
+    announcement including an appropriate copyright notice and a
+    notice that there is no warranty (or else, saying that you provide
+    a warranty) and that users may redistribute the program under
+    these conditions, and telling the user how to view a copy of this
+    License.  (Exception: if the Program itself is interactive but
+    does not normally print such an announcement, your work based on
+    the Program is not required to print an announcement.)
 
-  3. You may copy and distribute the Program (or a portion or derivative of
-it, under Paragraph 2) in object code or executable form under the terms of
-Paragraphs 1 and 2 above provided that you also do one of the following:
-
-    a) accompany it with the complete corresponding machine-readable
-    source code, which must be distributed under the terms of
-    Paragraphs 1 and 2 above; or,
-
-    b) accompany it with a written offer, valid for at least three
-    years, to give any third party free (except for a nominal charge
-    for the cost of distribution) a complete machine-readable copy of the
-    corresponding source code, to be distributed under the terms of
-    Paragraphs 1 and 2 above; or,
-
-    c) accompany it with the information you received as to where the
-    corresponding source code may be obtained.  (This alternative is
+These requirements apply to the modified work as a whole.  If
+identifiable sections of that work are not derived from the Program,
+and can be reasonably considered independent and separate works in
+themselves, then this License, and its terms, do not apply to those
+sections when you distribute them as separate works.  But when you
+distribute the same sections as part of a whole which is a work based
+on the Program, the distribution of the whole must be on the terms of
+this License, whose permissions for other licensees extend to the
+entire whole, and thus to each and every part regardless of who wrote it.
+
+Thus, it is not the intent of this section to claim rights or contest
+your rights to work written entirely by you; rather, the intent is to
+exercise the right to control the distribution of derivative or
+collective works based on the Program.
+
+In addition, mere aggregation of another work not based on the Program
+with the Program (or with a work based on the Program) on a volume of
+a storage or distribution medium does not bring the other work under
+the scope of this License.
+
+  3. You may copy and distribute the Program (or a work based on it,
+under Section 2) in object code or executable form under the terms of
+Sections 1 and 2 above provided that you also do one of the following:
+
+    a) Accompany it with the complete corresponding machine-readable
+    source code, which must be distributed under the terms of Sections
+    1 and 2 above on a medium customarily used for software interchange; or,
+
+    b) Accompany it with a written offer, valid for at least three
+    years, to give any third party, for a charge no more than your
+    cost of physically performing source distribution, a complete
+    machine-readable copy of the corresponding source code, to be
+    distributed under the terms of Sections 1 and 2 above on a medium
+    customarily used for software interchange; or,
+
+    c) Accompany it with the information you received as to the offer
+    to distribute corresponding source code.  (This alternative is
     allowed only for noncommercial distribution and only if you
-    received the program in object code or executable form alone.)
-
-Source code for a work means the preferred form of the work for making
-modifications to it.  For an executable file, complete source code means
-all the source code for all modules it contains; but, as a special
-exception, it need not include source code for modules which are standard
-libraries that accompany the operating system on which the executable
-file runs, or for standard header files or definitions files that
-accompany that operating system.
-
-  4. You may not copy, modify, sublicense, distribute or transfer the
-Program except as expressly provided under this General Public License.
-Any attempt otherwise to copy, modify, sublicense, distribute or transfer
-the Program is void, and will automatically terminate your rights to use
-the Program under this License.  However, parties who have received
-copies, or rights to use copies, from you under this General Public
-License will not have their licenses terminated so long as such parties
-remain in full compliance.
-
-  5. By copying, distributing or modifying the Program (or any work based
-on the Program) you indicate your acceptance of this license to do so,
-and all its terms and conditions.
+    received the program in object code or executable form with such
+    an offer, in accord with Subsection b above.)
+
+The source code for a work means the preferred form of the work for
+making modifications to it.  For an executable work, complete source
+code means all the source code for all modules it contains, plus any
+associated interface definition files, plus the scripts used to
+control compilation and installation of the executable.  However, as a
+special exception, the source code distributed need not include
+anything that is normally distributed (in either source or binary
+form) with the major components (compiler, kernel, and so on) of the
+operating system on which the executable runs, unless that component
+itself accompanies the executable.
+
+If distribution of executable or object code is made by offering
+access to copy from a designated place, then offering equivalent
+access to copy the source code from the same place counts as
+distribution of the source code, even though third parties are not
+compelled to copy the source along with the object code.
+
+  4. You may not copy, modify, sublicense, or distribute the Program
+except as expressly provided under this License.  Any attempt
+otherwise to copy, modify, sublicense or distribute the Program is
+void, and will automatically terminate your rights under this License.
+However, parties who have received copies, or rights, from you under
+this License will not have their licenses terminated so long as such
+parties remain in full compliance.
+
+  5. You are not required to accept this License, since you have not
+signed it.  However, nothing else grants you permission to modify or
+distribute the Program or its derivative works.  These actions are
+prohibited by law if you do not accept this License.  Therefore, by
+modifying or distributing the Program (or any work based on the
+Program), you indicate your acceptance of this License to do so, and
+all its terms and conditions for copying, distributing or modifying
+the Program or works based on it.
 
   6. Each time you redistribute the Program (or any work based on the
-Program), the recipient automatically receives a license from the original
-licensor to copy, distribute or modify the Program subject to these
-terms and conditions.  You may not impose any further restrictions on the
-recipients' exercise of the rights granted herein.
+Program), the recipient automatically receives a license from the
+original licensor to copy, distribute or modify the Program subject to
+these terms and conditions.  You may not impose any further
+restrictions on the recipients' exercise of the rights granted herein.
+You are not responsible for enforcing compliance by third parties to
+this License.
+
+  7. If, as a consequence of a court judgment or allegation of patent
+infringement or for any other reason (not limited to patent issues),
+conditions are imposed on you (whether by court order, agreement or
+otherwise) that contradict the conditions of this License, they do not
+excuse you from the conditions of this License.  If you cannot
+distribute so as to satisfy simultaneously your obligations under this
+License and any other pertinent obligations, then as a consequence you
+may not distribute the Program at all.  For example, if a patent
+license would not permit royalty-free redistribution of the Program by
+all those who receive copies directly or indirectly through you, then
+the only way you could satisfy both it and this License would be to
+refrain entirely from distribution of the Program.
+
+If any portion of this section is held invalid or unenforceable under
+any particular circumstance, the balance of the section is intended to
+apply and the section as a whole is intended to apply in other
+circumstances.
+
+It is not the purpose of this section to induce you to infringe any
+patents or other property right claims or to contest validity of any
+such claims; this section has the sole purpose of protecting the
+integrity of the free software distribution system, which is
+implemented by public license practices.  Many people have made
+generous contributions to the wide range of software distributed
+through that system in reliance on consistent application of that
+system; it is up to the author/donor to decide if he or she is willing
+to distribute software through any other system and a licensee cannot
+impose that choice.
+
+This section is intended to make thoroughly clear what is believed to
+be a consequence of the rest of this License.
 
-  7. The Free Software Foundation may publish revised and/or new versions
+  8. If the distribution and/or use of the Program is restricted in
+certain countries either by patents or by copyrighted interfaces, the
+original copyright holder who places the Program under this License
+may add an explicit geographical distribution limitation excluding
+those countries, so that distribution is permitted only in or among
+countries not thus excluded.  In such case, this License incorporates
+the limitation as if written in the body of this License.
+
+  9. The Free Software Foundation may publish revised and/or new versions
 of the General Public License from time to time.  Such new versions will
 be similar in spirit to the present version, but may differ in detail to
 address new problems or concerns.
 
 Each version is given a distinguishing version number.  If the Program
-specifies a version number of the license which applies to it and "any
+specifies a version number of this License which applies to it and "any
 later version", you have the option of following the terms and conditions
 either of that version or of any later version published by the Free
 Software Foundation.  If the Program does not specify a version number of
-the license, you may choose any version ever published by the Free Software
+this License, you may choose any version ever published by the Free Software
 Foundation.
 
-  8. If you wish to incorporate parts of the Program into other free
+  10. If you wish to incorporate parts of the Program into other free
 programs whose distribution conditions are different, write to the author
 to ask for permission.  For software which is copyrighted by the Free
 Software Foundation, write to the Free Software Foundation; we sometimes
@@ -166,9 +256,9 @@ make exceptions for this.  Our decision will be guided by the two goals
 of preserving the free status of all derivatives of our free software and
 of promoting the sharing and reuse of software generally.
 
-			    NO WARRANTY
+                            NO WARRANTY
 
-  9. BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY
+  11. BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY
 FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW.  EXCEPT WHEN
 OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES
 PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED
@@ -178,7 +268,7 @@ TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU.  SHOULD THE
 PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING,
 REPAIR OR CORRECTION.
 
-  10. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING
+  12. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING
 WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR
 REDISTRIBUTE THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES,
 INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING
@@ -188,27 +278,26 @@ YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER
 PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE
 POSSIBILITY OF SUCH DAMAGES.
 
-		     END OF TERMS AND CONDITIONS
+                     END OF TERMS AND CONDITIONS
 
-	Appendix: How to Apply These Terms to Your New Programs
+        Appendix: How to Apply These Terms to Your New Programs
 
   If you develop a new program, and you want it to be of the greatest
-possible use to humanity, the best way to achieve this is to make it
-free software which everyone can redistribute and change under these
-terms.
+possible use to the public, the best way to achieve this is to make it
+free software which everyone can redistribute and change under these terms.
 
-  To do so, attach the following notices to the program.  It is safest to
-attach them to the start of each source file to most effectively convey
-the exclusion of warranty; and each file should have at least the
-"copyright" line and a pointer to where the full notice is found.
+  To do so, attach the following notices to the program.  It is safest
+to attach them to the start of each source file to most effectively
+convey the exclusion of warranty; and each file should have at least
+the "copyright" line and a pointer to where the full notice is found.
 
     <one line to give the program's name and a brief idea of what it does.>
-    Copyright (C) 19yy  <name of author>
+    Copyright (C) <year>  <name of author>
 
     This program is free software; you can redistribute it and/or modify
     it under the terms of the GNU General Public License as published by
-    the Free Software Foundation; either version 1, or (at your option)
-    any later version.
+    the Free Software Foundation; either version 2 of the License, or
+    (at your option) any later version.
 
     This program is distributed in the hope that it will be useful,
     but WITHOUT ANY WARRANTY; without even the implied warranty of
@@ -217,33 +306,35 @@ the exclusion of warranty; and each file should have at least the
 
     You should have received a copy of the GNU General Public License
     along with this program; if not, write to the Free Software
-    Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
+    Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307, USA.
 
 Also add information on how to contact you by electronic and paper mail.
 
 If the program is interactive, make it output a short notice like this
 when it starts in an interactive mode:
 
-    Gnomovision version 69, Copyright (C) 19xx name of author
+    Gnomovision version 69, Copyright (C) year  name of author
     Gnomovision comes with ABSOLUTELY NO WARRANTY; for details type `show w'.
     This is free software, and you are welcome to redistribute it
     under certain conditions; type `show c' for details.
 
-The hypothetical commands `show w' and `show c' should show the
-appropriate parts of the General Public License.  Of course, the
-commands you use may be called something other than `show w' and `show
-c'; they could even be mouse-clicks or menu items--whatever suits your
-program.
+The hypothetical commands `show w' and `show c' should show the appropriate
+parts of the General Public License.  Of course, the commands you use may
+be called something other than `show w' and `show c'; they could even be
+mouse-clicks or menu items--whatever suits your program.
 
 You should also get your employer (if you work as a programmer) or your
 school, if any, to sign a "copyright disclaimer" for the program, if
-necessary.  Here a sample; alter the names:
+necessary.  Here is a sample; alter the names:
 
-  Yoyodyne, Inc., hereby disclaims all copyright interest in the
-  program `Gnomovision' (a program to direct compilers to make passes
-  at assemblers) written by James Hacker.
+  Yoyodyne, Inc., hereby disclaims all copyright interest in the program
+  `Gnomovision' (which makes passes at compilers) written by James Hacker.
 
   <signature of Ty Coon>, 1 April 1989
   Ty Coon, President of Vice
 
-That's all there is to it!
+This General Public License does not permit incorporating your program into
+proprietary programs.  If your program is a subroutine library, you may
+consider it more useful to permit linking proprietary applications with the
+library.  If this is what you want to do, use the GNU Library General
+Public License instead of this License.
diff --git a/contrib/gperf/ChangeLog b/contrib/gperf/ChangeLog
index f22484311903..95ba1fefb2ec 100644
--- a/contrib/gperf/ChangeLog
+++ b/contrib/gperf/ChangeLog
@@ -1,3 +1,229 @@
+2000-09-26  Bruno Haible  <bruno@linuix.math.u-bordeaux.fr>
+
+        * gperf-2.7.2 released.
+
+        * doc/gperf.texi: Add a second bug report address
+          <gperf-bugs@lists.sourceforge.net>.
+        * README: Updated.
+
+2000-08-28  Bruno Haible  <bruno@linuix.math.u-bordeaux.fr>
+
+        * lib/getopt.h (struct option): Use "const" also when compiling in
+          C++ mode. Avoids warnings from Sun CC and HP-UX aCC.
+
+        * doc/Makefile.in (docdir): Change from $(datadir)/doc/@PACKAGE@ to
+          $(prefix)/doc/@PACKAGE@, following the newest GNU standards.
+
+2000-08-20  Bruno Haible  <bruno@linuix.math.u-bordeaux.fr>
+
+        * src/version.cc: Bump version number to 2.7.2.
+        * doc/gperf.texi: Likewise.
+
+        * doc/texinfo.tex: Update to post-texinfo-4.0 version. @code in the
+          title page now chooses a larger font. The overall layout of the
+          text is denser.
+
+        * AUTHORS: New file.
+
+        * tests/Makefile.in (all): Add check-lang-utf8 and check-lang-ucs2.
+          (check-lang-utf8, check-lang-ucs2): New targets.
+          (clean): Remove lu8out and lu2out.
+        * tests/lang-utf8.gperf, tests/lang-utf8.exp: New files.
+        * tests/lang-ucs2.gperf, tests/test2.c, tests/lang-ucs2.in,
+          tests/lang-ucs2.exp: New files.
+
+        Allow the use of embedded NULs in keys.
+        * lib/hash.h (hashpjw): Add a length argument.
+        * lib/hash.cc (hashpjw): Likewise. Don't stop when encountering a NUL
+          character.
+        * src/hash-table.h (Hash_Table constructor): Add ignore_len argument.
+          (Hash_Table::ignore_length): New field.
+          (Hash_Table::insert): Renamed from Hash_Table::operator(). Remove
+          ignore_length argument.
+        * src/hash-table.cc (NIL): Remove macro.
+          (Hash_Table constructor): Add ignore_len argument. Use it to
+          initialize ignore_length.
+          (Hash_Table destructor): Specify explicit length of char_set and
+          key.
+          (Hash_Table::insert): Renamed from Hash_Table::operator(). Remove
+          ignore_length argument. Pass explicit length to hashpjw. Compare
+          char_set using memcmp, not strcmp.
+        * src/list-node.h (List_Node): Rename field length to key_length.
+          New field char_set_length.
+          (List_Node constructor): Accept key and rest, not the entire line.
+        * src/list-node.cc (List_Node constructor): Accept key and rest, not
+          the entire line. Don't NUL terminate key and char_set. Specify
+          explicit length of key. Initialize char_set_length field.
+        * src/key-list.cc: Include <ctype.h>.
+          (parse_line): New function.
+          (Key_List::read_keys): Call parse_line instead of new List_Node.
+          Pass option[NOLENGTH] to Hash_Table constructor, not
+          Hash_Table::insert. Specify explicit length of key and char_set.
+          (Key_List::get_occurrence): Use explicit length of char_set.
+          (Key_List::set_determined): Likewise.
+          (Key_List::already_determined): Likewise.
+          (output_string): Add length argument. Output unprintable characters
+          using octal escape sequence.
+          (output_keyword_entry): Use explicit length of key.
+          (Key_List::output_lookup_array): Specify explicit length of key.
+          (output_switch_case): Likewise.
+          (Key_List::dump): Likewise.
+        * src/gen-perf.h (Gen_Perf::compute_disjoint_union): Add two length
+          arguments.
+        * src/gen-perf.cc (Gen_Perf::compute_disjoint_union): Likewise. Don't
+          stop when encountering NUL characters. Don't NUL terminate the
+          result.
+          (Gen_Perf::hash): Use explicit length of char_set.
+          (Gen_Perf::change): Specify explicit length of key.
+        * doc/gperf.texi: Document it.
+
+        * doc/help2man: New file, help2man version 1.022.
+        * Makefile.devel (all): Add doc/gperf.1.
+          (doc/gperf.1): New target.
+        * doc/gperf.1: Automatically generated.
+
+        * mkinstalldirs: New file, from automake-1.4, grep-2.4.1, guile-1.4,
+          libtool-1.3.3, make-3.79.1, tar-1.13.
+        * src/Makefile.in (MKINSTALLDIRS): New variable.
+          (install, installdirs): Use it instead of mkdir.
+        * doc/Makefile.in (MKINSTALLDIRS): New variable.
+          (install, installdirs): Use it instead of mkdir.
+
+        * INSTALL: Update.
+
+2000-08-19  Bruno Haible  <bruno@linuix.math.u-bordeaux.fr>
+
+        * src/key-list.cc (Output_Compare_Memcmp): New class.
+          (Key_List::output_lookup_function): When option -l is given, use
+          memcmp instead of strcmp or strncmp.
+
+        * doc/gperf.texi: The bug report address is <bug-gnu-utils@gnu.org>.
+          The download address is ftp.gnu.org. Remove mention of -a and -g
+          options (now nops). Explain effect of -c option.
+
+        * doc/configure.in (PACKAGE): New variable.
+        * doc/Makefile.in (datadir, docdir): New variables.
+          (dvidir, htmldir): Change values.
+          (install, installdirs): Update.
+
+        * src/configure.in: Rename cache variable gp_cxx_throw_decl to
+          gp_cv_cxx_throw_decl.
+
+        * src/key-list.cc (Key_List::output_hash_function): When outputting
+          __inline, take advantage of C++ compilers which have inline.
+
+        * src/key-list.cc (Output_Compare_Strncmp::output_comparison):
+          After the call to strncmp, verify that expr2 is not longer than
+          `len'.
+          Reported by Carlo Wood <carlo@runaway.xs4all.nl>.
+
+        * src/key-list.cc (Key_List::output_lookup_function_body): Avoid
+          emitting the loop for dealing with duplicates if
+          total_duplicates == 0.
+
+        * src/key-list.cc (Key_List::read_keys): Don't accept an empty key.
+
+        * src/Makefile.in (install, installdirs, uninstall): Respect
+          $(DESTDIR).
+        * doc/Makefile.in (install, installdirs, uninstall): Likewise.
+
+        * src/options.cc (Options::print_options): Escape the arguments which
+          contain special characters.
+
+        * tests/c-parse.gperf: Updated from gcc-2.95.2/gcc/c-parse.gperf.
+        * tests/objc.gperf: New file, from gcc-2.95.2/gcc/objc/objc.gperf.
+        * tests/chill.gperf: New file, from gcc-2.95.2/gcc/ch/gperf.
+        * tests/cplusplus.gperf: New file, from gcc-2.95.2/gcc/cp/gxx.gperf.
+        * tests/gplus.gperf: Remove file.
+        * tests/java.gperf: New file, from gcc-2.95.2/gcc/java/keyword.gperf.
+        * tests/Makefile: Check them all.
+        * tests/c-parse.exp: Renamed from tests/test-1.exp.
+        * tests/modula2.exp: Renamed from tests/test-2.exp.
+        * tests/cplusplus.exp: Renamed from tests/test-3.exp.
+        * tests/gpc.exp: Renamed from tests/test-5.exp.
+
+        * src/key-list.cc (output_switch_case): Add trailing semicolon to
+          lengthptr assignment line. Fixes bug with -D and -S.
+          From Reini Urban <rurban@sbox.tu-graz.ac.at>. Also reported by
+          David Hunter.
+        * tests/Makefile.in (check-lang-syntax): Perform each test with -D
+          once without and once with duplicates.
+
+        * src/key-list.cc (output_keyword_entry): Avoid outputting a struct
+          initializer of the form {"key",}.
+
+        * src/iterator.cc: Don't include <stream.h>.
+          From Michael Deutschmann <ldeutsch@mail.netshop.net>.
+
+        * tests/Makefile.in (VALIDATE, check-lang-syntax): Use $(srcdir) where
+          appropriate.
+          Reported by Andreas Schwab <schwab@issan.informatik.uni-dortmund.de>.
+
+        * tests/validate: Don't run -traditional tests by default.
+
+        * src/main.cc (main): Check for write error on stdout before returning.
+
+        * src/Makefile.in (LDFLAGS): New definition, to catch the value given
+          at configure time.
+
+        Make the structure initializers customizable. Based on a patch by
+        Kaveh R. Ghazi <ghazi@caip.rutgers.edu>.
+        * src/options.h (Options::get_initializer_suffix,
+          Options::initializer_suffix): New declarations.
+        * src/options.icc (Options::get_initializer_suffix): New function.
+        * src/options.cc (DEFAULT_INITIALIZER_SUFFIX): New constant.
+          (Options::initializer_suffix): New variable.
+          (Options::short_usage): Document option "-F".
+          (Options::long_usage): Document option "-F".
+          (Options constructor): Initialize initializer_suffix.
+          (Options destructor): Dump initializer_suffix.
+          (long_options): Add option "-F".
+          (Options::operator()): Accept option "-F". Sets initializer_suffix.
+        * src/key-list.cc (output_keyword_blank_entries): Output
+          initializer_suffix.
+        * doc/gperf.texi: Document option "-F".
+
+        * COPYING: Replace with GPL version 2 (with new FSF address and Y2K
+          safe year format).
+
+        * doc/gpl.texinfo: New file.
+        * doc/gperf.texi: Document it.
+        * doc/Makefile.in (gperf.info, gperf.dvi, gperf.html, gperf_toc.html):
+          Update dependencies.
+
+        * doc/Makefile.in (MAKEINFO): Unset LANG while running makeinfo.
+
+1998-05-20  Bruno Haible  <bruno@linuix.mathematik.uni-karlsruhe.de>
+
+        * doc/Makefile.in (gperf.dvi, clean): Remove gperf.cps.
+          (install, installdirs, uninstall): Don't install gperf.dvi. The
+          info and HTML documentations are sufficient for on-line use, and
+          users who wish to print the documentation (in PS or DVI format)
+          can do this directly off the source distribution.
+          (DVIPS): Use "-D600" instead of "-Pljfour", for portability.
+
+1998-05-20  Akim Demaille  <demaille@inf.enst.fr>
+
+        * doc/gperf.texi: Many modifications:
+          (Output Format): Declare `hash' and `in_word_set' as functions.
+          (Concept Index): New section.
+          (Title page): Use standard presentation.
+          (Top): Use @top instead of @unnumbered so that automatic master
+          update works.
+          (Motivation): Avoid spaces in @var.
+          (Options): Use the standard name ``Invoking ...''.
+          (Options): Declare also the long form of the options.
+          (Options): Remove redundant @itemize when @table is used.
+
+1998-05-08  Andreas Schwab  <schwab@issan.informatik.uni-dortmund.de>
+
+        * aclocal.m4 (CL_PROG_INSTALL): Set cl_cv_path_install, not
+          ac_cv_path_install.
+
+Sat May  2 13:20:54 1998  Bruno Haible  <bruno@linuix.mathematik.uni-karlsruhe.de>
+
+        * gperf-2.7 released.
+
 Sat May  2 12:31:51 1998  Bruno Haible  <bruno@linuix.mathematik.uni-karlsruhe.de>
 
         * src/version.cc (version_string): Remove the "(C++ version)" suffix.
diff --git a/contrib/gperf/INSTALL b/contrib/gperf/INSTALL
index a2c8722ccaff..350b32f51311 100644
--- a/contrib/gperf/INSTALL
+++ b/contrib/gperf/INSTALL
@@ -31,7 +31,7 @@ The simplest way to compile this package is:
      `sh ./configure' instead to prevent `csh' from trying to execute
      `configure' itself.
 
-     Running `configure' takes awhile.  While running, it prints some
+     Running `configure' takes a while.  While running, it prints some
      messages telling which features it is checking for.
 
   2. Type `make' to compile the package.
@@ -167,7 +167,9 @@ operates.
 `--quiet'
 `--silent'
 `-q'
-     Do not print messages saying which checks are being made.
+     Do not print messages saying which checks are being made.  To
+     suppress all normal output, redirect it to `/dev/null' (any error
+     messages will still be shown).
 
 `--srcdir=DIR'
      Look for the package's source code in directory DIR.  Usually
diff --git a/contrib/gperf/Makefile.devel b/contrib/gperf/Makefile.devel
index e145bd87fdfc..cd6b6282acfe 100644
--- a/contrib/gperf/Makefile.devel
+++ b/contrib/gperf/Makefile.devel
@@ -4,7 +4,7 @@
 SHELL = /bin/sh
 MAKE = make
 
-all : configures src/config.h.in
+all : configures src/config.h.in doc/gperf.1
 
 CONFIGURES = configure lib/configure src/configure tests/configure doc/configure
 
@@ -31,5 +31,8 @@ check-configures : $(CONFIGURES)
 src/config.h.in : src/configure.in aclocal.m4
 	cd src && autoheader -l ..
 
+doc/gperf.1 : force
+	prog=`PATH=build/src:src:$$PATH which gperf`; if test -n "$$prog"; then doc/help2man --name='generate a perfect hash function from a key set' --section=1 $$prog > doc/gperf.1; fi
+
 force :
 
diff --git a/contrib/gperf/NEWS b/contrib/gperf/NEWS
index 28ddd4c14147..9a2877f05384 100644
--- a/contrib/gperf/NEWS
+++ b/contrib/gperf/NEWS
@@ -1,3 +1,13 @@
+New in 2.7.2:
+
+* Keywords may now be enclosed in double quotes; this permits the use of
+  '#', ',', space or NUL inside keywords.
+* Bug fixes.
+
+New in 2.7.1:
+
+* Added option "-F" for gcc.
+
 New in 2.7:
 
 * gperf is now a stand-alone package, untied from libg++.
diff --git a/contrib/gperf/README b/contrib/gperf/README
index bd9d14ea680c..92ec7d20f1ff 100644
--- a/contrib/gperf/README
+++ b/contrib/gperf/README
@@ -1,24 +1,29 @@
-While teaching a data structures course at University of California,
-Irvine, I developed a program called GPERF that generates perfect hash
-functions for sets of key words.  A perfect hash function is simply:
+This is GNU gperf. It is a program that generates perfect hash
+functions for sets of key words.  A perfect hash function is:
 
           A hash function and a data structure that allows 
           recognition of a key word in a set of words using 
           exactly 1 probe into the data structure.
 
-The gperf.texinfo file explains how the program works, the form of the
-input, what options are available, and hints on choosing the best
-options for particular key word sets.  The texinfo file is readable
-both via the GNU emacs `info' command, and is also suitable for
-typesetting with TeX.
+The doc/gperf.html file explains how the program works, the form of
+the input, what options are available, and hints on choosing the best
+options for particular key words set.
 
-The enclosed Makefile creates the executable program ``gperf'' and
-also runs some tests.
+See the file NEWS for a list of major changes in the current release.
+
+See the file INSTALL for compilation and installation instructions.
 
 Output from the GPERF program is used to recognize reserved words in
 the GNU C, GNU C++, and GNU Pascal compilers, as well as with the GNU
 indent program.
 
-Happy hacking!
+For general documentation on the coding and usage standards
+this distribution follows, see the GNU standards document
+ftp://ftp.gnu.org/pub/gnu/standards.*, especially the 'Makefile
+Conventions', 'Configuration', and 'User Interfaces' sections.
+
+Mail suggestions and bug reports to both <bug-gnu-utils@gnu.org> and
+<gperf-bugs@lists.sourceforge.net>.  When reporting bugs, please
+include in the subject line the package name and version (output of
+'gperf --version') for which you found a problem.
 
-Douglas C. Schmidt
diff --git a/contrib/gperf/aclocal.m4 b/contrib/gperf/aclocal.m4
index b39197a46e85..bb22edca357d 100644
--- a/contrib/gperf/aclocal.m4
+++ b/contrib/gperf/aclocal.m4
@@ -37,7 +37,7 @@ AC_CACHE_VAL(cl_cv_path_install,
 	      # AIX installbsd doesn't work without option "-g".
 	      :
 	    else
-	      ac_cv_path_install="$ac_dir/$ac_prog -c"
+	      cl_cv_path_install="$ac_dir/$ac_prog -c"
 	      break 2
 	    fi
 	  fi
diff --git a/contrib/gperf/configure b/contrib/gperf/configure
index 7ac6dd58da6e..77123803d955 100755
--- a/contrib/gperf/configure
+++ b/contrib/gperf/configure
@@ -1,7 +1,7 @@
 #! /bin/sh
 
 # Guess values for system-dependent variables and create Makefiles.
-# Generated automatically using autoconf version 2.12 
+# Generated automatically using autoconf version 2.13 
 # Copyright (C) 1992, 93, 94, 95, 96 Free Software Foundation, Inc.
 #
 # This configure script is free software; the Free Software Foundation
@@ -49,6 +49,7 @@ mandir='${prefix}/man'
 # Initialize some other variables.
 subdirs=
 MFLAGS= MAKEFLAGS=
+SHELL=${CONFIG_SHELL-/bin/sh}
 # Maximum number of lines to put in a shell here document.
 ac_max_here_lines=12
 
@@ -332,7 +333,7 @@ EOF
     verbose=yes ;;
 
   -version | --version | --versio | --versi | --vers)
-    echo "configure generated by autoconf version 2.12"
+    echo "configure generated by autoconf version 2.13"
     exit 0 ;;
 
   -with-* | --with-*)
@@ -502,9 +503,11 @@ ac_ext=c
 # CFLAGS is not in ac_cpp because -g, -O, etc. are not valid cpp options.
 ac_cpp='$CPP $CPPFLAGS'
 ac_compile='${CC-cc} -c $CFLAGS $CPPFLAGS conftest.$ac_ext 1>&5'
-ac_link='${CC-cc} -o conftest $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
+ac_link='${CC-cc} -o conftest${ac_exeext} $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
 cross_compiling=$ac_cv_prog_cc_cross
 
+ac_exeext=
+ac_objext=o
 if (echo "testing\c"; echo 1,2,3) | grep c >/dev/null; then
   # Stardent Vistra SVR4 grep lacks -e, says ghazi@caip.rutgers.edu.
   if (echo -n testing; echo 1,2,3) | sed s/-n/xn/ | grep xn >/dev/null; then
@@ -519,7 +522,7 @@ fi
 
 
 echo $ac_n "checking whether ${MAKE-make} sets \${MAKE}""... $ac_c" 1>&6
-echo "configure:523: checking whether ${MAKE-make} sets \${MAKE}" >&5
+echo "configure:526: checking whether ${MAKE-make} sets \${MAKE}" >&5
 set dummy ${MAKE-make}; ac_make=`echo "$2" | sed 'y%./+-%__p_%'`
 if eval "test \"`echo '$''{'ac_cv_prog_make_${ac_make}_set'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
@@ -586,7 +589,7 @@ EOF
 # Ultrix sh set writes to stderr and can't be redirected directly,
 # and sets the high bit in the cache file unless we assign to the vars.
 (set) 2>&1 |
-  case `(ac_space=' '; set) 2>&1` in
+  case `(ac_space=' '; set | grep ac_space) 2>&1` in
   *ac_space=\ *)
     # `set' does not quote correctly, so add quotes (double-quote substitution
     # turns \\\\ into \\, and sed turns \\ into \).
@@ -665,7 +668,7 @@ do
     echo "running \${CONFIG_SHELL-/bin/sh} $0 $ac_configure_args --no-create --no-recursion"
     exec \${CONFIG_SHELL-/bin/sh} $0 $ac_configure_args --no-create --no-recursion ;;
   -version | --version | --versio | --versi | --vers | --ver | --ve | --v)
-    echo "$CONFIG_STATUS generated by autoconf version 2.12"
+    echo "$CONFIG_STATUS generated by autoconf version 2.13"
     exit 0 ;;
   -help | --help | --hel | --he | --h)
     echo "\$ac_cs_usage"; exit 0 ;;
@@ -684,9 +687,11 @@ sed 's/%@/@@/; s/@%/@@/; s/%g\$/@g/; /@g\$/s/[\\\\&%]/\\\\&/g;
  s/@@/%@/; s/@@/@%/; s/@g\$/%g/' > conftest.subs <<\\CEOF
 $ac_vpsub
 $extrasub
+s%@SHELL@%$SHELL%g
 s%@CFLAGS@%$CFLAGS%g
 s%@CPPFLAGS@%$CPPFLAGS%g
 s%@CXXFLAGS@%$CXXFLAGS%g
+s%@FFLAGS@%$FFLAGS%g
 s%@DEFS@%$DEFS%g
 s%@LDFLAGS@%$LDFLAGS%g
 s%@LIBS@%$LIBS%g
diff --git a/contrib/gperf/doc/Makefile.in b/contrib/gperf/doc/Makefile.in
index bfacd16a13f9..99ac03d1b68c 100644
--- a/contrib/gperf/doc/Makefile.in
+++ b/contrib/gperf/doc/Makefile.in
@@ -1,6 +1,6 @@
 # Makefile for gperf/doc
 
-# Copyright (C) 1998 Free Software Foundation, Inc.
+# Copyright (C) 1998, 2000 Free Software Foundation, Inc.
 #
 # This file is part of GNU GPERF.
 #
@@ -27,11 +27,14 @@ srcdir = @srcdir@
 prefix = @prefix@
 local_prefix = /usr/local
 exec_prefix = @exec_prefix@
+datadir = @datadir@
 infodir = @infodir@
 mandir = @mandir@
 man1dir = $(mandir)/man1
-dvidir = $(mandir)/dvi
-htmldir = $(mandir)/html
+docdir = $(prefix)/doc/@PACKAGE@
+dvidir = $(docdir)
+psdir = $(docdir)
+htmldir = $(docdir)
 
 # Programs used by "make":
 RM = rm -f
@@ -40,14 +43,15 @@ RM = rm -f
 # Programs used by "make" if you have changed the documentation files:
 TEX = tex
 TEXI2DVI = texi2dvi
-DVIPS = dvips -Pljfour
-MAKEINFO = LANGUAGE= makeinfo
+DVIPS = dvips -D600
+MAKEINFO = LANG= LANGUAGE= makeinfo
 TEXI2HTML = texi2html
 
 # Programs used by "make install":
 INSTALL = @INSTALL@
 INSTALL_PROGRAM = @INSTALL_PROGRAM@
 INSTALL_DATA = @INSTALL_DATA@
+MKINSTALLDIRS = $(SHELL) $(srcdir)/../mkinstalldirs
 
 #### End of system configuration section. ####
 
@@ -60,16 +64,16 @@ all : info dvi ps html
 
 info : $(srcdir)/gperf.info
 
-$(srcdir)/gperf.info : $(srcdir)/gperf.texi
+$(srcdir)/gperf.info : $(srcdir)/gperf.texi $(srcdir)/gpl.texinfo
 	cd $(srcdir) && $(MAKEINFO) --no-split gperf.texi
 
 
 dvi : $(srcdir)/gperf.dvi
 
-$(srcdir)/gperf.dvi : $(srcdir)/gperf.texi
-	cd $(srcdir) && $(RM) gperf.aux gperf.toc gperf.cp gperf.fn gperf.ky gperf.pg gperf.tp gperf.vr gperf.log
+$(srcdir)/gperf.dvi : $(srcdir)/gperf.texi $(srcdir)/gpl.texinfo
+	cd $(srcdir) && $(RM) gperf.aux gperf.toc gperf.cp gperf.fn gperf.ky gperf.pg gperf.tp gperf.vr gperf.log gperf.cps
 	cd $(srcdir) && $(TEXI2DVI) gperf.texi
-	cd $(srcdir) && $(RM) gperf.aux gperf.toc gperf.cp gperf.fn gperf.ky gperf.pg gperf.tp gperf.vr gperf.log
+	cd $(srcdir) && $(RM) gperf.aux gperf.toc gperf.cp gperf.fn gperf.ky gperf.pg gperf.tp gperf.vr gperf.log gperf.cps
 
 
 ps : $(srcdir)/gperf.ps
@@ -80,44 +84,46 @@ $(srcdir)/gperf.ps : $(srcdir)/gperf.dvi
 
 html : gperf.html gperf_toc.html
 
-gperf.html : $(srcdir)/gperf.texi
+gperf.html : $(srcdir)/gperf.texi $(srcdir)/gpl.texinfo
 	cd $(srcdir) && $(TEXI2HTML) -expandinfo -number -monolithic gperf.texi
 
-gperf_toc.html : $(srcdir)/gperf.texi
+gperf_toc.html : $(srcdir)/gperf.texi $(srcdir)/gpl.texinfo
 	cd $(srcdir) && $(RM) gperf_*.html
 	cd $(srcdir) && $(TEXI2HTML) -expandinfo -number -split_chapter gperf.texi
 
 
 install : all force
-	if [ ! -d $(infodir) ] ; then mkdir $(infodir) ; fi
-	$(INSTALL_DATA) $(srcdir)/gperf.info $(infodir)/gperf.info
-	if [ ! -d $(mandir) ] ; then mkdir $(mandir) ; fi
-	if [ ! -d $(man1dir) ] ; then mkdir $(man1dir) ; fi
-	$(INSTALL_DATA) $(srcdir)/gperf.1 $(man1dir)/gperf.1
-	if [ ! -d $(dvidir) ] ; then mkdir $(dvidir) ; fi
-	$(INSTALL_DATA) $(srcdir)/gperf.dvi $(dvidir)/gperf.dvi
-	if [ ! -d $(htmldir) ] ; then mkdir $(htmldir) ; fi
-	$(INSTALL_DATA) $(srcdir)/gperf.html $(htmldir)/gperf.html
+	$(MKINSTALLDIRS) $(DESTDIR)$(infodir)
+	$(INSTALL_DATA) $(srcdir)/gperf.info $(DESTDIR)$(infodir)/gperf.info
+	$(MKINSTALLDIRS) $(DESTDIR)$(man1dir)
+	$(INSTALL_DATA) $(srcdir)/gperf.1 $(DESTDIR)$(man1dir)/gperf.1
+#	$(MKINSTALLDIRS) $(DESTDIR)$(dvidir)
+#	$(INSTALL_DATA) $(srcdir)/gperf.dvi $(DESTDIR)$(dvidir)/gperf.dvi
+#	$(MKINSTALLDIRS) $(DESTDIR)$(psdir)
+#	$(INSTALL_DATA) $(srcdir)/gperf.ps $(DESTDIR)$(dvidir)/gperf.ps
+	$(MKINSTALLDIRS) $(DESTDIR)$(htmldir)
+	$(INSTALL_DATA) $(srcdir)/gperf.html $(DESTDIR)$(htmldir)/gperf.html
 
 installdirs : force
-	if [ ! -d $(infodir) ] ; then mkdir $(infodir) ; fi
-	if [ ! -d $(mandir) ] ; then mkdir $(mandir) ; fi
-	if [ ! -d $(man1dir) ] ; then mkdir $(man1dir) ; fi
-	if [ ! -d $(dvidir) ] ; then mkdir $(dvidir) ; fi
-	if [ ! -d $(htmldir) ] ; then mkdir $(htmldir) ; fi
+	$(MKINSTALLDIRS) $(DESTDIR)$(infodir)
+	$(MKINSTALLDIRS) $(DESTDIR)$(man1dir)
+#	$(MKINSTALLDIRS) $(DESTDIR)$(dvidir)
+#	$(MKINSTALLDIRS) $(DESTDIR)$(psdir)
+	$(MKINSTALLDIRS) $(DESTDIR)$(htmldir)
 
 uninstall : force
-	$(RM) $(infodir)/gperf.info
-	$(RM) $(man1dir)/gperf.1
-	$(RM) $(dvidir)/gperf.dvi
-	$(RM) $(htmldir)/gperf.html
+	$(RM) $(DESTDIR)$(infodir)/gperf.info
+	$(RM) $(DESTDIR)$(man1dir)/gperf.1
+#	$(RM) $(DESTDIR)$(dvidir)/gperf.dvi
+#	$(RM) $(DESTDIR)$(psdir)/gperf.ps
+	$(RM) $(DESTDIR)$(htmldir)/gperf.html
 
 check : all
 
 mostlyclean : clean
 
 clean : force
-	$(RM) *~ *.aux *.toc *.cp *.fn *.ky *.pg *.tp *.vr *.my *.log core
+	$(RM) *~ *.aux *.toc *.cp *.fn *.ky *.pg *.tp *.vr *.my *.log *.cps core
 
 distclean : clean
 	$(RM) config.status config.log config.cache Makefile
diff --git a/contrib/gperf/doc/configure b/contrib/gperf/doc/configure
index 7b6e3c3a43b7..a10049f45e63 100755
--- a/contrib/gperf/doc/configure
+++ b/contrib/gperf/doc/configure
@@ -1,7 +1,7 @@
 #! /bin/sh
 
 # Guess values for system-dependent variables and create Makefiles.
-# Generated automatically using autoconf version 2.12 
+# Generated automatically using autoconf version 2.13 
 # Copyright (C) 1992, 93, 94, 95, 96 Free Software Foundation, Inc.
 #
 # This configure script is free software; the Free Software Foundation
@@ -49,6 +49,7 @@ mandir='${prefix}/man'
 # Initialize some other variables.
 subdirs=
 MFLAGS= MAKEFLAGS=
+SHELL=${CONFIG_SHELL-/bin/sh}
 # Maximum number of lines to put in a shell here document.
 ac_max_here_lines=12
 
@@ -332,7 +333,7 @@ EOF
     verbose=yes ;;
 
   -version | --version | --versio | --versi | --vers)
-    echo "configure generated by autoconf version 2.12"
+    echo "configure generated by autoconf version 2.13"
     exit 0 ;;
 
   -with-* | --with-*)
@@ -502,9 +503,11 @@ ac_ext=c
 # CFLAGS is not in ac_cpp because -g, -O, etc. are not valid cpp options.
 ac_cpp='$CPP $CPPFLAGS'
 ac_compile='${CC-cc} -c $CFLAGS $CPPFLAGS conftest.$ac_ext 1>&5'
-ac_link='${CC-cc} -o conftest $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
+ac_link='${CC-cc} -o conftest${ac_exeext} $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
 cross_compiling=$ac_cv_prog_cc_cross
 
+ac_exeext=
+ac_objext=o
 if (echo "testing\c"; echo 1,2,3) | grep c >/dev/null; then
   # Stardent Vistra SVR4 grep lacks -e, says ghazi@caip.rutgers.edu.
   if (echo -n testing; echo 1,2,3) | sed s/-n/xn/ | grep xn >/dev/null; then
@@ -518,8 +521,10 @@ else
 fi
 
 
+PACKAGE=gperf
+
 echo $ac_n "checking whether ${MAKE-make} sets \${MAKE}""... $ac_c" 1>&6
-echo "configure:523: checking whether ${MAKE-make} sets \${MAKE}" >&5
+echo "configure:528: checking whether ${MAKE-make} sets \${MAKE}" >&5
 set dummy ${MAKE-make}; ac_make=`echo "$2" | sed 'y%./+-%__p_%'`
 if eval "test \"`echo '$''{'ac_cv_prog_make_${ac_make}_set'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
@@ -556,7 +561,7 @@ fi
 # SVR4 /usr/ucb/install, which tries to use the nonexistent group "staff"
 # ./install, which can be erroneously created by make from ./install.sh.
 echo $ac_n "checking for a BSD compatible install""... $ac_c" 1>&6
-echo "configure:560: checking for a BSD compatible install" >&5
+echo "configure:565: checking for a BSD compatible install" >&5
 if test -z "$INSTALL"; then
 if eval "test \"`echo '$''{'cl_cv_path_install'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
@@ -581,7 +586,7 @@ else
 	      # AIX installbsd doesn't work without option "-g".
 	      :
 	    else
-	      ac_cv_path_install="$ac_dir/$ac_prog -c"
+	      cl_cv_path_install="$ac_dir/$ac_prog -c"
 	      break 2
 	    fi
 	  fi
@@ -631,7 +636,7 @@ EOF
 # Ultrix sh set writes to stderr and can't be redirected directly,
 # and sets the high bit in the cache file unless we assign to the vars.
 (set) 2>&1 |
-  case `(ac_space=' '; set) 2>&1` in
+  case `(ac_space=' '; set | grep ac_space) 2>&1` in
   *ac_space=\ *)
     # `set' does not quote correctly, so add quotes (double-quote substitution
     # turns \\\\ into \\, and sed turns \\ into \).
@@ -710,7 +715,7 @@ do
     echo "running \${CONFIG_SHELL-/bin/sh} $0 $ac_configure_args --no-create --no-recursion"
     exec \${CONFIG_SHELL-/bin/sh} $0 $ac_configure_args --no-create --no-recursion ;;
   -version | --version | --versio | --versi | --vers | --ver | --ve | --v)
-    echo "$CONFIG_STATUS generated by autoconf version 2.12"
+    echo "$CONFIG_STATUS generated by autoconf version 2.13"
     exit 0 ;;
   -help | --help | --hel | --he | --h)
     echo "\$ac_cs_usage"; exit 0 ;;
@@ -729,9 +734,11 @@ sed 's/%@/@@/; s/@%/@@/; s/%g\$/@g/; /@g\$/s/[\\\\&%]/\\\\&/g;
  s/@@/%@/; s/@@/@%/; s/@g\$/%g/' > conftest.subs <<\\CEOF
 $ac_vpsub
 $extrasub
+s%@SHELL@%$SHELL%g
 s%@CFLAGS@%$CFLAGS%g
 s%@CPPFLAGS@%$CPPFLAGS%g
 s%@CXXFLAGS@%$CXXFLAGS%g
+s%@FFLAGS@%$FFLAGS%g
 s%@DEFS@%$DEFS%g
 s%@LDFLAGS@%$LDFLAGS%g
 s%@LIBS@%$LIBS%g
@@ -750,6 +757,7 @@ s%@includedir@%$includedir%g
 s%@oldincludedir@%$oldincludedir%g
 s%@infodir@%$infodir%g
 s%@mandir@%$mandir%g
+s%@PACKAGE@%$PACKAGE%g
 s%@SET_MAKE@%$SET_MAKE%g
 s%@INSTALL@%$INSTALL%g
 s%@INSTALL_PROGRAM@%$INSTALL_PROGRAM%g
diff --git a/contrib/gperf/doc/configure.in b/contrib/gperf/doc/configure.in
index f72482cba649..944b862cf315 100644
--- a/contrib/gperf/doc/configure.in
+++ b/contrib/gperf/doc/configure.in
@@ -1,6 +1,6 @@
 dnl autoconf configuration for gperf/doc
 
-dnl Copyright (C) 1998 Free Software Foundation, Inc.
+dnl Copyright (C) 1998, 2000 Free Software Foundation, Inc.
 dnl written by Douglas C. Schmidt (schmidt@ics.uci.edu)
 dnl
 dnl This file is part of GNU GPERF.
@@ -21,6 +21,8 @@ dnl Free Software Foundation, 59 Temple Place - Suite 330, Boston,
 dnl MA 02111-1307, USA.
 
 AC_INIT(gperf.1)
+PACKAGE=gperf
+AC_SUBST(PACKAGE)
 AC_PROG_MAKE_SET
 dnl
 dnl           checks for programs
diff --git a/contrib/gperf/doc/gperf.1 b/contrib/gperf/doc/gperf.1
index 1e4a2aa5558c..dd425e0c7e2c 100644
--- a/contrib/gperf/doc/gperf.1
+++ b/contrib/gperf/doc/gperf.1
@@ -1,23 +1,187 @@
-.TH GPERF 1 "March 7, 1998
-.UC 4
+.\" DO NOT MODIFY THIS FILE!  It was generated by help2man 1.022.
+.TH GPERF "1" "September 2000" "GNU gperf 2.7.2" FSF
 .SH NAME
 gperf \- generate a perfect hash function from a key set
 .SH SYNOPSIS
-.B gperf 
-[ 
-.B \-adghijklnoprsStv
-] [ 
-.I keyfile
-]
+.B gperf
+[\fIOPTION\fR]... [\fIINPUT-FILE\fR]
 .SH DESCRIPTION
-
-\fIgperf\fP reads a set of ``keys'' from \fIkeyfile\fP (or, by
-default, from the standard input) and attempts to find a non-minimal
-perfect hashing function that recognizes a member of the key set in
-constant, i.e., O(1), time.  If such a function is found the program
-generates a pair of \fIC\fP source code routines that perform the
-hashing and table lookup.  All generated code is directed to the
-standard output.
-
-Please refer to the \fIgperf.texi\fP file for more information.
-This file is distributed with \fIgperf\fP release.
+GNU `gperf' generates perfect hash functions.
+.PP
+If a long option shows an argument as mandatory, then it is mandatory
+for the equivalent short option also.
+.SS "Input file interpretation:"
+.TP
+\fB\-e\fR, \fB\-\-delimiters\fR=\fIDELIMITER\-LIST\fR
+Allow user to provide a string containing delimiters
+used to separate keywords from their attributes.
+Default is ",\en".
+.TP
+\fB\-t\fR, \fB\-\-struct\-type\fR
+Allows the user to include a structured type
+declaration for generated code. Any text before %%
+is considered part of the type declaration. Key
+words and additional fields may follow this, one
+group of fields per line.
+.SS "Language for the output code:"
+.TP
+\fB\-L\fR, \fB\-\-language\fR=\fILANGUAGE\-NAME\fR
+Generates code in the specified language. Languages
+handled are currently C++, ANSI-C, C, and KR-C. The
+default is C.
+.SS "Details in the output code:"
+.TP
+\fB\-K\fR, \fB\-\-slot\-name\fR=\fINAME\fR
+Select name of the keyword component in the keyword
+structure.
+.TP
+\fB\-F\fR, \fB\-\-initializer\-suffix\fR=\fIINITIALIZERS\fR
+Initializers for additional components in the keyword
+structure.
+.TP
+\fB\-H\fR, \fB\-\-hash\-fn\-name\fR=\fINAME\fR
+Specify name of generated hash function. Default is
+`hash'.
+.TP
+\fB\-N\fR, \fB\-\-lookup\-fn\-name\fR=\fINAME\fR
+Specify name of generated lookup function. Default
+name is `in_word_set'.
+.TP
+\fB\-Z\fR, \fB\-\-class\-name\fR=\fINAME\fR
+Specify name of generated C++ class. Default name is
+`Perfect_Hash'.
+.TP
+\fB\-7\fR, \fB\-\-seven\-bit\fR
+Assume 7-bit characters.
+.TP
+\fB\-c\fR, \fB\-\-compare\-strncmp\fR
+Generate comparison code using strncmp rather than
+strcmp.
+.TP
+\fB\-C\fR, \fB\-\-readonly\-tables\fR
+Make the contents of generated lookup tables
+constant, i.e., readonly.
+.TP
+\fB\-E\fR, \fB\-\-enum\fR
+Define constant values using an enum local to the
+lookup function rather than with defines.
+.TP
+\fB\-I\fR, \fB\-\-includes\fR
+Include the necessary system include file <string.h>
+at the beginning of the code.
+.TP
+\fB\-G\fR, \fB\-\-global\fR
+Generate the static table of keywords as a static
+global variable, rather than hiding it inside of the
+lookup function (which is the default behavior).
+.TP
+\fB\-W\fR, \fB\-\-word\-array\-name\fR=\fINAME\fR
+Specify name of word list array. Default name is
+`wordlist'.
+.TP
+\fB\-S\fR, \fB\-\-switch\fR=\fICOUNT\fR
+Causes the generated C code to use a switch
+statement scheme, rather than an array lookup table.
+This can lead to a reduction in both time and space
+requirements for some keyfiles. The COUNT argument
+determines how many switch statements are generated.
+A value of 1 generates 1 switch containing all the
+elements, a value of 2 generates 2 tables with 1/2
+the elements in each table, etc. If COUNT is very
+large, say 1000000, the generated C code does a
+binary search.
+.TP
+\fB\-T\fR, \fB\-\-omit\-struct\-type\fR
+Prevents the transfer of the type declaration to the
+output file. Use this option if the type is already
+defined elsewhere.
+.SS "Algorithm employed by gperf:"
+.TP
+\fB\-k\fR, \fB\-\-key\-positions\fR=\fIKEYS\fR
+Select the key positions used in the hash function.
+The allowable choices range between 1-126, inclusive.
+The positions are separated by commas, ranges may be
+used, and key positions may occur in any order.
+Also, the meta-character '*' causes the generated
+hash function to consider ALL key positions, and $
+indicates the ``final character'' of a key, e.g.,
+$,1,2,4,6-10.
+.TP
+\fB\-l\fR, \fB\-\-compare\-strlen\fR
+Compare key lengths before trying a string
+comparison. This helps cut down on the number of
+string comparisons made during the lookup.
+.TP
+\fB\-D\fR, \fB\-\-duplicates\fR
+Handle keywords that hash to duplicate values. This
+is useful for certain highly redundant keyword sets.
+.TP
+\fB\-f\fR, \fB\-\-fast\fR=\fIITERATIONS\fR
+Generate the gen-perf.hash function ``fast''. This
+decreases gperf's running time at the cost of
+minimizing generated table size. The numeric
+argument represents the number of times to iterate
+when resolving a collision. `0' means ``iterate by
+the number of keywords''.
+.TP
+\fB\-i\fR, \fB\-\-initial\-asso\fR=\fIN\fR
+Provide an initial value for the associate values
+array. Default is 0. Setting this value larger helps
+inflate the size of the final table.
+.TP
+\fB\-j\fR, \fB\-\-jump\fR=\fIJUMP\-VALUE\fR
+Affects the ``jump value'', i.e., how far to advance
+the associated character value upon collisions. Must
+be an odd number, default is 5.
+.TP
+\fB\-n\fR, \fB\-\-no\-strlen\fR
+Do not include the length of the keyword when
+computing the hash function.
+.TP
+\fB\-o\fR, \fB\-\-occurrence\-sort\fR
+Reorders input keys by frequency of occurrence of
+the key sets. This should decrease the search time
+dramatically.
+.TP
+\fB\-r\fR, \fB\-\-random\fR
+Utilizes randomness to initialize the associated
+values table.
+.TP
+\fB\-s\fR, \fB\-\-size\-multiple\fR=\fIN\fR
+Affects the size of the generated hash table. The
+numeric argument N indicates ``how many times larger
+or smaller'' the associated value range should be,
+in relationship to the number of keys, e.g. a value
+of 3 means ``allow the maximum associated value to
+be about 3 times larger than the number of input
+keys.'' Conversely, a value of \fB\-3\fR means ``make the
+maximum associated value about 3 times smaller than
+the number of input keys. A larger table should
+decrease the time required for an unsuccessful
+search, at the expense of extra table space. Default
+value is 1.
+.SS "Informative output:"
+.TP
+\fB\-h\fR, \fB\-\-help\fR
+Print this message.
+.TP
+\fB\-v\fR, \fB\-\-version\fR
+Print the gperf version number.
+.TP
+\fB\-d\fR, \fB\-\-debug\fR
+Enables the debugging option (produces verbose
+output to the standard error).
+.SH "REPORTING BUGS"
+Report bugs to <bug-gnu-utils@gnu.org>.
+.SH "SEE ALSO"
+The full documentation for
+.B gperf
+is maintained as a Texinfo manual.  If the
+.B info
+and
+.B gperf
+programs are properly installed at your site, the command
+.IP
+.B info gperf
+.PP
+should give you access to the complete manual.
diff --git a/contrib/gperf/doc/gperf.texi b/contrib/gperf/doc/gperf.texi
index c4a35754c9f0..e510ac97021f 100644
--- a/contrib/gperf/doc/gperf.texi
+++ b/contrib/gperf/doc/gperf.texi
@@ -1,28 +1,42 @@
-\input texinfo  @c -*-texinfo-*-
-
-@settitle User's Guide to @code{gperf}
+\input texinfo @c -*- texinfo -*-
+@c %**start of header
 @setfilename gperf.info
-
-@ifinfo
-@format
-START-INFO-DIR-ENTRY
+@settitle Perfect Hash Function Generator
+@c @setchapternewpage odd
+@c %**end of header
+
+@c some day we should @include version.texi instead of defining
+@c these values at hand.
+@set UPDATED 26 September 2000
+@set EDITION 2.7.2
+@set VERSION 2.7.2
+@c ---------------------
+
+@c remove the black boxes generated in the GPL appendix.
+@finalout
+
+@c Merge functions into the concept index
+@syncodeindex fn cp
+@c @synindex pg cp
+
+@dircategory Programming Tools
+@direntry
 * Gperf: (gperf).                Perfect Hash Function Generator.
-END-INFO-DIR-ENTRY
-@end format
-@end ifinfo
+@end direntry
 
 @ifinfo
-This file documents the features of the GNU Perfect Hash Function Generator
+This file documents the features of the GNU Perfect Hash Function
+Generator @value{VERSION}.
 
-Copyright (C) 1989-1998 Free Software Foundation, Inc.
+Copyright @copyright{} 1989-2000 Free Software Foundation, Inc.
 
-Permission is granted to make and distribute verbatim copies of
-this manual provided the copyright notice and this permission notice
-are preserved on all copies.
+Permission is granted to make and distribute verbatim copies of this
+manual provided the copyright notice and this permission notice are
+preserved on all copies.
 
 @ignore
-Permission is granted to process this file through @TeX{} and print the
-results, provided the printed document carries copying permission
+Permission is granted to process this file through TeX and print the
+results, provided the printed document carries a copying permission
 notice identical to this one except for the removal of this paragraph
 (this paragraph not being relevant to the printed manual).
 
@@ -32,32 +46,26 @@ Permission is granted to copy and distribute modified versions of this
 manual under the conditions for verbatim copying, provided also that the
 section entitled ``GNU General Public License'' is included exactly as
 in the original, and provided that the entire resulting derived work is
-distributed under the terms of a permission notice identical to this one.
+distributed under the terms of a permission notice identical to this
+one.
 
 Permission is granted to copy and distribute translations of this manual
 into another language, under the above conditions for modified versions,
-except that the section entitled ``GNU @code{gperf} General Public License'' and
-this permission notice may be included in translations approved by the
-Free Software Foundation instead of in the original English.
-@end ifinfo
+except that the section entitled ``GNU General Public License'' and this
+permission notice may be included in translations approved by the Free
+Software Foundation instead of in the original English.
 
-@setchapternewpage odd
+@end ifinfo
 
 @titlepage
-@center @titlefont{User's Guide}
-@sp 2
-@center @titlefont{for the}
-@sp 2
-@center @titlefont{GNU GPERF Utility}
-@sp 4
-@center Douglas C. Schmidt
-@sp 3
-@center last updated 14 April 1998
-@sp 1
-@center for version 2.7
+@title User's Guide to @code{gperf} @value{VERSION}
+@subtitle The GNU Perfect Hash Function Generator
+@subtitle Edition @value{EDITION}, @value{UPDATED}
+@author Douglas C. Schmidt
+
 @page
 @vskip 0pt plus 1filll
-Copyright @copyright{} 1989-1998 Free Software Foundation, Inc.
+Copyright @copyright{} 1989-2000 Free Software Foundation, Inc.
 
 
 Permission is granted to make and distribute verbatim copies of
@@ -66,364 +74,78 @@ are preserved on all copies.
 
 Permission is granted to copy and distribute modified versions of this
 manual under the conditions for verbatim copying, provided also that the
-section entitled ``GNU @code{gperf} General Public License'' is included exactly as
-in the original, and provided that the entire resulting derived work is
-distributed under the terms of a permission notice identical to this one.
+section entitled ``GNU General Public License'' is included
+exactly as in the original, and provided that the entire resulting
+derived work is distributed under the terms of a permission notice
+identical to this one.
 
 Permission is granted to copy and distribute translations of this manual
 into another language, under the above conditions for modified versions,
-except that the section entitled ``GNU @code{gperf} General Public License'' may be
-included in a translation approved by the author instead of in the original
-English.
+except that the section entitled ``GNU General Public License'' may be
+included in a translation approved by the author instead of in the
+original English.
 @end titlepage
 
 @ifinfo
 @node Top, Copying, (dir), (dir)
-@unnumbered Introduction
+@top Introduction
 
 This manual documents the GNU @code{gperf} perfect hash function generator
 utility, focusing on its features and how to use them, and how to report
 bugs.
 
-@end ifinfo
 @menu
-* Copying::			GNU @code{gperf} General Public License says
+* Copying::                     GNU @code{gperf} General Public License says
                                 how you can copy and share @code{gperf}.
-* Contributors::		People who have contributed to @code{gperf}.
-* Motivation::			Static search structures and GNU GPERF.
-* Search Structures::		Static search structures and GNU @code{gperf}
-* Description::			High-level discussion of how GPERF functions.
-* Options::			A description of options to the program.
-* Bugs::			Known bugs and limitations with GPERF.
-* Projects::			Things still left to do.
-* Implementation::		Implementation Details for GNU GPERF.
-* Bibliography::		Material Referenced in this Report.
-
- --- The Detailed Node Listing ---
-
-High-Level Description of GNU @code{gperf}
-
-* Input Format::		Input Format to @code{gperf}
-* Output Format::		Output Format for Generated C Code with @code{gperf}
-
-Input Format to @code{gperf}
-
-* Declarations::		@code{struct} Declarations and C Code Inclusion.
-* Keywords::			Format for Keyword Entries.
-* Functions::			Including Additional C Functions.
-
-Options to the @code{gperf} Utility
-
-* Input Details::          Options that affect Interpretation of the Input File
-* Output Language::        Specifying the Language for the Output Code
-* Output Details::         Fine tuning Details in the Output Code
-* Algorithmic Details::    Changing the Algorithms employed by @code{gperf}
-* Verbosity::              Informative Output
-@end menu
-
-@node Copying, Contributors, Top, Top
-@unnumbered GNU GENERAL PUBLIC LICENSE
-@center Version 1, February 1989
-
-@display
-Copyright @copyright{} 1989 Free Software Foundation, Inc.
-675 Mass Ave, Cambridge, MA 02139, USA
-
-Everyone is permitted to copy and distribute verbatim copies
-of this license document, but changing it is not allowed.
-@end display
-
-@unnumberedsec Preamble
-
-  The license agreements of most software companies try to keep users
-at the mercy of those companies.  By contrast, our General Public
-License is intended to guarantee your freedom to share and change free
-software---to make sure the software is free for all its users.  The
-General Public License applies to the Free Software Foundation's
-software and to any other program whose authors commit to using it.
-You can use it for your programs, too.
-
-  When we speak of free software, we are referring to freedom, not
-price.  Specifically, the General Public License is designed to make
-sure that you have the freedom to give away or sell copies of free
-software, that you receive source code or can get it if you want it,
-that you can change the software or use pieces of it in new free
-programs; and that you know you can do these things.
-
-  To protect your rights, we need to make restrictions that forbid
-anyone to deny you these rights or to ask you to surrender the rights.
-These restrictions translate to certain responsibilities for you if you
-distribute copies of the software, or if you modify it.
-
-  For example, if you distribute copies of a such a program, whether
-gratis or for a fee, you must give the recipients all the rights that
-you have.  You must make sure that they, too, receive or can get the
-source code.  And you must tell them their rights.
-
-  We protect your rights with two steps: (1) copyright the software, and
-(2) offer you this license which gives you legal permission to copy,
-distribute and/or modify the software.
-
-  Also, for each author's protection and ours, we want to make certain
-that everyone understands that there is no warranty for this free
-software.  If the software is modified by someone else and passed on, we
-want its recipients to know that what they have is not the original, so
-that any problems introduced by others will not reflect on the original
-authors' reputations.
-
-  The precise terms and conditions for copying, distribution and
-modification follow.
-
-@iftex
-@unnumberedsec TERMS AND CONDITIONS
-@end iftex
-@ifinfo
-@center TERMS AND CONDITIONS
-@end ifinfo
+* Contributors::                People who have contributed to @code{gperf}.
+* Motivation::                  Static search structures and GNU GPERF.
+* Search Structures::           Static search structures and GNU @code{gperf}
+* Description::                 High-level discussion of how GPERF functions.
+* Options::                     A description of options to the program.
+* Bugs::                        Known bugs and limitations with GPERF.
+* Projects::                    Things still left to do.
+* Implementation::              Implementation Details for GNU GPERF.
+* Bibliography::                Material Referenced in this Report.
 
-@enumerate
-@item
-This License Agreement applies to any program or other work which
-contains a notice placed by the copyright holder saying it may be
-distributed under the terms of this General Public License.  The
-``Program'', below, refers to any such program or work, and a ``work based
-on the Program'' means either the Program or any work containing the
-Program or a portion of it, either verbatim or with modifications.  Each
-licensee is addressed as ``you''.
-
-@item
-You may copy and distribute verbatim copies of the Program's source
-code as you receive it, in any medium, provided that you conspicuously and
-appropriately publish on each copy an appropriate copyright notice and
-disclaimer of warranty; keep intact all the notices that refer to this
-General Public License and to the absence of any warranty; and give any
-other recipients of the Program a copy of this General Public License
-along with the Program.  You may charge a fee for the physical act of
-transferring a copy.
-
-@item
-You may modify your copy or copies of the Program or any portion of
-it, and copy and distribute such modifications under the terms of Paragraph
-1 above, provided that you also do the following:
-
-@itemize @bullet
-@item
-cause the modified files to carry prominent notices stating that
-you changed the files and the date of any change; and
-
-@item
-cause the whole of any work that you distribute or publish, that
-in whole or in part contains the Program or any part thereof, either
-with or without modifications, to be licensed at no charge to all
-third parties under the terms of this General Public License (except
-that you may choose to grant warranty protection to some or all
-third parties, at your option).
-
-@item
-If the modified program normally reads commands interactively when
-run, you must cause it, when started running for such interactive use
-in the simplest and most usual way, to print or display an
-announcement including an appropriate copyright notice and a notice
-that there is no warranty (or else, saying that you provide a
-warranty) and that users may redistribute the program under these
-conditions, and telling the user how to view a copy of this General
-Public License.
-
-@item
-You may charge a fee for the physical act of transferring a
-copy, and you may at your option offer warranty protection in
-exchange for a fee.
-@end itemize
+* Concept Index::               
 
-Mere aggregation of another independent work with the Program (or its
-derivative) on a volume of a storage or distribution medium does not bring
-the other work under the scope of these terms.
-
-@item
-You may copy and distribute the Program (or a portion or derivative of
-it, under Paragraph 2) in object code or executable form under the terms of
-Paragraphs 1 and 2 above provided that you also do one of the following:
-
-@itemize @bullet
-@item
-accompany it with the complete corresponding machine-readable
-source code, which must be distributed under the terms of
-Paragraphs 1 and 2 above; or,
+@detailmenu --- The Detailed Node Listing ---
 
-@item
-accompany it with a written offer, valid for at least three
-years, to give any third party free (except for a nominal charge
-for the cost of distribution) a complete machine-readable copy of the
-corresponding source code, to be distributed under the terms of
-Paragraphs 1 and 2 above; or,
-
-@item
-accompany it with the information you received as to where the
-corresponding source code may be obtained.  (This alternative is
-allowed only for noncommercial distribution and only if you
-received the program in object code or executable form alone.)
-@end itemize
+High-Level Description of GNU @code{gperf}
 
-Source code for a work means the preferred form of the work for making
-modifications to it.  For an executable file, complete source code means
-all the source code for all modules it contains; but, as a special
-exception, it need not include source code for modules which are standard
-libraries that accompany the operating system on which the executable
-file runs, or for standard header files or definitions files that
-accompany that operating system.
+* Input Format::                Input Format to @code{gperf}
+* Output Format::               Output Format for Generated C Code with @code{gperf}
+* Binary Strings::              Use of NUL characters
 
-@item
-You may not copy, modify, sublicense, distribute or transfer the
-Program except as expressly provided under this General Public License.
-Any attempt otherwise to copy, modify, sublicense, distribute or transfer
-the Program is void, and will automatically terminate your rights to use
-the Program under this License.  However, parties who have received
-copies, or rights to use copies, from you under this General Public
-License will not have their licenses terminated so long as such parties
-remain in full compliance.
+Input Format to @code{gperf}
 
-@item
-By copying, distributing or modifying the Program (or any work based
-on the Program) you indicate your acceptance of this license to do so,
-and all its terms and conditions.
+* Declarations::                @code{struct} Declarations and C Code Inclusion.
+* Keywords::                    Format for Keyword Entries.
+* Functions::                   Including Additional C Functions.
 
-@item
-Each time you redistribute the Program (or any work based on the
-Program), the recipient automatically receives a license from the original
-licensor to copy, distribute or modify the Program subject to these
-terms and conditions.  You may not impose any further restrictions on the
-recipients' exercise of the rights granted herein.
+Invoking @code{gperf}
 
-@item
-The Free Software Foundation may publish revised and/or new versions
-of the General Public License from time to time.  Such new versions will
-be similar in spirit to the present version, but may differ in detail to
-address new problems or concerns.
-
-Each version is given a distinguishing version number.  If the Program
-specifies a version number of the license which applies to it and ``any
-later version'', you have the option of following the terms and conditions
-either of that version or of any later version published by the Free
-Software Foundation.  If the Program does not specify a version number of
-the license, you may choose any version ever published by the Free Software
-Foundation.
+* Input Details::               Options that affect Interpretation of the Input File
+* Output Language::             Specifying the Language for the Output Code
+* Output Details::              Fine tuning Details in the Output Code
+* Algorithmic Details::         Changing the Algorithms employed by @code{gperf}
+* Verbosity::                   Informative Output
 
-@item
-If you wish to incorporate parts of the Program into other free
-programs whose distribution conditions are different, write to the author
-to ask for permission.  For software which is copyrighted by the Free
-Software Foundation, write to the Free Software Foundation; we sometimes
-make exceptions for this.  Our decision will be guided by the two goals
-of preserving the free status of all derivatives of our free software and
-of promoting the sharing and reuse of software generally.
-
-@iftex
-@heading NO WARRANTY
-@end iftex
-@ifinfo
-@center NO WARRANTY
-@end ifinfo
-
-@item
-BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY
-FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW.  EXCEPT WHEN
-OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES
-PROVIDE THE PROGRAM ``AS IS'' WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED
-OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
-MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.  THE ENTIRE RISK AS
-TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU.  SHOULD THE
-PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING,
-REPAIR OR CORRECTION.
+@end detailmenu
+@end menu
 
-@item
-IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING WILL
-ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR
-REDISTRIBUTE THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES,
-INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES
-ARISING OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT
-LIMITED TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES
-SUSTAINED BY YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE
-WITH ANY OTHER PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN
-ADVISED OF THE POSSIBILITY OF SUCH DAMAGES.
-@end enumerate
-
-@iftex
-@heading END OF TERMS AND CONDITIONS
-@end iftex
-@ifinfo
-@center END OF TERMS AND CONDITIONS
 @end ifinfo
 
-@page
-@unnumberedsec Appendix: How to Apply These Terms to Your New Programs
-
-  If you develop a new program, and you want it to be of the greatest
-possible use to humanity, the best way to achieve this is to make it
-free software which everyone can redistribute and change under these
-terms.
-
-  To do so, attach the following notices to the program.  It is safest to
-attach them to the start of each source file to most effectively convey
-the exclusion of warranty; and each file should have at least the
-``copyright'' line and a pointer to where the full notice is found.
-
-@smallexample
-@var{one line to give the program's name and a brief idea of what it does.}
-Copyright (C) 19@var{yy}  @var{name of author}
-
-This program is free software; you can redistribute it and/or modify
-it under the terms of the GNU General Public License as published by
-the Free Software Foundation; either version 1, or (at your option)
-any later version.
-
-This program is distributed in the hope that it will be useful,
-but WITHOUT ANY WARRANTY; without even the implied warranty of
-MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
-GNU General Public License for more details.
-
-You should have received a copy of the GNU General Public License
-along with this program; if not, write to the Free Software
-Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
-@end smallexample
-
-Also add information on how to contact you by electronic and paper mail.
-
-If the program is interactive, make it output a short notice like this
-when it starts in an interactive mode:
-
-@smallexample
-Gnomovision version 69, Copyright (C) 19@var{yy} @var{name of author}
-Gnomovision comes with ABSOLUTELY NO WARRANTY; for details type `show w'.
-This is free software, and you are welcome to redistribute it
-under certain conditions; type `show c' for details.
-@end smallexample
-
-The hypothetical commands `show w' and `show c' should show the
-appropriate parts of the General Public License.  Of course, the
-commands you use may be called something other than `show w' and `show
-c'; they could even be mouse-clicks or menu items---whatever suits your
-program.
-
-You should also get your employer (if you work as a programmer) or your
-school, if any, to sign a ``copyright disclaimer'' for the program, if
-necessary.  Here a sample; alter the names:
-
-@example
-Yoyodyne, Inc., hereby disclaims all copyright interest in the
-program `Gnomovision' (a program to direct compilers to make passes
-at assemblers) written by James Hacker.
-
-@var{signature of Ty Coon}, 1 April 1989
-Ty Coon, President of Vice
-@end example
-
-That's all there is to it!
+@node Copying, Contributors, Top, Top
+@unnumbered GNU GENERAL PUBLIC LICENSE
+@include gpl.texinfo
 
 @node Contributors, Motivation, Copying, Top
 @unnumbered Contributors to GNU @code{gperf} Utility
 
 @itemize @bullet
 @item
+@cindex Bugs 
 The GNU @code{gperf} perfect hash function generator utility was
 originally written in GNU C++ by Douglas C. Schmidt.  It is now also
 available in a highly-portable ``old-style'' C version.  The general
@@ -432,7 +154,8 @@ Bostic's algorithm written in C, and distributed to net.sources around
 1984.  The current program is a heavily modified, enhanced, and extended
 implementation of Keith's basic idea, created at the University of
 California, Irvine.  Bugs, patches, and suggestions should be reported
-to @code{<bug-gnu-utils@@gnu.org>} and @code{<schmidt@@ics.uci.edu>}.
+to both @code{<bug-gnu-utils@@gnu.org>} and
+@code{<gperf-bugs@@lists.sourceforge.net>}.
 
 @item
 Special thanks is extended to Michael Tiemann and Doug Lea, for
@@ -441,35 +164,36 @@ creation.
 
 In addition, Adam de Boor and Nels Olson provided many tips and insights
 that greatly helped improve the quality and functionality of @code{gperf}.
+
+@item
+A testsuite was added by Bruno Haible. He also rewrote the output
+routines for better reliability.
 @end itemize
 
 @node Motivation, Search Structures, Contributors, Top
 @chapter Introduction
 
 @code{gperf} is a perfect hash function generator written in C++.  It
-transforms an @var{n} element user-specified keyword set @var{W} into
-a perfect hash function @var{F}.  @var{F} uniquely maps keywords in
-@var{W} onto the range 0..@var{k}, where @var{k >= n}.  If
-@var{k = n} then @var{F} is a @emph{minimal} perfect hash function.
+transforms an @var{n} element user-specified keyword set @var{W} into a
+perfect hash function @var{F}.  @var{F} uniquely maps keywords in
+@var{W} onto the range 0..@var{k}, where @var{k} >= @var{n}.  If @var{k}
+= @var{n} then @var{F} is a @emph{minimal} perfect hash function.
 @code{gperf} generates a 0..@var{k} element static lookup table and a
 pair of C functions.  These functions determine whether a given
-character string @var{s} occurs in @var{W}, using at most one probe
-into the lookup table.
+character string @var{s} occurs in @var{W}, using at most one probe into
+the lookup table.
 
 @code{gperf} currently generates the reserved keyword recognizer for
 lexical analyzers in several production and research compilers and
 language processing tools, including GNU C, GNU C++, GNU Pascal, GNU
 Modula 3, and GNU indent.  Complete C++ source code for @code{gperf} is
-available via anonymous ftp from @code{ics.uci.edu} and
-@code{ftp.santafe.edu}.  @code{gperf} was also distributed along with
-the GNU libg++ library for several years.  A highly portable,
-functionally equivalent K&R C version of @code{gperf} is archived in
-comp.sources.unix, volume 20.  Finally, a paper describing
-@code{gperf}'s design and implementation in greater detail is available
-in the Second USENIX C++ Conference proceedings.
+available via anonymous ftp from @code{ftp://ftp.gnu.org/pub/gnu/gperf/}.
+A paper describing @code{gperf}'s design and implementation in greater
+detail is available in the Second USENIX C++ Conference proceedings.
 
 @node Search Structures, Description, Motivation, Top
 @chapter Static search structures and GNU @code{gperf}
+@cindex Static search structure
 
 A @dfn{static search structure} is an Abstract Data Type with certain
 fundamental operations, e.g., @emph{initialize}, @emph{insert},
@@ -495,7 +219,7 @@ proportional to log @var{n}.  Conversely, hash table implementations
 often locate a table entry in constant time, but typically impose
 additional memory overhead and exhibit poor worst case performance.
 
-
+@cindex Minimal perfect hash functions
 @emph{Minimal perfect hash functions} provide an optimal solution for a
 particular class of static search sets.  A minimal perfect hash
 function is defined by two properties:
@@ -539,8 +263,9 @@ efficiently identify their respective reserved keywords.
 @chapter High-Level Description of GNU @code{gperf}
 
 @menu
-* Input Format::		Input Format to @code{gperf}
-* Output Format::		Output Format for Generated C Code with @code{gperf}
+* Input Format::                Input Format to @code{gperf}
+* Output Format::               Output Format for Generated C Code with @code{gperf}
+* Binary Strings::              Use of NUL characters
 @end menu
 
 The perfect hash function generator @code{gperf} reads a set of
@@ -571,7 +296,10 @@ Experimentation is the key to getting the most from @code{gperf}.
 
 @node Input Format, Output Format, Description, Description
 @section Input Format to @code{gperf}
-
+@cindex Format
+@cindex Declaration section
+@cindex Keywords section
+@cindex Functions section
 You can control the input keyfile format by varying certain command-line
 arguments, in particular the @samp{-t} option.  The input's appearance
 is similar to GNU utilities @code{flex} and @code{bison} (or UNIX
@@ -588,14 +316,14 @@ functions
 @end group
 @end example
 
-@emph{Unlike} @code{flex} or @code{bison}, all sections of @code{gperf}'s input
-are optional.  The following sections describe the input format for each
-section.
+@emph{Unlike} @code{flex} or @code{bison}, all sections of
+@code{gperf}'s input are optional.  The following sections describe the
+input format for each section.
 
 @menu
-* Declarations::		@code{struct} Declarations and C Code Inclusion.
-* Keywords::			Format for Keyword Entries.
-* Functions::			Including Additional C Functions.
+* Declarations::                @code{struct} Declarations and C Code Inclusion.
+* Keywords::                    Format for Keyword Entries.
+* Functions::                   Including Additional C Functions.
 @end menu
 
 @node Declarations, Keywords, Input Format, Input Format
@@ -606,11 +334,11 @@ arbitrary C declarations and definitions, as well as provisions for
 providing a user-supplied @code{struct}.  If the @samp{-t} option
 @emph{is} enabled, you @emph{must} provide a C @code{struct} as the last
 component in the declaration section from the keyfile file.  The first
-field in this struct must be a @code{char *} identifier called @samp{name},
-although it is possible to modify this field's name with the @samp{-K}
-option described below.
+field in this struct must be a @code{char *} or @code{const char *}
+identifier called @samp{name}, although it is possible to modify this
+field's name with the @samp{-K} option described below.
 
-Here is simple example, using months of the year and their attributes as
+Here is a simple example, using months of the year and their attributes as
 input:
 
 @example
@@ -632,15 +360,18 @@ december, 12, 31, 31
 @end group
 @end example
 
-Separating the @code{struct} declaration from the list of key words and
-other fields are a pair of consecutive percent signs, @code{%%},
+@cindex @samp{%%}
+Separating the @code{struct} declaration from the list of keywords and
+other fields are a pair of consecutive percent signs, @samp{%%},
 appearing left justified in the first column, as in the UNIX utility
 @code{lex}.
 
+@cindex @samp{%@{}
+@cindex @samp{%@}}
 Using a syntax similar to GNU utilities @code{flex} and @code{bison}, it
 is possible to directly include C source text and comments verbatim into
 the generated output file.  This is accomplished by enclosing the region
-inside left-justified surrounding @code{%@{}, @code{%@}} pairs.  Here is
+inside left-justified surrounding @samp{%@{}, @samp{%@}} pairs.  Here is
 an input fragment based on the previous example that illustrates this
 feature:
 
@@ -682,15 +413,18 @@ in the first column is considered a comment.  Everything following the
 @samp{#} is ignored, up to and including the following newline.
 
 The first field of each non-comment line is always the key itself.  It
-should be given as a simple name, i.e., without surrounding
-string quotation marks, and be left-justified flush against the first
-column.  In this context, a ``field'' is considered to extend up to, but
+can be given in two ways: as a simple name, i.e., without surrounding
+string quotation marks, or as a string enclosed in double-quotes, in
+C syntax, possibly with backslash escapes like @code{\"} or @code{\234}
+or @code{\xa8}. In either case, it must start right at the beginning
+of the line, without leading whitespace.
+In this context, a ``field'' is considered to extend up to, but
 not include, the first blank, comma, or newline.  Here is a simple
 example taken from a partial list of C reserved words:
 
 @example
 @group
-# These are a few C reserved words, see the c.@code{gperf} file 
+# These are a few C reserved words, see the c.gperf file 
 # for a complete list of ANSI C reserved words.
 unsigned
 sizeof
@@ -704,7 +438,7 @@ return
 @end group
 @end example
 
-Note that unlike @code{flex} or @code{bison} the first @code{%%} marker
+Note that unlike @code{flex} or @code{bison} the first @samp{%%} marker
 may be elided if the declaration section is empty.
 
 Additional fields may optionally follow the leading keyword.  Fields
@@ -720,89 +454,116 @@ one contain keyword attributes.
 
 The optional third section also corresponds closely with conventions
 found in @code{flex} and @code{bison}.  All text in this section,
-starting at the final @code{%%} and extending to the end of the input
+starting at the final @samp{%%} and extending to the end of the input
 file, is included verbatim into the generated output file.  Naturally,
 it is your responsibility to ensure that the code contained in this
 section is valid C.
 
-@node Output Format,  , Input Format, Description
+@node Output Format, Binary Strings, Input Format, Description
 @section Output Format for Generated C Code with @code{gperf}
+@cindex hash table
 
-Several options control how the generated C code appears on the standard
-output.  Two C function are generated.  They are called @code{hash} and
-@code{in_word_set}, although you may modify the name for
-@code{in_word_set} with a command-line option.  Both functions require
-two arguments, a string, @code{char *} @var{str}, and a length
-parameter, @code{int} @var{len}.  Their default function prototypes are
-as follows:
-
-@example
-@group
-static int hash (char *str, int len);
-int in_word_set (char *str, int len);
-@end group
-@end example
+Several options control how the generated C code appears on the standard 
+output.  Two C function are generated.  They are called @code{hash} and 
+@code{in_word_set}, although you may modify their names with a command-line 
+option.  Both functions require two arguments, a string, @code{char *} 
+@var{str}, and a length parameter, @code{int} @var{len}.  Their default 
+function prototypes are as follows:
 
+@deftypefun {unsigned int} hash (const char * @var{str}, unsigned int @var{len})
 By default, the generated @code{hash} function returns an integer value
 created by adding @var{len} to several user-specified @var{str} key
 positions indexed into an @dfn{associated values} table stored in a
 local static array.  The associated values table is constructed
-internally by @code{gperf} and later output as a static local C array called
-@var{hash_table}; its meaning and properties are described below.
-@xref{Implementation}. The relevant key positions are specified via the
-@samp{-k} option when running @code{gperf}, as detailed in the @emph{Options}
-section below. @xref{Options}.
-
-Two options, @samp{-g} (assume you are compiling with GNU C and its
-@code{inline} feature) and @samp{-a} (assume ANSI C-style function
-prototypes), alter the content of both the generated @code{hash} and
-@code{in_word_set} routines.  However, function @code{in_word_set} may
-be modified more extensively, in response to your option settings.  The
-options that affect the @code{in_word_set} structure are:
+internally by @code{gperf} and later output as a static local C array
+called @samp{hash_table}; its meaning and properties are described below
+(@pxref{Implementation}). The relevant key positions are specified via
+the @samp{-k} option when running @code{gperf}, as detailed in the
+@emph{Options} section below(@pxref{Options}).
+@end deftypefun
+
+@deftypefun {} in_word_set (const char * @var{str}, unsigned int @var{len})
+If @var{str} is in the keyword set, returns a pointer to that
+keyword. More exactly, if the option @samp{-t} was given, it returns
+a pointer to the matching keyword's structure. Otherwise it returns
+@code{NULL}.
+@end deftypefun
+
+If the option @samp{-c} is not used, @var{str} must be a NUL terminated
+string of exactly length @var{len}. If @samp{-c} is used, @var{str} must
+simply be an array of @var{len} characters and does not need to be NUL
+terminated.
+
+The code generated for these two functions is affected by the following
+options:
 
-@itemize @bullet
 @table @samp
 @item -t
+@itemx --struct-type
 Make use of the user-defined @code{struct}.
 
-@item -S @var{total switch statements}
+@item -S @var{total-switch-statements}
+@itemx --switch=@var{total-switch-statements}
+@cindex @code{switch}
 Generate 1 or more C @code{switch} statement rather than use a large,
 (and potentially sparse) static array.  Although the exact time and
 space savings of this approach vary according to your C compiler's
 degree of optimization, this method often results in smaller and faster
 code.
 @end table
-@end itemize
-
-If the @samp{-t} and @samp{-S} options are omitted, the
-default action is to generate a @code{char *} array containing the keys,
-together with additional null strings used for padding the array.  By
-experimenting with the various input and output options, and timing the
-resulting C code, you can determine the best option choices for
-different keyword set characteristics.
 
-@node Options, Input Details, Description, Top
-@chapter Options to the @code{gperf} Utility
+If the @samp{-t} and @samp{-S} options are omitted, the default action
+is to generate a @code{char *} array containing the keys, together with
+additional null strings used for padding the array.  By experimenting
+with the various input and output options, and timing the resulting C
+code, you can determine the best option choices for different keyword
+set characteristics.
+
+@node Binary Strings,  , Output Format, Description
+@section Use of NUL characters
+@cindex NUL
+
+By default, the code generated by @code{gperf} operates on zero
+terminated strings, the usual representation of strings in C. This means
+that the keywords in the input file must not contain NUL characters,
+and the @var{str} argument passed to @code{hash} or @code{in_word_set}
+must be NUL terminated and have exactly length @var{len}.
+
+If option @samp{-c} is used, then the @var{str} argument does not need
+to be NUL terminated. The code generated by @code{gperf} will only
+access the first @var{len}, not @var{len+1}, bytes starting at @var{str}.
+However, the keywords in the input file still must not contain NUL
+characters.
+
+If option @samp{-l} is used, then the hash table performs binary
+comparison. The keywords in the input file may contain NUL characters,
+written in string syntax as @code{\000} or @code{\x00}, and the code
+generated by @code{gperf} will treat NUL like any other character.
+Also, in this case the @samp{-c} option is ignored.
+
+@node Options, Bugs, Description, Top
+@chapter Invoking @code{gperf}
 
 There are @emph{many} options to @code{gperf}.  They were added to make
 the program more convenient for use with real applications.  ``On-line''
-help is readily available via the @samp{-h} option.  Here is the complete
-list of options.
+help is readily available via the @samp{-h} option.  Here is the
+complete list of options.
 
 @menu
-* Input Details::          Options that affect Interpretation of the Input File
-* Output Language::        Specifying the Language for the Output Code
-* Output Details::         Fine tuning Details in the Output Code
-* Algorithmic Details::    Changing the Algorithms employed by @code{gperf}
-* Verbosity::              Informative Output
+* Input Details::               Options that affect Interpretation of the Input File
+* Output Language::             Specifying the Language for the Output Code
+* Output Details::              Fine tuning Details in the Output Code
+* Algorithmic Details::         Changing the Algorithms employed by @code{gperf}
+* Verbosity::                   Informative Output
 @end menu
 
 @node Input Details, Output Language, Options, Options
 @section Options that affect Interpretation of the Input File
 
-@itemize @bullet
 @table @samp
-@item -e @var{keyword delimiter list}
+@item -e @var{keyword-delimiter-list}
+@itemx --delimiters=@var{keyword-delimiter-list}
+@cindex Delimiters
 Allows the user to provide a string containing delimiters used to
 separate keywords from their attributes.  The default is ",\n".  This
 option is essential if you want to use keywords that have embedded
@@ -810,25 +571,24 @@ commas or newlines.  One useful trick is to use -e'TAB', where TAB is
 the literal tab character.
 
 @item -t
+@itemx --struct-type
 Allows you to include a @code{struct} type declaration for generated
-code.  Any text before a pair of consecutive %% is consider part of the
-type declaration.  Key words and additional fields may follow this, one
-group of fields per line.  A set of examples for generating perfect hash
-tables and functions for Ada, C, and G++, Pascal, and Modula 2 and 3
-reserved words are distributed with this release.
+code.  Any text before a pair of consecutive @samp{%%} is considered
+part of the type declaration.  Keywords and additional fields may follow
+this, one group of fields per line.  A set of examples for generating
+perfect hash tables and functions for Ada, C, C++, Pascal, Modula 2,
+Modula 3 and JavaScript reserved words are distributed with this release.
 @end table
-@end itemize
 
 @node Output Language, Output Details, Input Details, Options
 @section Options to specify the Language for the Output Code
 
-@itemize @bullet
 @table @samp
-@item -L @var{generated language name}
+@item -L @var{generated-language-name}
+@itemx --language=@var{generated-language-name}
 Instructs @code{gperf} to generate code in the language specified by the
 option's argument.  Languages handled are currently:
 
-@itemize @bullet
 @table @samp
 @item KR-C
 Old-style K&R C. This language is understood by old-style C compilers and
@@ -846,7 +606,6 @@ ANSI C. This language is understood by ANSI C compilers and C++ compilers.
 @item C++
 C++. This language is understood by C++ compilers.
 @end table
-@end itemize
 
 The default is C.
 
@@ -858,77 +617,100 @@ This option is supported for compatibility with previous releases of
 This option is supported for compatibility with previous releases of
 @code{gperf}. It does not do anything.
 @end table
-@end itemize
 
 @node Output Details, Algorithmic Details, Output Language, Options
 @section Options for fine tuning Details in the Output Code
 
-@itemize @bullet
 @table @samp
-@item -K @var{key name}
+@item -K @var{key-name}
+@itemx --slot-name=@var{key-name}
+@cindex Slot name
 This option is only useful when option @samp{-t} has been given.
 By default, the program assumes the structure component identifier for
 the keyword is @samp{name}.  This option allows an arbitrary choice of
 identifier for this component, although it still must occur as the first
 field in your supplied @code{struct}.
 
-@item -H @var{hash function name}
+@item -F @var{initializers}
+@itemx --initializer-suffix=@var{initializers}
+@cindex Initializers
+This option is only useful when option @samp{-t} has been given.
+It permits to specify initializers for the structure members following
+@var{key name} in empty hash table entries.  The list of initializers
+should start with a comma.  By default, the emitted code will
+zero-initialize structure members following @var{key name}.
+
+@item -H @var{hash-function-name}
+@itemx --hash-fn-name=@var{hash-function-name}
 Allows you to specify the name for the generated hash function.  Default
-name is @samp{hash}.  This option permits the use of two hash tables in the
-same file.
+name is @samp{hash}.  This option permits the use of two hash tables in
+the same file.
 
-@item -N @var{lookup function name}
+@item -N @var{lookup-function-name}
+@itemx --lookup-fn-name=@var{lookup-function-name}
 Allows you to specify the name for the generated lookup function.
-Default name is @samp{in_word_set}.  This option permits completely automatic
-generation of perfect hash functions, especially when multiple generated
-hash functions are used in the same application.
-
-@item -Z @var{class name}
-This option is only useful when option @samp{-L C++} has been given.
-It allows you to specify the name of generated C++ class.  Default name is
+Default name is @samp{in_word_set}.  This option permits completely
+automatic generation of perfect hash functions, especially when multiple
+generated hash functions are used in the same application.
+
+@item -Z @var{class-name}
+@itemx --class-name=@var{class-name}
+@cindex Class name
+This option is only useful when option @samp{-L C++} has been given.  It
+allows you to specify the name of generated C++ class.  Default name is
 @code{Perfect_Hash}.
 
 @item -7
+@itemx --seven-bit
 This option specifies that all strings that will be passed as arguments
 to the generated hash function and the generated lookup function will
 solely consist of 7-bit ASCII characters (characters in the range 0..127).
 (Note that the ANSI C functions @code{isalnum} and @code{isgraph} do
 @emph{not} guarantee that a character is in this range. Only an explicit
 test like @samp{c >= 'A' && c <= 'Z'} guarantees this.) This was the
-default in earlier versions of @code{gperf}; now the default is to assume
-8-bit characters.
+default in versions of @code{gperf} earlier than 2.7; now the default is
+to assume 8-bit characters.
 
 @item -c
+@itemx --compare-strncmp
 Generates C code that uses the @code{strncmp} function to perform
 string comparisons.  The default action is to use @code{strcmp}.
 
 @item -C
+@itemx --readonly-tables
 Makes the contents of all generated lookup tables constant, i.e.,
 ``readonly''.  Many compilers can generate more efficient code for this
 by putting the tables in readonly memory.
 
 @item -E
+@itemx  --enum
 Define constant values using an enum local to the lookup function rather
 than with #defines.  This also means that different lookup functions can
 reside in the same file.  Thanks to James Clark @code{<jjc@@ai.mit.edu>}.
 
 @item -I
+@itemx --includes
 Include the necessary system include file, @code{<string.h>}, at the
 beginning of the code.  By default, this is not done; the user must
 include this header file himself to allow compilation of the code.
 
 @item -G
+@itemx --global
 Generate the static table of keywords as a static global variable,
 rather than hiding it inside of the lookup function (which is the
 default behavior).
 
-@item -W @var{hash table array name}
+@item -W @var{hash-table-array-name}
+@itemx --word-array-name=@var{hash-table-array-name}
+@cindex Array name
 Allows you to specify the name for the generated array containing the
 hash table.  Default name is @samp{wordlist}.  This option permits the
 use of two hash tables in the same file, even when the option @samp{-G}
 is given.
 
-@item -S @var{total switch statements}
+@item -S @var{total-switch-statements}
+@itemx --switch=@var{total-switch-statements}
+@cindex @code{switch}
 Causes the generated C code to use a @code{switch} statement scheme,
 rather than an array lookup table.  This can lead to a reduction in both
 time and space requirements for some keyfiles.  The argument to this
@@ -940,6 +722,7 @@ correctly generate code for large @code{switch} statements. This option
 was inspired in part by Keith Bostic's original C program.
 
 @item -T
+@itemx --omit-struct-type
 Prevents the transfer of the type declaration to the output file.  Use
 this option if the type is already defined elsewhere.
 
@@ -947,14 +730,13 @@ this option if the type is already defined elsewhere.
 This option is supported for compatibility with previous releases of
 @code{gperf}. It does not do anything.
 @end table
-@end itemize
 
 @node Algorithmic Details, Verbosity, Output Details, Options
 @section Options for changing the Algorithms employed by @code{gperf}
 
-@itemize @bullet
 @table @samp
 @item -k @var{keys}
+@itemx --key-positions=@var{keys}
 Allows selection of the character key positions used in the keywords'
 hash function. The allowable choices range between 1-126, inclusive.
 The positions are separated by commas, e.g., @samp{-k 9,4,13,14};
@@ -973,6 +755,7 @@ selected key positions exceeding the key length are simply not
 referenced in the hash function.
 
 @item -l
+@itemx --compare-strlen
 Compare key lengths before trying a string comparison.  This might cut
 down on the number of string comparisons made during the lookup, since
 keys with different lengths are never compared via @code{strcmp}.
@@ -980,8 +763,11 @@ However, using @samp{-l} might greatly increase the size of the
 generated C code if the lookup table range is large (which implies that
 the switch option @samp{-S} is not enabled), since the length table
 contains as many elements as there are entries in the lookup table.
+This option is mandatory for binary comparisons (@pxref{Binary Strings}).
 
 @item -D
+@itemx --duplicates
+@cindex Duplicates
 Handle keywords whose key position sets hash to duplicate values.
 Duplicate hash values occur for two reasons:
 
@@ -991,11 +777,12 @@ Since @code{gperf} does not backtrack it is possible for it to process
 all your input keywords without finding a unique mapping for each word.
 However, frequently only a very small number of duplicates occur, and 
 the majority of keys still require one probe into the table.
+
 @item
 Sometimes a set of keys may have the same names, but possess different
-attributes.  With the -D option @code{gperf} treats all these keys as part of
-an equivalence class and generates a perfect hash function with multiple
-comparisons for duplicate keys.  It is up to you to completely
+attributes.  With the -D option @code{gperf} treats all these keys as
+part of an equivalence class and generates a perfect hash function with
+multiple comparisons for duplicate keys.  It is up to you to completely
 disambiguate the keywords by modifying the generated C code.  However,
 @code{gperf} helps you out by organizing the output.
 @end itemize
@@ -1006,33 +793,39 @@ Using this option usually means that the generated hash function is no
 longer perfect.  On the other hand, it permits @code{gperf} to work on
 keyword sets that it otherwise could not handle.
 
-@item -f @var{iteration amount}
-Generate the perfect hash function ``fast''.  This decreases @code{gperf}'s
-running time at the cost of minimizing generated table-size.  The
-iteration amount represents the number of times to iterate when
-resolving a collision.  `0' means iterate by the number of keywords.
-This option is probably most useful when used in conjunction with options
-@samp{-D} and/or @samp{-S} for @emph{large} keyword sets.
-
-@item -i @var{initial value}
+@item -f @var{iteration-amount}
+@itemx --fast=@var{iteration-amount}
+Generate the perfect hash function ``fast''.  This decreases
+@code{gperf}'s running time at the cost of minimizing generated
+table-size.  The iteration amount represents the number of times to
+iterate when resolving a collision.  `0' means iterate by the number of
+keywords.  This option is probably most useful when used in conjunction
+with options @samp{-D} and/or @samp{-S} for @emph{large} keyword sets.
+
+@item -i @var{initial-value}
+@itemx --initial-asso=@var{initial-value}
 Provides an initial @var{value} for the associate values array.  Default
 is 0.  Increasing the initial value helps inflate the final table size,
 possibly leading to more time efficient keyword lookups.  Note that this
 option is not particularly useful when @samp{-S} is used.  Also,
-@samp{-i} is overriden when the @samp{-r} option is used.
+@samp{-i} is overridden when the @samp{-r} option is used.
 
-@item -j @var{jump value}
-Affects the ``jump value'', i.e., how far to advance the
-associated character value upon collisions.  @var{Jump value} is rounded
-up to an odd number, the default is 5.  If the @var{jump value} is 0
-@code{gperf} jumps by random amounts.
+@item -j @var{jump-value}
+@itemx --jump=@var{jump-value}
+@cindex Jump value
+Affects the ``jump value'', i.e., how far to advance the associated
+character value upon collisions.  @var{Jump-value} is rounded up to an
+odd number, the default is 5.  If the @var{jump-value} is 0 @code{gperf}
+jumps by random amounts.
 
 @item -n
+@itemx --no-strlen
 Instructs the generator not to include the length of a keyword when
 computing its hash value.  This may save a few assembly instructions in
 the generated lookup table.
 
 @item -o
+@itemx --occurrence-sort
 Reorders the keywords by sorting the keywords so that frequently
 occuring key position set components appear first.  A second reordering
 pass follows so that keys with ``already determined values'' are placed
@@ -1042,12 +835,13 @@ produce more minimal perfect hash functions.  The reason for this is
 that the reordering helps prune the search time by handling inevitable
 collisions early in the search process.  On the other hand, if the
 number of keywords is @emph{very} large using @samp{-o} may
-@emph{increase} @code{gperf}'s execution time, since collisions will begin
-earlier and continue throughout the remainder of keyword processing.
-See Cichelli's paper from the January 1980 Communications of the ACM for
-details.
+@emph{increase} @code{gperf}'s execution time, since collisions will
+begin earlier and continue throughout the remainder of keyword
+processing.  See Cichelli's paper from the January 1980 Communications
+of the ACM for details.
 
 @item -r
+@itemx --random
 Utilizes randomness to initialize the associated values table.  This
 frequently generates solutions faster than using deterministic
 initialization (which starts all associated values at 0).  Furthermore,
@@ -1056,6 +850,7 @@ table.  If @code{gperf} has difficultly with a certain keyword set try using
 @samp{-r} or @samp{-D}.
 
 @item -s @var{size-multiple}
+@itemx --size-multiple=@var{size-multiple}
 Affects the size of the generated hash table.  The numeric argument for
 this option indicates ``how many times larger or smaller'' the maximum
 associated value range should be, in relationship to the number of keys.
@@ -1084,21 +879,22 @@ heuristic.  In particular, setting this value too high slows down
 of values.  Judicious use of the @samp{-f} option helps alleviate this
 overhead, however.
 @end table
-@end itemize
 
-@node Verbosity, Bugs, Algorithmic Details, Options
+@node Verbosity,  , Algorithmic Details, Options
 @section Informative Output
 
-@itemize @bullet
 @table @samp
 @item -h
+@itemx --help
 Prints a short summary on the meaning of each program option.  Aborts
 further program execution.
 
 @item -v
+@itemx --version
 Prints out the current version number.
 
 @item -d
+@itemx --debug
 Enables the debugging option.  This produces verbose diagnostics to
 ``standard error'' when @code{gperf} is executing.  It is useful both for
 maintaining the program and for determining whether a given set of
@@ -1106,9 +902,8 @@ options is actually speeding up the search for a solution.  Some useful
 information is dumped at the end of the program when the @samp{-d}
 option is enabled.
 @end table
-@end itemize
 
-@node Bugs, Projects, Verbosity, Top
+@node Bugs, Projects, Options, Top
 @chapter Known Bugs and Limitations with @code{gperf}
 
 The following are some limitations with the current release of
@@ -1200,7 +995,7 @@ determining both the minimum and maximum elements in a list.
 
 @page
 
-@node Bibliography,  , Implementation, Top
+@node Bibliography, Concept Index, Implementation, Top
 @chapter Bibliography
 
 [1] Chang, C.C.: @i{A Scheme for Constructing Ordered Minimal Perfect
@@ -1247,5 +1042,10 @@ Retrieving Method for Static Sets} Communications of the ACM, 20
 [14] Tiemann, Michael D. @i{User's Guide to GNU C++} Free Software
 Foundation, 1989.
 
+@node Concept Index,  , Bibliography, Top
+@unnumbered Concept Index
+
+@printindex cp
+
 @contents
 @bye
diff --git a/contrib/gperf/doc/gpl.texinfo b/contrib/gperf/doc/gpl.texinfo
new file mode 100644
index 000000000000..be2dc253c470
--- /dev/null
+++ b/contrib/gperf/doc/gpl.texinfo
@@ -0,0 +1,398 @@
+@c This GPL is meant to be included from other files.
+@c To format a standalone GPL, use license.texi.
+
+@center Version 2, June 1991
+
+@display
+Copyright @copyright{} 1989, 1991 Free Software Foundation, Inc.,
+59 Temple Place, Suite 330, Boston, MA 02111-1307, USA.
+
+Everyone is permitted to copy and distribute verbatim copies
+of this license document, but changing it is not allowed.
+@end display
+
+@unnumberedsec Preamble
+
+  The licenses for most software are designed to take away your
+freedom to share and change it.  By contrast, the GNU General Public
+License is intended to guarantee your freedom to share and change free
+software---to make sure the software is free for all its users.  This
+General Public License applies to most of the Free Software
+Foundation's software and to any other program whose authors commit to
+using it.  (Some other Free Software Foundation software is covered by
+the GNU Library General Public License instead.)  You can apply it to
+your programs, too.
+
+  When we speak of free software, we are referring to freedom, not
+price.  Our General Public Licenses are designed to make sure that you
+have the freedom to distribute copies of free software (and charge for
+this service if you wish), that you receive source code or can get it
+if you want it, that you can change the software or use pieces of it
+in new free programs; and that you know you can do these things.
+
+  To protect your rights, we need to make restrictions that forbid
+anyone to deny you these rights or to ask you to surrender the rights.
+These restrictions translate to certain responsibilities for you if you
+distribute copies of the software, or if you modify it.
+
+  For example, if you distribute copies of such a program, whether
+gratis or for a fee, you must give the recipients all the rights that
+you have.  You must make sure that they, too, receive or can get the
+source code.  And you must show them these terms so they know their
+rights.
+
+  We protect your rights with two steps: (1) copyright the software, and
+(2) offer you this license which gives you legal permission to copy,
+distribute and/or modify the software.
+
+  Also, for each author's protection and ours, we want to make certain
+that everyone understands that there is no warranty for this free
+software.  If the software is modified by someone else and passed on, we
+want its recipients to know that what they have is not the original, so
+that any problems introduced by others will not reflect on the original
+authors' reputations.
+
+  Finally, any free program is threatened constantly by software
+patents.  We wish to avoid the danger that redistributors of a free
+program will individually obtain patent licenses, in effect making the
+program proprietary.  To prevent this, we have made it clear that any
+patent must be licensed for everyone's free use or not licensed at all.
+
+  The precise terms and conditions for copying, distribution and
+modification follow.
+
+@iftex
+@unnumberedsec TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
+@end iftex
+@ifinfo
+@center TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
+@end ifinfo
+
+@enumerate 0
+@item
+This License applies to any program or other work which contains
+a notice placed by the copyright holder saying it may be distributed
+under the terms of this General Public License.  The ``Program'', below,
+refers to any such program or work, and a ``work based on the Program''
+means either the Program or any derivative work under copyright law:
+that is to say, a work containing the Program or a portion of it,
+either verbatim or with modifications and/or translated into another
+language.  (Hereinafter, translation is included without limitation in
+the term ``modification''.)  Each licensee is addressed as ``you''.
+
+Activities other than copying, distribution and modification are not
+covered by this License; they are outside its scope.  The act of
+running the Program is not restricted, and the output from the Program
+is covered only if its contents constitute a work based on the
+Program (independent of having been made by running the Program).
+Whether that is true depends on what the Program does.
+
+@item
+You may copy and distribute verbatim copies of the Program's
+source code as you receive it, in any medium, provided that you
+conspicuously and appropriately publish on each copy an appropriate
+copyright notice and disclaimer of warranty; keep intact all the
+notices that refer to this License and to the absence of any warranty;
+and give any other recipients of the Program a copy of this License
+along with the Program.
+
+You may charge a fee for the physical act of transferring a copy, and
+you may at your option offer warranty protection in exchange for a fee.
+
+@item
+You may modify your copy or copies of the Program or any portion
+of it, thus forming a work based on the Program, and copy and
+distribute such modifications or work under the terms of Section 1
+above, provided that you also meet all of these conditions:
+
+@enumerate a
+@item
+You must cause the modified files to carry prominent notices
+stating that you changed the files and the date of any change.
+
+@item
+You must cause any work that you distribute or publish, that in
+whole or in part contains or is derived from the Program or any
+part thereof, to be licensed as a whole at no charge to all third
+parties under the terms of this License.
+
+@item
+If the modified program normally reads commands interactively
+when run, you must cause it, when started running for such
+interactive use in the most ordinary way, to print or display an
+announcement including an appropriate copyright notice and a
+notice that there is no warranty (or else, saying that you provide
+a warranty) and that users may redistribute the program under
+these conditions, and telling the user how to view a copy of this
+License.  (Exception: if the Program itself is interactive but
+does not normally print such an announcement, your work based on
+the Program is not required to print an announcement.)
+@end enumerate
+
+These requirements apply to the modified work as a whole.  If
+identifiable sections of that work are not derived from the Program,
+and can be reasonably considered independent and separate works in
+themselves, then this License, and its terms, do not apply to those
+sections when you distribute them as separate works.  But when you
+distribute the same sections as part of a whole which is a work based
+on the Program, the distribution of the whole must be on the terms of
+this License, whose permissions for other licensees extend to the
+entire whole, and thus to each and every part regardless of who wrote it.
+
+Thus, it is not the intent of this section to claim rights or contest
+your rights to work written entirely by you; rather, the intent is to
+exercise the right to control the distribution of derivative or
+collective works based on the Program.
+
+In addition, mere aggregation of another work not based on the Program
+with the Program (or with a work based on the Program) on a volume of
+a storage or distribution medium does not bring the other work under
+the scope of this License.
+
+@item
+You may copy and distribute the Program (or a work based on it,
+under Section 2) in object code or executable form under the terms of
+Sections 1 and 2 above provided that you also do one of the following:
+
+@enumerate a
+@item
+Accompany it with the complete corresponding machine-readable
+source code, which must be distributed under the terms of Sections
+1 and 2 above on a medium customarily used for software interchange; or,
+
+@item
+Accompany it with a written offer, valid for at least three
+years, to give any third party, for a charge no more than your
+cost of physically performing source distribution, a complete
+machine-readable copy of the corresponding source code, to be
+distributed under the terms of Sections 1 and 2 above on a medium
+customarily used for software interchange; or,
+
+@item
+Accompany it with the information you received as to the offer
+to distribute corresponding source code.  (This alternative is
+allowed only for noncommercial distribution and only if you
+received the program in object code or executable form with such
+an offer, in accord with Subsection b above.)
+@end enumerate
+
+The source code for a work means the preferred form of the work for
+making modifications to it.  For an executable work, complete source
+code means all the source code for all modules it contains, plus any
+associated interface definition files, plus the scripts used to
+control compilation and installation of the executable.  However, as a
+special exception, the source code distributed need not include
+anything that is normally distributed (in either source or binary
+form) with the major components (compiler, kernel, and so on) of the
+operating system on which the executable runs, unless that component
+itself accompanies the executable.
+
+If distribution of executable or object code is made by offering
+access to copy from a designated place, then offering equivalent
+access to copy the source code from the same place counts as
+distribution of the source code, even though third parties are not
+compelled to copy the source along with the object code.
+
+@item
+You may not copy, modify, sublicense, or distribute the Program
+except as expressly provided under this License.  Any attempt
+otherwise to copy, modify, sublicense or distribute the Program is
+void, and will automatically terminate your rights under this License.
+However, parties who have received copies, or rights, from you under
+this License will not have their licenses terminated so long as such
+parties remain in full compliance.
+
+@item
+You are not required to accept this License, since you have not
+signed it.  However, nothing else grants you permission to modify or
+distribute the Program or its derivative works.  These actions are
+prohibited by law if you do not accept this License.  Therefore, by
+modifying or distributing the Program (or any work based on the
+Program), you indicate your acceptance of this License to do so, and
+all its terms and conditions for copying, distributing or modifying
+the Program or works based on it.
+
+@item
+Each time you redistribute the Program (or any work based on the
+Program), the recipient automatically receives a license from the
+original licensor to copy, distribute or modify the Program subject to
+these terms and conditions.  You may not impose any further
+restrictions on the recipients' exercise of the rights granted herein.
+You are not responsible for enforcing compliance by third parties to
+this License.
+
+@item
+If, as a consequence of a court judgment or allegation of patent
+infringement or for any other reason (not limited to patent issues),
+conditions are imposed on you (whether by court order, agreement or
+otherwise) that contradict the conditions of this License, they do not
+excuse you from the conditions of this License.  If you cannot
+distribute so as to satisfy simultaneously your obligations under this
+License and any other pertinent obligations, then as a consequence you
+may not distribute the Program at all.  For example, if a patent
+license would not permit royalty-free redistribution of the Program by
+all those who receive copies directly or indirectly through you, then
+the only way you could satisfy both it and this License would be to
+refrain entirely from distribution of the Program.
+
+If any portion of this section is held invalid or unenforceable under
+any particular circumstance, the balance of the section is intended to
+apply and the section as a whole is intended to apply in other
+circumstances.
+
+It is not the purpose of this section to induce you to infringe any
+patents or other property right claims or to contest validity of any
+such claims; this section has the sole purpose of protecting the
+integrity of the free software distribution system, which is
+implemented by public license practices.  Many people have made
+generous contributions to the wide range of software distributed
+through that system in reliance on consistent application of that
+system; it is up to the author/donor to decide if he or she is willing
+to distribute software through any other system and a licensee cannot
+impose that choice.
+
+This section is intended to make thoroughly clear what is believed to
+be a consequence of the rest of this License.
+
+@item
+If the distribution and/or use of the Program is restricted in
+certain countries either by patents or by copyrighted interfaces, the
+original copyright holder who places the Program under this License
+may add an explicit geographical distribution limitation excluding
+those countries, so that distribution is permitted only in or among
+countries not thus excluded.  In such case, this License incorporates
+the limitation as if written in the body of this License.
+
+@item
+The Free Software Foundation may publish revised and/or new versions
+of the General Public License from time to time.  Such new versions will
+be similar in spirit to the present version, but may differ in detail to
+address new problems or concerns.
+
+Each version is given a distinguishing version number.  If the Program
+specifies a version number of this License which applies to it and ``any
+later version'', you have the option of following the terms and conditions
+either of that version or of any later version published by the Free
+Software Foundation.  If the Program does not specify a version number of
+this License, you may choose any version ever published by the Free Software
+Foundation.
+
+@item
+If you wish to incorporate parts of the Program into other free
+programs whose distribution conditions are different, write to the author
+to ask for permission.  For software which is copyrighted by the Free
+Software Foundation, write to the Free Software Foundation; we sometimes
+make exceptions for this.  Our decision will be guided by the two goals
+of preserving the free status of all derivatives of our free software and
+of promoting the sharing and reuse of software generally.
+
+@iftex
+@vskip -@baselineskip
+@vskip -@baselineskip
+@heading NO WARRANTY
+@end iftex
+@ifinfo
+@center NO WARRANTY
+@end ifinfo
+
+@item
+BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY
+FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW.  EXCEPT WHEN
+OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES
+PROVIDE THE PROGRAM ``AS IS'' WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED
+OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
+MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.  THE ENTIRE RISK AS
+TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU.  SHOULD THE
+PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING,
+REPAIR OR CORRECTION.
+
+@item
+IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING
+WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR
+REDISTRIBUTE THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES,
+INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING
+OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED
+TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY
+YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER
+PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE
+POSSIBILITY OF SUCH DAMAGES.
+@end enumerate
+
+@iftex
+@heading END OF TERMS AND CONDITIONS
+@end iftex
+@ifinfo
+@center END OF TERMS AND CONDITIONS
+@end ifinfo
+
+@page
+@unnumberedsec How to Apply These Terms to Your New Programs
+
+  If you develop a new program, and you want it to be of the greatest
+possible use to the public, the best way to achieve this is to make it
+free software which everyone can redistribute and change under these terms.
+
+  To do so, attach the following notices to the program.  It is safest
+to attach them to the start of each source file to most effectively
+convey the exclusion of warranty; and each file should have at least
+the ``copyright'' line and a pointer to where the full notice is found.
+
+@smallexample
+@var{one line to give the program's name and an idea of what it does.}
+Copyright (C) @var{year}  @var{name of author}
+
+This program is free software; you can redistribute it and/or
+modify it under the terms of the GNU General Public License
+as published by the Free Software Foundation; either version 2
+of the License, or (at your option) any later version.
+
+This program is distributed in the hope that it will be useful,
+but WITHOUT ANY WARRANTY; without even the implied warranty of
+MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+GNU General Public License for more details.
+
+You should have received a copy of the GNU General Public License
+along with this program; if not, write to the Free Software
+Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307, USA.
+@end smallexample
+
+Also add information on how to contact you by electronic and paper mail.
+
+If the program is interactive, make it output a short notice like this
+when it starts in an interactive mode:
+
+@smallexample
+Gnomovision version 69, Copyright (C) @var{year}  @var{name of author}
+Gnomovision comes with ABSOLUTELY NO WARRANTY; for details
+type `show w'.  This is free software, and you are welcome
+to redistribute it under certain conditions; type `show c' 
+for details.
+@end smallexample
+
+The hypothetical commands @samp{show w} and @samp{show c} should show
+the appropriate parts of the General Public License.  Of course, the
+commands you use may be called something other than @samp{show w} and
+@samp{show c}; they could even be mouse-clicks or menu items---whatever
+suits your program.
+
+You should also get your employer (if you work as a programmer) or your
+school, if any, to sign a ``copyright disclaimer'' for the program, if
+necessary.  Here is a sample; alter the names:
+
+@example
+@group
+Yoyodyne, Inc., hereby disclaims all copyright
+interest in the program `Gnomovision'
+(which makes passes at compilers) written 
+by James Hacker.
+
+@var{signature of Ty Coon}, 1 April 1989
+Ty Coon, President of Vice
+@end group
+@end example
+
+This General Public License does not permit incorporating your program into
+proprietary programs.  If your program is a subroutine library, you may
+consider it more useful to permit linking proprietary applications with the
+library.  If this is what you want to do, use the GNU Library General
+Public License instead of this License.
diff --git a/contrib/gperf/doc/help2man b/contrib/gperf/doc/help2man
new file mode 100755
index 000000000000..2c8164789aa7
--- /dev/null
+++ b/contrib/gperf/doc/help2man
@@ -0,0 +1,513 @@
+#!/usr/bin/perl -w
+
+# Generate a short man page from --help and --version output.
+# Copyright © 1997, 1998, 1999, 2000 Free Software Foundation, Inc.
+
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 2, or (at your option)
+# any later version.
+
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the Free Software Foundation,
+# Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+
+# Written by Brendan O'Dea <bod@compusol.com.au>
+# Available from ftp://ftp.gnu.org/gnu/help2man/
+
+use 5.004;
+use strict;
+use Getopt::Long;
+use Text::Tabs qw(expand);
+use POSIX qw(strftime setlocale LC_TIME);
+
+my $this_program = 'help2man';
+my $this_version = '1.022';
+my $version_info = <<EOT;
+GNU $this_program $this_version
+
+Copyright (C) 1997, 1998, 1999, 2000 Free Software Foundation, Inc.
+This is free software; see the source for copying conditions.  There is NO
+warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
+
+Written by Brendan O'Dea <bod\@compusol.com.au>
+EOT
+
+my $help_info = <<EOT;
+`$this_program' generates a man page out of `--help' and `--version' output.
+
+Usage: $this_program [OPTION]... EXECUTABLE
+
+ -n, --name=STRING       use `STRING' as the description for the NAME paragraph
+ -s, --section=SECTION   use `SECTION' as the section for the man page
+ -i, --include=FILE      include material from `FILE'
+ -I, --opt-include=FILE  include material from `FILE' if it exists
+ -o, --output=FILE       send output to `FILE'
+ -N, --no-info           suppress pointer to Texinfo manual
+     --help              print this help, then exit
+     --version           print version number, then exit
+
+EXECUTABLE should accept `--help' and `--version' options.
+
+Report bugs to <bug-help2man\@gnu.org>.
+EOT
+
+my $section = 1;
+my ($opt_name, @opt_include, $opt_output, $opt_no_info);
+
+# Parse options.
+Getopt::Long::config('bundling');
+GetOptions (
+    'n|name=s'		=> \$opt_name,
+    's|section=s'	=> \$section,
+    'i|include=s'	=> sub { push @opt_include, [ pop, 1 ] },
+    'I|opt-include=s'	=> sub { push @opt_include, [ pop, 0 ] },
+    'o|output=s'	=> \$opt_output,
+    'N|no-info'		=> \$opt_no_info,
+    help		=> sub { print $help_info; exit },
+    version		=> sub { print $version_info; exit },
+) or die $help_info;
+
+die $help_info unless @ARGV == 1;
+
+my %include = ();
+my %append = ();
+my @include = (); # retain order given in include file
+
+# Provide replacement `quote-regex' operator for pre-5.005.
+BEGIN { eval q(sub qr { '' =~ $_[0]; $_[0] }) if $] < 5.005 }
+
+# Process include file (if given).  Format is:
+#
+#   [section name]
+#   verbatim text
+#
+# or
+#
+#   /pattern/
+#   verbatim text
+#
+
+for (@opt_include)
+{
+    my ($inc, $required) = @$_;
+
+    next unless -f $inc or $required;
+    die "$this_program: can't open `$inc' ($!)\n"
+	unless open INC, $inc;
+
+    my $key;
+    my $hash = \%include;
+
+    while (<INC>)
+    {
+	# [section]
+	if (/^\[([^]]+)\]/)
+	{
+	    $key = uc $1;
+	    $key =~ s/^\s+//;
+	    $key =~ s/\s+$//;
+	    $hash = \%include;
+	    push @include, $key unless $include{$key};
+	    next;
+	}
+
+	# /pattern/
+	if (m!^/(.*)/([ims]*)!)
+	{
+	    my $pat = $2 ? "(?$2)$1" : $1;
+
+	    # Check pattern.
+	    eval { $key = qr($pat) };
+	    if ($@)
+	    {
+		$@ =~ s/ at .*? line \d.*//;
+		die "$inc:$.:$@";
+	    }
+
+	    $hash = \%append;
+	    next;
+	}
+
+	# Silently ignore anything before the first
+	# section--allows for comments and revision info.
+	next unless $key;
+
+	$hash->{$key} ||= '';
+	$hash->{$key} .= $_;
+    }
+
+    close INC;
+
+    die "$this_program: no valid information found in `$inc'\n"
+	unless $key;
+}
+
+# Compress trailing blank lines.
+for my $hash (\(%include, %append))
+{
+    for (keys %$hash) { $hash->{$_} =~ s/\n+$/\n/ }
+}
+
+# Turn off localisation of executable's ouput.
+@ENV{qw(LANGUAGE LANG LC_ALL)} = ('C') x 3;
+
+# Turn off localisation of date (for strftime).
+setlocale LC_TIME, 'C';
+
+# Grab help and version info from executable.
+my ($help_text, $version_text) = map {
+    join '', map { s/ +$//; expand $_ } `$ARGV[0] --$_ 2>/dev/null`
+	or die "$this_program: can't get `--$_' info from $ARGV[0]\n"
+} qw(help version);
+
+my $date = strftime "%B %Y", localtime;
+(my $program = $ARGV[0]) =~ s!.*/!!;
+my $package = $program;
+my $version;
+
+if ($opt_output)
+{
+    unlink $opt_output
+	or die "$this_program: can't unlink $opt_output ($!)\n"
+	if -e $opt_output;
+
+    open STDOUT, ">$opt_output"
+	or die "$this_program: can't create $opt_output ($!)\n";
+}
+
+# The first line of the --version information is assumed to be in one
+# of the following formats:
+#
+#   <version>
+#   <program> <version>
+#   {GNU,Free} <program> <version>
+#   <program> ({GNU,Free} <package>) <version>
+#   <program> - {GNU,Free} <package> <version>
+#
+# and seperated from any copyright/author details by a blank line.
+
+($_, $version_text) = split /\n+/, $version_text, 2;
+
+if (/^(\S+) +\(((?:GNU|Free) +[^)]+)\) +(.*)/ or
+    /^(\S+) +- *((?:GNU|Free) +\S+) +(.*)/)
+{
+    $program = $1;
+    $package = $2;
+    $version = $3;
+}
+elsif (/^((?:GNU|Free) +)?(\S+) +(.*)/)
+{
+    $program = $2;
+    $package = $1 ? "$1$2" : $2;
+    $version = $3;
+}
+else
+{
+    $version = $_;
+}
+
+$program =~ s!.*/!!;
+
+# No info for `info' itself.
+$opt_no_info = 1 if $program eq 'info';
+
+# --name overrides --include contents.
+$include{NAME} = "$program \\- $opt_name\n" if $opt_name;
+
+# Default (useless) NAME paragraph.
+$include{NAME} ||= "$program \\- manual page for $program $version\n";
+
+# Man pages traditionally have the page title in caps.
+my $PROGRAM = uc $program;
+
+# Extract usage clause(s) [if any] for SYNOPSIS.
+if ($help_text =~ s/^Usage:( +(\S+))(.*)((?:\n(?: {6}\1| *or: +\S).*)*)//m)
+{
+    my @syn = $2 . $3;
+
+    if ($_ = $4)
+    {
+	s/^\n//;
+	for (split /\n/) { s/^ *(or: +)?//; push @syn, $_ }
+    }
+
+    my $synopsis = '';
+    for (@syn)
+    {
+	$synopsis .= ".br\n" if $synopsis;
+	s!^\S*/!!;
+	s/^(\S+) *//;
+	$synopsis .= ".B $1\n";
+	s/\s+$//;
+	s/(([][]|\.\.+)+)/\\fR$1\\fI/g;
+	s/^/\\fI/ unless s/^\\fR//;
+	$_ .= '\fR';
+	s/(\\fI)( *)/$2$1/g;
+	s/\\fI\\fR//g;
+	s/^\\fR//;
+	s/\\fI$//;
+	s/^\./\\&./;
+
+	$synopsis .= "$_\n";
+    }
+
+    $include{SYNOPSIS} ||= $synopsis;
+}
+
+# Process text, initial section is DESCRIPTION.
+my $sect = 'DESCRIPTION';
+$_ = "$help_text\n\n$version_text";
+
+# Normalise paragraph breaks.
+s/^\n+//;
+s/\n*$/\n/;
+s/\n\n+/\n\n/g;
+
+# Temporarily exchange leading dots and backslashes for tokens.
+s/^\./\x80/mg;
+s/\\/\x81/g;
+
+# Start a new paragraph (if required) for these.
+s/([^\n])\n(Report +bugs|Email +bug +reports +to|Written +by)/$1\n\n$2/g;
+
+sub convert_option;
+
+while (length)
+{
+    # Convert some standard paragraph names.
+    if (s/^(Options|Examples): *\n//)
+    {
+	$sect = uc $1;
+	next;
+    }
+
+    # Copyright section
+    if (/^Copyright +[(\xa9]/)
+    {
+	$sect = 'COPYRIGHT';
+	$include{$sect} ||= '';
+	$include{$sect} .= ".PP\n" if $include{$sect};
+
+	my $copy;
+	($copy, $_) = split /\n\n/, $_, 2;
+
+	for ($copy)
+	{
+	    # Add back newline
+	    s/\n*$/\n/;
+
+	    # Convert iso9959-1 copyright symbol or (c) to nroff
+	    # character.
+	    s/^Copyright +(?:\xa9|\([Cc]\))/Copyright \\(co/mg;
+
+	    # Insert line breaks before additional copyright messages
+	    # and the disclaimer.
+	    s/(.)\n(Copyright |This +is +free +software)/$1\n.br\n$2/g;
+
+	    # Join hyphenated lines.
+	    s/([A-Za-z])-\n */$1/g;
+	}
+
+	$include{$sect} .= $copy;
+	$_ ||= '';
+	next;
+    }
+
+    # Catch bug report text.
+    if (/^(Report +bugs|Email +bug +reports +to) /)
+    {
+	$sect = 'REPORTING BUGS';
+    }
+
+    # Author section.
+    elsif (/^Written +by/)
+    {
+	$sect = 'AUTHOR';
+    }
+
+    # Examples, indicated by an indented leading $, % or > are
+    # rendered in a constant width font.
+    if (/^( +)([\$\%>] )\S/)
+    {
+	my $indent = $1;
+	my $prefix = $2;
+	my $break = '.IP';
+	$include{$sect} ||= '';
+	while (s/^$indent\Q$prefix\E(\S.*)\n*//)
+	{
+	    $include{$sect} .= "$break\n\\f(CW$prefix$1\\fR\n";
+	    $break = '.br';
+	}
+
+	next;
+    }
+
+    my $matched = '';
+    $include{$sect} ||= '';
+
+    # Sub-sections have a trailing colon and the second line indented.
+    if (s/^(\S.*:) *\n / /)
+    {
+	$matched .= $& if %append;
+	$include{$sect} .= qq(.SS "$1"\n);
+    }
+
+    my $indent = 0;
+    my $content = '';
+
+    # Option with description.
+    if (s/^( {1,10}([+-]\S.*?))(?:(  +)|\n( {20,}))(\S.*)\n//)
+    {
+	$matched .= $& if %append;
+	$indent = length ($4 || "$1$3");
+	$content = ".TP\n\x82$2\n\x82$5\n";
+	unless ($4)
+	{
+	    # Indent may be different on second line.
+	    $indent = length $& if /^ {20,}/;
+	}
+    }
+
+    # Option without description.
+    elsif (s/^ {1,10}([+-]\S.*)\n//)
+    {
+	$matched .= $& if %append;
+	$content = ".HP\n\x82$1\n";
+	$indent = 80; # not continued
+    }
+
+    # Indented paragraph with tag.
+    elsif (s/^( +(\S.*?)  +)(\S.*)\n//)
+    {
+	$matched .= $& if %append;
+	$indent = length $1;
+	$content = ".TP\n\x82$2\n\x82$3\n";
+    }
+
+    # Indented paragraph.
+    elsif (s/^( +)(\S.*)\n//)
+    {
+	$matched .= $& if %append;
+	$indent = length $1;
+	$content = ".IP\n\x82$2\n";
+    }
+
+    # Left justified paragraph.
+    else
+    {
+	s/(.*)\n//;
+	$matched .= $& if %append;
+	$content = ".PP\n" if $include{$sect};
+	$content .= "$1\n";
+    }
+
+    # Append continuations.
+    while (s/^ {$indent}(\S.*)\n//)
+    {
+	$matched .= $& if %append;
+	$content .= "\x82$1\n"
+    }
+
+    # Move to next paragraph.
+    s/^\n+//;
+
+    for ($content)
+    {
+	# Leading dot protection.
+	s/\x82\./\x80/g;
+	s/\x82//g;
+
+	# Convert options.
+	s/(^| )(-[][\w=-]+)/$1 . convert_option $2/mge;
+    }
+
+    # Check if matched paragraph contains /pat/.
+    if (%append)
+    {
+	for my $pat (keys %append)
+	{
+	    if ($matched =~ $pat)
+	    {
+		$content .= ".PP\n" unless $append{$pat} =~ /^\./;
+		$content .= $append{$pat};
+	    }
+	}
+    }
+
+    $include{$sect} .= $content;
+}
+
+# Refer to the real documentation.
+unless ($opt_no_info)
+{
+    $sect = 'SEE ALSO';
+    $include{$sect} ||= '';
+    $include{$sect} .= ".PP\n" if $include{$sect};
+    $include{$sect} .= <<EOT;
+The full documentation for
+.B $program
+is maintained as a Texinfo manual.  If the
+.B info
+and
+.B $program
+programs are properly installed at your site, the command
+.IP
+.B info $program
+.PP
+should give you access to the complete manual.
+EOT
+}
+
+# Output header.
+print <<EOT;
+.\\" DO NOT MODIFY THIS FILE!  It was generated by $this_program $this_version.
+.TH $PROGRAM "$section" "$date" "$package $version" FSF
+EOT
+
+# Section ordering.
+my @pre = qw(NAME SYNOPSIS DESCRIPTION OPTIONS EXAMPLES);
+my @post = ('AUTHOR', 'REPORTING BUGS', 'COPYRIGHT', 'SEE ALSO');
+my $filter = join '|', @pre, @post;
+
+# Output content.
+for (@pre, (grep ! /^($filter)$/o, @include), @post)
+{
+    if ($include{$_})
+    {
+	my $quote = /\W/ ? '"' : '';
+	print ".SH $quote$_$quote\n";
+
+	for ($include{$_})
+	{
+	    # Replace leading dot an backslash tokens.
+	    s/\x80/\\&./g;
+	    s/\x81/\\e/g;
+	    print;
+	}
+    }
+}
+
+exit;
+
+# Convert option dashes to \- to stop nroff from hyphenating 'em, and
+# embolden.  Option arguments get italicised.
+sub convert_option
+{
+    local $_ = '\fB' . shift;
+
+    s/-/\\-/g;
+    unless (s/\[=(.*)\]$/\\fR[=\\fI$1\\fR]/)
+    {
+	s/=(.)/\\fR=\\fI$1/;
+	s/ (.)/ \\fI$1/;
+	$_ .= '\fR';
+    }
+
+    $_;
+}
diff --git a/contrib/gperf/doc/texinfo.tex b/contrib/gperf/doc/texinfo.tex
new file mode 100644
index 000000000000..faad86b0e247
--- /dev/null
+++ b/contrib/gperf/doc/texinfo.tex
@@ -0,0 +1,5999 @@
+% texinfo.tex -- TeX macros to handle Texinfo files.
+%
+% Load plain if necessary, i.e., if running under initex.
+\expandafter\ifx\csname fmtname\endcsname\relax\input plain\fi
+%
+\def\texinfoversion{2000-05-28.15}
+%
+% Copyright (C) 1985, 86, 88, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99
+% Free Software Foundation, Inc.
+%
+% This texinfo.tex file is free software; you can redistribute it and/or
+% modify it under the terms of the GNU General Public License as
+% published by the Free Software Foundation; either version 2, or (at
+% your option) any later version.
+%
+% This texinfo.tex file is distributed in the hope that it will be
+% useful, but WITHOUT ANY WARRANTY; without even the implied warranty
+% of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+% General Public License for more details.
+%
+% You should have received a copy of the GNU General Public License
+% along with this texinfo.tex file; see the file COPYING.  If not, write
+% to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
+% Boston, MA 02111-1307, USA.
+%
+% In other words, you are welcome to use, share and improve this program.
+% You are forbidden to forbid anyone else to use, share and improve
+% what you give them.   Help stamp out software-hoarding!
+%
+% Please try the latest version of texinfo.tex before submitting bug
+% reports; you can get the latest version from:
+%   ftp://ftp.gnu.org/gnu/texinfo.tex
+%   (and all GNU mirrors, see http://www.gnu.org/order/ftp.html)
+%   ftp://texinfo.org/tex/texinfo.tex
+%   ftp://us.ctan.org/macros/texinfo/texinfo.tex
+%   (and all CTAN mirrors, finger ctan@us.ctan.org for a list).
+%   /home/gd/gnu/doc/texinfo.tex on the GNU machines.
+% The texinfo.tex in any given Texinfo distribution could well be out
+% of date, so if that's what you're using, please check.
+% Texinfo has a small home page at http://texinfo.org/.
+%
+% Send bug reports to bug-texinfo@gnu.org.  Please include including a
+% complete document in each bug report with which we can reproduce the
+% problem.  Patches are, of course, greatly appreciated.
+%
+% To process a Texinfo manual with TeX, it's most reliable to use the
+% texi2dvi shell script that comes with the distribution.  For a simple
+% manual foo.texi, however, you can get away with this:
+%   tex foo.texi
+%   texindex foo.??
+%   tex foo.texi
+%   tex foo.texi
+%   dvips foo.dvi -o # or whatever, to process the dvi file; this makes foo.ps.
+% The extra runs of TeX get the cross-reference information correct.
+% Sometimes one run after texindex suffices, and sometimes you need more
+% than two; texi2dvi does it as many times as necessary.
+%
+% It is possible to adapt texinfo.tex for other languages.  You can get
+% the existing language-specific files from ftp://ftp.gnu.org/gnu/texinfo/.
+
+\message{Loading texinfo [version \texinfoversion]:}
+
+% If in a .fmt file, print the version number
+% and turn on active characters that we couldn't do earlier because
+% they might have appeared in the input file name.
+\everyjob{\message{[Texinfo version \texinfoversion]}%
+  \catcode`+=\active \catcode`\_=\active}
+
+% Save some parts of plain tex whose names we will redefine.
+\let\ptexb=\b
+\let\ptexbullet=\bullet
+\let\ptexc=\c
+\let\ptexcomma=\,
+\let\ptexdot=\.
+\let\ptexdots=\dots
+\let\ptexend=\end
+\let\ptexequiv=\equiv
+\let\ptexexclam=\!
+\let\ptexi=\i
+\let\ptexlbrace=\{
+\let\ptexrbrace=\}
+\let\ptexstar=\*
+\let\ptext=\t
+
+% We never want plain's outer \+ definition in Texinfo.
+% For @tex, we can use \tabalign.
+\let\+ = \relax
+
+\message{Basics,}
+\chardef\other=12
+
+% If this character appears in an error message or help string, it
+% starts a new line in the output.
+\newlinechar = `^^J
+
+% Set up fixed words for English if not already set.
+\ifx\putwordAppendix\undefined  \gdef\putwordAppendix{Appendix}\fi
+\ifx\putwordChapter\undefined   \gdef\putwordChapter{Chapter}\fi
+\ifx\putwordfile\undefined      \gdef\putwordfile{file}\fi
+\ifx\putwordin\undefined        \gdef\putwordin{in}\fi
+\ifx\putwordIndexIsEmpty\undefined     \gdef\putwordIndexIsEmpty{(Index is empty)}\fi
+\ifx\putwordIndexNonexistent\undefined \gdef\putwordIndexNonexistent{(Index is nonexistent)}\fi
+\ifx\putwordInfo\undefined      \gdef\putwordInfo{Info}\fi
+\ifx\putwordInstanceVariableof\undefined \gdef\putwordInstanceVariableof{Instance Variable of}\fi
+\ifx\putwordMethodon\undefined  \gdef\putwordMethodon{Method on}\fi
+\ifx\putwordNoTitle\undefined   \gdef\putwordNoTitle{No Title}\fi
+\ifx\putwordof\undefined        \gdef\putwordof{of}\fi
+\ifx\putwordon\undefined        \gdef\putwordon{on}\fi
+\ifx\putwordpage\undefined      \gdef\putwordpage{page}\fi
+\ifx\putwordsection\undefined   \gdef\putwordsection{section}\fi
+\ifx\putwordSection\undefined   \gdef\putwordSection{Section}\fi
+\ifx\putwordsee\undefined       \gdef\putwordsee{see}\fi
+\ifx\putwordSee\undefined       \gdef\putwordSee{See}\fi
+\ifx\putwordShortTOC\undefined  \gdef\putwordShortTOC{Short Contents}\fi
+\ifx\putwordTOC\undefined       \gdef\putwordTOC{Table of Contents}\fi
+%
+\ifx\putwordMJan\undefined \gdef\putwordMJan{January}\fi
+\ifx\putwordMFeb\undefined \gdef\putwordMFeb{February}\fi
+\ifx\putwordMMar\undefined \gdef\putwordMMar{March}\fi
+\ifx\putwordMApr\undefined \gdef\putwordMApr{April}\fi
+\ifx\putwordMMay\undefined \gdef\putwordMMay{May}\fi
+\ifx\putwordMJun\undefined \gdef\putwordMJun{June}\fi
+\ifx\putwordMJul\undefined \gdef\putwordMJul{July}\fi
+\ifx\putwordMAug\undefined \gdef\putwordMAug{August}\fi
+\ifx\putwordMSep\undefined \gdef\putwordMSep{September}\fi
+\ifx\putwordMOct\undefined \gdef\putwordMOct{October}\fi
+\ifx\putwordMNov\undefined \gdef\putwordMNov{November}\fi
+\ifx\putwordMDec\undefined \gdef\putwordMDec{December}\fi
+%
+\ifx\putwordDefmac\undefined    \gdef\putwordDefmac{Macro}\fi
+\ifx\putwordDefspec\undefined   \gdef\putwordDefspec{Special Form}\fi
+\ifx\putwordDefvar\undefined    \gdef\putwordDefvar{Variable}\fi
+\ifx\putwordDefopt\undefined    \gdef\putwordDefopt{User Option}\fi
+\ifx\putwordDeftypevar\undefined\gdef\putwordDeftypevar{Variable}\fi
+\ifx\putwordDeffunc\undefined   \gdef\putwordDeffunc{Function}\fi
+\ifx\putwordDeftypefun\undefined\gdef\putwordDeftypefun{Function}\fi
+
+% Ignore a token.
+%
+\def\gobble#1{}
+
+\hyphenation{ap-pen-dix}
+\hyphenation{mini-buf-fer mini-buf-fers}
+\hyphenation{eshell}
+\hyphenation{white-space}
+
+% Margin to add to right of even pages, to left of odd pages.
+\newdimen \bindingoffset
+\newdimen \normaloffset
+\newdimen\pagewidth \newdimen\pageheight
+
+% Sometimes it is convenient to have everything in the transcript file
+% and nothing on the terminal.  We don't just call \tracingall here,
+% since that produces some useless output on the terminal.
+%
+\def\gloggingall{\begingroup \globaldefs = 1 \loggingall \endgroup}%
+\ifx\eTeXversion\undefined
+\def\loggingall{\tracingcommands2 \tracingstats2
+   \tracingpages1 \tracingoutput1 \tracinglostchars1
+   \tracingmacros2 \tracingparagraphs1 \tracingrestores1
+   \showboxbreadth\maxdimen\showboxdepth\maxdimen
+}%
+\else
+\def\loggingall{\tracingcommands3 \tracingstats2
+   \tracingpages1 \tracingoutput1 \tracinglostchars1
+   \tracingmacros2 \tracingparagraphs1 \tracingrestores1
+   \tracingscantokens1 \tracingassigns1 \tracingifs1
+   \tracinggroups1 \tracingnesting2
+   \showboxbreadth\maxdimen\showboxdepth\maxdimen
+}%
+\fi
+
+% For @cropmarks command.
+% Do @cropmarks to get crop marks.
+%
+\newif\ifcropmarks
+\let\cropmarks = \cropmarkstrue
+%
+% Dimensions to add cropmarks at corners.
+% Added by P. A. MacKay, 12 Nov. 1986
+%
+\newdimen\outerhsize \newdimen\outervsize % set by the paper size routines
+\newdimen\cornerlong  \cornerlong=1pc
+\newdimen\cornerthick \cornerthick=.3pt
+\newdimen\topandbottommargin \topandbottommargin=.75in
+
+% Main output routine.
+\chardef\PAGE = 255
+\output = {\onepageout{\pagecontents\PAGE}}
+
+\newbox\headlinebox
+\newbox\footlinebox
+
+% \onepageout takes a vbox as an argument.  Note that \pagecontents
+% does insertions, but you have to call it yourself.
+\def\onepageout#1{%
+  \ifcropmarks \hoffset=0pt \else \hoffset=\normaloffset \fi
+  %
+  \ifodd\pageno  \advance\hoffset by \bindingoffset
+  \else \advance\hoffset by -\bindingoffset\fi
+  %
+  % Do this outside of the \shipout so @code etc. will be expanded in
+  % the headline as they should be, not taken literally (outputting ''code).
+  \setbox\headlinebox = \vbox{\let\hsize=\pagewidth \makeheadline}%
+  \setbox\footlinebox = \vbox{\let\hsize=\pagewidth \makefootline}%
+  %
+  {%
+    % Have to do this stuff outside the \shipout because we want it to
+    % take effect in \write's, yet the group defined by the \vbox ends
+    % before the \shipout runs.
+    %
+    \escapechar = `\\     % use backslash in output files.
+    \indexdummies         % don't expand commands in the output.
+    \normalturnoffactive  % \ in index entries must not stay \, e.g., if
+                   % the page break happens to be in the middle of an example.
+    \shipout\vbox{%
+      % Do this early so pdf references go to the beginning of the page.
+      \ifpdfmakepagedest \pdfmkdest{\the\pageno} \fi
+      %
+      \ifcropmarks \vbox to \outervsize\bgroup
+        \hsize = \outerhsize
+        \vskip-\topandbottommargin
+        \vtop to0pt{%
+          \line{\ewtop\hfil\ewtop}%
+          \nointerlineskip
+          \line{%
+            \vbox{\moveleft\cornerthick\nstop}%
+            \hfill
+            \vbox{\moveright\cornerthick\nstop}%
+          }%
+          \vss}%
+        \vskip\topandbottommargin
+        \line\bgroup
+          \hfil % center the page within the outer (page) hsize.
+          \ifodd\pageno\hskip\bindingoffset\fi
+          \vbox\bgroup
+      \fi
+      %
+      \unvbox\headlinebox
+      \pagebody{#1}%
+      \ifdim\ht\footlinebox > 0pt
+        % Only leave this space if the footline is nonempty.
+        % (We lessened \vsize for it in \oddfootingxxx.)
+        % The \baselineskip=24pt in plain's \makefootline has no effect.
+        \vskip 2\baselineskip
+        \unvbox\footlinebox
+      \fi
+      %
+      \ifcropmarks
+          \egroup % end of \vbox\bgroup
+        \hfil\egroup % end of (centering) \line\bgroup
+        \vskip\topandbottommargin plus1fill minus1fill
+        \boxmaxdepth = \cornerthick
+        \vbox to0pt{\vss
+          \line{%
+            \vbox{\moveleft\cornerthick\nsbot}%
+            \hfill
+            \vbox{\moveright\cornerthick\nsbot}%
+          }%
+          \nointerlineskip
+          \line{\ewbot\hfil\ewbot}%
+        }%
+      \egroup % \vbox from first cropmarks clause
+      \fi
+    }% end of \shipout\vbox
+  }% end of group with \turnoffactive
+  \advancepageno
+  \ifnum\outputpenalty>-20000 \else\dosupereject\fi
+}
+
+\newinsert\margin \dimen\margin=\maxdimen
+
+\def\pagebody#1{\vbox to\pageheight{\boxmaxdepth=\maxdepth #1}}
+{\catcode`\@ =11
+\gdef\pagecontents#1{\ifvoid\topins\else\unvbox\topins\fi
+% marginal hacks, juha@viisa.uucp (Juha Takala)
+\ifvoid\margin\else % marginal info is present
+  \rlap{\kern\hsize\vbox to\z@{\kern1pt\box\margin \vss}}\fi
+\dimen@=\dp#1 \unvbox#1
+\ifvoid\footins\else\vskip\skip\footins\footnoterule \unvbox\footins\fi
+\ifr@ggedbottom \kern-\dimen@ \vfil \fi}
+}
+
+% Here are the rules for the cropmarks.  Note that they are
+% offset so that the space between them is truly \outerhsize or \outervsize
+% (P. A. MacKay, 12 November, 1986)
+%
+\def\ewtop{\vrule height\cornerthick depth0pt width\cornerlong}
+\def\nstop{\vbox
+  {\hrule height\cornerthick depth\cornerlong width\cornerthick}}
+\def\ewbot{\vrule height0pt depth\cornerthick width\cornerlong}
+\def\nsbot{\vbox
+  {\hrule height\cornerlong depth\cornerthick width\cornerthick}}
+
+% Parse an argument, then pass it to #1.  The argument is the rest of
+% the input line (except we remove a trailing comment).  #1 should be a
+% macro which expects an ordinary undelimited TeX argument.
+%
+\def\parsearg#1{%
+  \let\next = #1%
+  \begingroup
+    \obeylines
+    \futurelet\temp\parseargx
+}
+
+% If the next token is an obeyed space (from an @example environment or
+% the like), remove it and recurse.  Otherwise, we're done.
+\def\parseargx{%
+  % \obeyedspace is defined far below, after the definition of \sepspaces.
+  \ifx\obeyedspace\temp
+    \expandafter\parseargdiscardspace
+  \else
+    \expandafter\parseargline
+  \fi
+}
+
+% Remove a single space (as the delimiter token to the macro call).
+{\obeyspaces %
+ \gdef\parseargdiscardspace {\futurelet\temp\parseargx}}
+
+{\obeylines %
+  \gdef\parseargline#1^^M{%
+    \endgroup % End of the group started in \parsearg.
+    %
+    % First remove any @c comment, then any @comment.
+    % Result of each macro is put in \toks0.
+    \argremovec #1\c\relax %
+    \expandafter\argremovecomment \the\toks0 \comment\relax %
+    %
+    % Call the caller's macro, saved as \next in \parsearg.
+    \expandafter\next\expandafter{\the\toks0}%
+  }%
+}
+
+% Since all \c{,omment} does is throw away the argument, we can let TeX
+% do that for us.  The \relax here is matched by the \relax in the call
+% in \parseargline; it could be more or less anything, its purpose is
+% just to delimit the argument to the \c.
+\def\argremovec#1\c#2\relax{\toks0 = {#1}}
+\def\argremovecomment#1\comment#2\relax{\toks0 = {#1}}
+
+% \argremovec{,omment} might leave us with trailing spaces, though; e.g.,
+%    @end itemize  @c foo
+% will have two active spaces as part of the argument with the
+% `itemize'.  Here we remove all active spaces from #1, and assign the
+% result to \toks0.
+%
+% This loses if there are any *other* active characters besides spaces
+% in the argument -- _ ^ +, for example -- since they get expanded.
+% Fortunately, Texinfo does not define any such commands.  (If it ever
+% does, the catcode of the characters in questionwill have to be changed
+% here.)  But this means we cannot call \removeactivespaces as part of
+% \argremovec{,omment}, since @c uses \parsearg, and thus the argument
+% that \parsearg gets might well have any character at all in it.
+%
+\def\removeactivespaces#1{%
+  \begingroup
+    \ignoreactivespaces
+    \edef\temp{#1}%
+    \global\toks0 = \expandafter{\temp}%
+  \endgroup
+}
+
+% Change the active space to expand to nothing.
+%
+\begingroup
+  \obeyspaces
+  \gdef\ignoreactivespaces{\obeyspaces\let =\empty}
+\endgroup
+
+
+\def\flushcr{\ifx\par\lisppar \def\next##1{}\else \let\next=\relax \fi \next}
+
+%% These are used to keep @begin/@end levels from running away
+%% Call \inENV within environments (after a \begingroup)
+\newif\ifENV \ENVfalse \def\inENV{\ifENV\relax\else\ENVtrue\fi}
+\def\ENVcheck{%
+\ifENV\errmessage{Still within an environment; press RETURN to continue}
+\endgroup\fi} % This is not perfect, but it should reduce lossage
+
+% @begin foo  is the same as @foo, for now.
+\newhelp\EMsimple{Press RETURN to continue.}
+
+\outer\def\begin{\parsearg\beginxxx}
+
+\def\beginxxx #1{%
+\expandafter\ifx\csname #1\endcsname\relax
+{\errhelp=\EMsimple \errmessage{Undefined command @begin #1}}\else
+\csname #1\endcsname\fi}
+
+% @end foo executes the definition of \Efoo.
+%
+\def\end{\parsearg\endxxx}
+\def\endxxx #1{%
+  \removeactivespaces{#1}%
+  \edef\endthing{\the\toks0}%
+  %
+  \expandafter\ifx\csname E\endthing\endcsname\relax
+    \expandafter\ifx\csname \endthing\endcsname\relax
+      % There's no \foo, i.e., no ``environment'' foo.
+      \errhelp = \EMsimple
+      \errmessage{Undefined command `@end \endthing'}%
+    \else
+      \unmatchedenderror\endthing
+    \fi
+  \else
+    % Everything's ok; the right environment has been started.
+    \csname E\endthing\endcsname
+  \fi
+}
+
+% There is an environment #1, but it hasn't been started.  Give an error.
+%
+\def\unmatchedenderror#1{%
+  \errhelp = \EMsimple
+  \errmessage{This `@end #1' doesn't have a matching `@#1'}%
+}
+
+% Define the control sequence \E#1 to give an unmatched @end error.
+%
+\def\defineunmatchedend#1{%
+  \expandafter\def\csname E#1\endcsname{\unmatchedenderror{#1}}%
+}
+
+
+% Single-spacing is done by various environments (specifically, in
+% \nonfillstart and \quotations).
+\newskip\singlespaceskip \singlespaceskip = 12.5pt
+\def\singlespace{%
+  % Why was this kern here?  It messes up equalizing space above and below
+  % environments.  --karl, 6may93
+  %{\advance \baselineskip by -\singlespaceskip
+  %\kern \baselineskip}%
+  \setleading \singlespaceskip
+}
+
+%% Simple single-character @ commands
+
+% @@ prints an @
+% Kludge this until the fonts are right (grr).
+\def\@{{\tt\char64}}
+
+% This is turned off because it was never documented
+% and you can use @w{...} around a quote to suppress ligatures.
+%% Define @` and @' to be the same as ` and '
+%% but suppressing ligatures.
+%\def\`{{`}}
+%\def\'{{'}}
+
+% Used to generate quoted braces.
+\def\mylbrace {{\tt\char123}}
+\def\myrbrace {{\tt\char125}}
+\let\{=\mylbrace
+\let\}=\myrbrace
+\begingroup
+  % Definitions to produce actual \{ & \} command in an index.
+  \catcode`\{ = 12 \catcode`\} = 12
+  \catcode`\[ = 1 \catcode`\] = 2
+  \catcode`\@ = 0 \catcode`\\ = 12
+  @gdef@lbracecmd[\{]%
+  @gdef@rbracecmd[\}]%
+@endgroup
+
+% Accents: @, @dotaccent @ringaccent @ubaraccent @udotaccent
+% Others are defined by plain TeX: @` @' @" @^ @~ @= @v @H.
+\let\, = \c
+\let\dotaccent = \.
+\def\ringaccent#1{{\accent23 #1}}
+\let\tieaccent = \t
+\let\ubaraccent = \b
+\let\udotaccent = \d
+
+% Other special characters: @questiondown @exclamdown
+% Plain TeX defines: @AA @AE @O @OE @L (and lowercase versions) @ss.
+\def\questiondown{?`}
+\def\exclamdown{!`}
+
+% Dotless i and dotless j, used for accents.
+\def\imacro{i}
+\def\jmacro{j}
+\def\dotless#1{%
+  \def\temp{#1}%
+  \ifx\temp\imacro \ptexi
+  \else\ifx\temp\jmacro \j
+  \else \errmessage{@dotless can be used only with i or j}%
+  \fi\fi
+}
+
+% Be sure we're in horizontal mode when doing a tie, since we make space
+% equivalent to this in @example-like environments. Otherwise, a space
+% at the beginning of a line will start with \penalty -- and
+% since \penalty is valid in vertical mode, we'd end up putting the
+% penalty on the vertical list instead of in the new paragraph.
+{\catcode`@ = 11
+ % Avoid using \@M directly, because that causes trouble
+ % if the definition is written into an index file.
+ \global\let\tiepenalty = \@M
+ \gdef\tie{\leavevmode\penalty\tiepenalty\ }
+}
+
+% @: forces normal size whitespace following.
+\def\:{\spacefactor=1000 }
+
+% @* forces a line break.
+\def\*{\hfil\break\hbox{}\ignorespaces}
+
+% @. is an end-of-sentence period.
+\def\.{.\spacefactor=3000 }
+
+% @! is an end-of-sentence bang.
+\def\!{!\spacefactor=3000 }
+
+% @? is an end-of-sentence query.
+\def\?{?\spacefactor=3000 }
+
+% @w prevents a word break.  Without the \leavevmode, @w at the
+% beginning of a paragraph, when TeX is still in vertical mode, would
+% produce a whole line of output instead of starting the paragraph.
+\def\w#1{\leavevmode\hbox{#1}}
+
+% @group ... @end group forces ... to be all on one page, by enclosing
+% it in a TeX vbox.  We use \vtop instead of \vbox to construct the box
+% to keep its height that of a normal line.  According to the rules for
+% \topskip (p.114 of the TeXbook), the glue inserted is
+% max (\topskip - \ht (first item), 0).  If that height is large,
+% therefore, no glue is inserted, and the space between the headline and
+% the text is small, which looks bad.
+%
+\def\group{\begingroup
+  \ifnum\catcode13=\active \else
+    \errhelp = \groupinvalidhelp
+    \errmessage{@group invalid in context where filling is enabled}%
+  \fi
+  %
+  % The \vtop we start below produces a box with normal height and large
+  % depth; thus, TeX puts \baselineskip glue before it, and (when the
+  % next line of text is done) \lineskip glue after it.  (See p.82 of
+  % the TeXbook.)  Thus, space below is not quite equal to space
+  % above.  But it's pretty close.
+  \def\Egroup{%
+    \egroup           % End the \vtop.
+    \endgroup         % End the \group.
+  }%
+  %
+  \vtop\bgroup
+    % We have to put a strut on the last line in case the @group is in
+    % the midst of an example, rather than completely enclosing it.
+    % Otherwise, the interline space between the last line of the group
+    % and the first line afterwards is too small.  But we can't put the
+    % strut in \Egroup, since there it would be on a line by itself.
+    % Hence this just inserts a strut at the beginning of each line.
+    \everypar = {\strut}%
+    %
+    % Since we have a strut on every line, we don't need any of TeX's
+    % normal interline spacing.
+    \offinterlineskip
+    %
+    % OK, but now we have to do something about blank
+    % lines in the input in @example-like environments, which normally
+    % just turn into \lisppar, which will insert no space now that we've
+    % turned off the interline space.  Simplest is to make them be an
+    % empty paragraph.
+    \ifx\par\lisppar
+      \edef\par{\leavevmode \par}%
+      %
+      % Reset ^^M's definition to new definition of \par.
+      \obeylines
+    \fi
+    %
+    % Do @comment since we are called inside an environment such as
+    % @example, where each end-of-line in the input causes an
+    % end-of-line in the output.  We don't want the end-of-line after
+    % the `@group' to put extra space in the output.  Since @group
+    % should appear on a line by itself (according to the Texinfo
+    % manual), we don't worry about eating any user text.
+    \comment
+}
+%
+% TeX puts in an \escapechar (i.e., `@') at the beginning of the help
+% message, so this ends up printing `@group can only ...'.
+%
+\newhelp\groupinvalidhelp{%
+group can only be used in environments such as @example,^^J%
+where each line of input produces a line of output.}
+
+% @need space-in-mils
+% forces a page break if there is not space-in-mils remaining.
+
+\newdimen\mil  \mil=0.001in
+
+\def\need{\parsearg\needx}
+
+% Old definition--didn't work.
+%\def\needx #1{\par %
+%% This method tries to make TeX break the page naturally
+%% if the depth of the box does not fit.
+%{\baselineskip=0pt%
+%\vtop to #1\mil{\vfil}\kern -#1\mil\nobreak
+%\prevdepth=-1000pt
+%}}
+
+\def\needx#1{%
+  % Ensure vertical mode, so we don't make a big box in the middle of a
+  % paragraph.
+  \par
+  %
+  % If the @need value is less than one line space, it's useless.
+  \dimen0 = #1\mil
+  \dimen2 = \ht\strutbox
+  \advance\dimen2 by \dp\strutbox
+  \ifdim\dimen0 > \dimen2
+    %
+    % Do a \strut just to make the height of this box be normal, so the
+    % normal leading is inserted relative to the preceding line.
+    % And a page break here is fine.
+    \vtop to #1\mil{\strut\vfil}%
+    %
+    % TeX does not even consider page breaks if a penalty added to the
+    % main vertical list is 10000 or more.  But in order to see if the
+    % empty box we just added fits on the page, we must make it consider
+    % page breaks.  On the other hand, we don't want to actually break the
+    % page after the empty box.  So we use a penalty of 9999.
+    %
+    % There is an extremely small chance that TeX will actually break the
+    % page at this \penalty, if there are no other feasible breakpoints in
+    % sight.  (If the user is using lots of big @group commands, which
+    % almost-but-not-quite fill up a page, TeX will have a hard time doing
+    % good page breaking, for example.)  However, I could not construct an
+    % example where a page broke at this \penalty; if it happens in a real
+    % document, then we can reconsider our strategy.
+    \penalty9999
+    %
+    % Back up by the size of the box, whether we did a page break or not.
+    \kern -#1\mil
+    %
+    % Do not allow a page break right after this kern.
+    \nobreak
+  \fi
+}
+
+% @br   forces paragraph break
+
+\let\br = \par
+
+% @dots{} output an ellipsis using the current font.
+% We do .5em per period so that it has the same spacing in a typewriter
+% font as three actual period characters.
+%
+\def\dots{%
+  \leavevmode
+  \hbox to 1.5em{%
+    \hskip 0pt plus 0.25fil minus 0.25fil
+    .\hss.\hss.%
+    \hskip 0pt plus 0.5fil minus 0.5fil
+  }%
+}
+
+% @enddots{} is an end-of-sentence ellipsis.
+%
+\def\enddots{%
+  \leavevmode
+  \hbox to 2em{%
+    \hskip 0pt plus 0.25fil minus 0.25fil
+    .\hss.\hss.\hss.%
+    \hskip 0pt plus 0.5fil minus 0.5fil
+  }%
+  \spacefactor=3000
+}
+
+
+% @page    forces the start of a new page
+%
+\def\page{\par\vfill\supereject}
+
+% @exdent text....
+% outputs text on separate line in roman font, starting at standard page margin
+
+% This records the amount of indent in the innermost environment.
+% That's how much \exdent should take out.
+\newskip\exdentamount
+
+% This defn is used inside fill environments such as @defun.
+\def\exdent{\parsearg\exdentyyy}
+\def\exdentyyy #1{{\hfil\break\hbox{\kern -\exdentamount{\rm#1}}\hfil\break}}
+
+% This defn is used inside nofill environments such as @example.
+\def\nofillexdent{\parsearg\nofillexdentyyy}
+\def\nofillexdentyyy #1{{\advance \leftskip by -\exdentamount
+\leftline{\hskip\leftskip{\rm#1}}}}
+
+% @inmargin{TEXT} puts TEXT in the margin next to the current paragraph.
+
+\def\inmargin#1{%
+\strut\vadjust{\nobreak\kern-\strutdepth
+  \vtop to \strutdepth{\baselineskip\strutdepth\vss
+  \llap{\rightskip=\inmarginspacing \vbox{\noindent #1}}\null}}}
+\newskip\inmarginspacing \inmarginspacing=1cm
+\def\strutdepth{\dp\strutbox}
+
+%\hbox{{\rm#1}}\hfil\break}}
+
+% @include file    insert text of that file as input.
+% Allow normal characters that  we make active in the argument (a file name).
+\def\include{\begingroup
+  \catcode`\\=12
+  \catcode`~=12
+  \catcode`^=12
+  \catcode`_=12
+  \catcode`|=12
+  \catcode`<=12
+  \catcode`>=12
+  \catcode`+=12
+  \parsearg\includezzz}
+% Restore active chars for included file.
+\def\includezzz#1{\endgroup\begingroup
+  % Read the included file in a group so nested @include's work.
+  \def\thisfile{#1}%
+  \input\thisfile
+\endgroup}
+
+\def\thisfile{}
+
+% @center line   outputs that line, centered
+
+\def\center{\parsearg\centerzzz}
+\def\centerzzz #1{{\advance\hsize by -\leftskip
+\advance\hsize by -\rightskip
+\centerline{#1}}}
+
+% @sp n   outputs n lines of vertical space
+
+\def\sp{\parsearg\spxxx}
+\def\spxxx #1{\vskip #1\baselineskip}
+
+% @comment ...line which is ignored...
+% @c is the same as @comment
+% @ignore ... @end ignore  is another way to write a comment
+
+\def\comment{\begingroup \catcode`\^^M=\other%
+\catcode`\@=\other \catcode`\{=\other \catcode`\}=\other%
+\commentxxx}
+{\catcode`\^^M=\other \gdef\commentxxx#1^^M{\endgroup}}
+
+\let\c=\comment
+
+% @paragraphindent NCHARS
+% We'll use ems for NCHARS, close enough.
+% We cannot implement @paragraphindent asis, though.
+% 
+\def\asisword{asis} % no translation, these are keywords
+\def\noneword{none}
+%
+\def\paragraphindent{\parsearg\doparagraphindent}
+\def\doparagraphindent#1{%
+  \def\temp{#1}%
+  \ifx\temp\asisword
+  \else
+    \ifx\temp\noneword
+      \defaultparindent = 0pt
+    \else
+      \defaultparindent = #1em
+    \fi
+  \fi
+  \parindent = \defaultparindent
+}
+
+% @exampleindent NCHARS
+% We'll use ems for NCHARS like @paragraphindent.
+% It seems @exampleindent asis isn't necessary, but
+% I preserve it to make it similar to @paragraphindent.
+\def\exampleindent{\parsearg\doexampleindent}
+\def\doexampleindent#1{%
+  \def\temp{#1}%
+  \ifx\temp\asisword
+  \else
+    \ifx\temp\noneword
+      \lispnarrowing = 0pt
+    \else
+      \lispnarrowing = #1em
+    \fi
+  \fi
+}
+
+% @asis just yields its argument.  Used with @table, for example.
+%
+\def\asis#1{#1}
+
+% @math means output in math mode.
+% We don't use $'s directly in the definition of \math because control
+% sequences like \math are expanded when the toc file is written.  Then,
+% we read the toc file back, the $'s will be normal characters (as they
+% should be, according to the definition of Texinfo).  So we must use a
+% control sequence to switch into and out of math mode.
+%
+% This isn't quite enough for @math to work properly in indices, but it
+% seems unlikely it will ever be needed there.
+%
+\let\implicitmath = $
+\def\math#1{\implicitmath #1\implicitmath}
+
+% @bullet and @minus need the same treatment as @math, just above.
+\def\bullet{\implicitmath\ptexbullet\implicitmath}
+\def\minus{\implicitmath-\implicitmath}
+
+% @refill is a no-op.
+\let\refill=\relax
+
+% If working on a large document in chapters, it is convenient to
+% be able to disable indexing, cross-referencing, and contents, for test runs.
+% This is done with @novalidate (before @setfilename).
+%
+\newif\iflinks \linkstrue % by default we want the aux files.
+\let\novalidate = \linksfalse
+
+% @setfilename is done at the beginning of every texinfo file.
+% So open here the files we need to have open while reading the input.
+% This makes it possible to make a .fmt file for texinfo.
+\def\setfilename{%
+   \iflinks
+     \readauxfile
+   \fi % \openindices needs to do some work in any case.
+   \openindices
+   \fixbackslash  % Turn off hack to swallow `\input texinfo'.
+   \global\let\setfilename=\comment % Ignore extra @setfilename cmds.
+   %
+   % If texinfo.cnf is present on the system, read it.
+   % Useful for site-wide @afourpaper, etc.
+   % Just to be on the safe side, close the input stream before the \input.
+   \openin 1 texinfo.cnf
+   \ifeof1 \let\temp=\relax \else \def\temp{\input texinfo.cnf }\fi
+   \closein1
+   \temp
+   %
+   \comment % Ignore the actual filename.
+}
+
+% Called from \setfilename.
+%
+\def\openindices{%
+  \newindex{cp}%
+  \newcodeindex{fn}%
+  \newcodeindex{vr}%
+  \newcodeindex{tp}%
+  \newcodeindex{ky}%
+  \newcodeindex{pg}%
+}
+
+% @bye.
+\outer\def\bye{\pagealignmacro\tracingstats=1\ptexend}
+
+
+\message{pdf,}
+% adobe `portable' document format
+\newcount\tempnum
+\newcount\lnkcount
+\newtoks\filename
+\newcount\filenamelength
+\newcount\pgn
+\newtoks\toksA
+\newtoks\toksB
+\newtoks\toksC
+\newtoks\toksD
+\newbox\boxA
+\newcount\countA
+\newif\ifpdf
+\newif\ifpdfmakepagedest
+
+\ifx\pdfoutput\undefined
+  \pdffalse
+  \let\pdfmkdest = \gobble
+  \let\pdfurl = \gobble
+  \let\endlink = \relax
+  \let\linkcolor = \relax
+  \let\pdfmakeoutlines = \relax
+\else
+  \pdftrue
+  \pdfoutput = 1
+  \input pdfcolor
+  \def\dopdfimage#1#2#3{%
+    \def\imagewidth{#2}%
+    \def\imageheight{#3}%
+    \ifnum\pdftexversion < 14
+      \pdfimage
+    \else
+      \pdfximage
+    \fi
+      \ifx\empty\imagewidth\else width \imagewidth \fi
+      \ifx\empty\imageheight\else height \imageheight \fi
+      {#1.pdf}%
+    \ifnum\pdftexversion < 14 \else
+      \pdfrefximage \pdflastximage
+    \fi}
+  \def\pdfmkdest#1{\pdfdest name{#1@} xyz}
+  \def\pdfmkpgn#1{#1@}
+  \let\linkcolor = \Blue  % was Cyan, but that seems light?
+  \def\endlink{\Black\pdfendlink}
+  % Adding outlines to PDF; macros for calculating structure of outlines
+  % come from Petr Olsak
+  \def\expnumber#1{\expandafter\ifx\csname#1\endcsname\relax 0%
+    \else \csname#1\endcsname \fi}
+  \def\advancenumber#1{\tempnum=\expnumber{#1}\relax
+    \advance\tempnum by1
+    \expandafter\xdef\csname#1\endcsname{\the\tempnum}}
+  \def\pdfmakeoutlines{{%
+    \openin 1 \jobname.toc
+    \ifeof 1\else\bgroup
+      \closein 1 
+      \indexnofonts
+      \def\tt{}
+      \let\_ = \normalunderscore
+      % Thanh's hack / proper braces in bookmarks  
+      \edef\mylbrace{\iftrue \string{\else}\fi}\let\{=\mylbrace
+      \edef\myrbrace{\iffalse{\else\string}\fi}\let\}=\myrbrace
+      %
+      \def\chapentry ##1##2##3{}
+      \def\unnumbchapentry ##1##2{}
+      \def\secentry ##1##2##3##4{\advancenumber{chap##2}}
+      \def\unnumbsecentry ##1##2{}
+      \def\subsecentry ##1##2##3##4##5{\advancenumber{sec##2.##3}}
+      \def\unnumbsubsecentry ##1##2{}
+      \def\subsubsecentry ##1##2##3##4##5##6{\advancenumber{subsec##2.##3.##4}}
+      \def\unnumbsubsubsecentry ##1##2{}
+      \input \jobname.toc
+      \def\chapentry ##1##2##3{%
+        \pdfoutline goto name{\pdfmkpgn{##3}}count-\expnumber{chap##2}{##1}}
+      \def\unnumbchapentry ##1##2{%
+        \pdfoutline goto name{\pdfmkpgn{##2}}{##1}}
+      \def\secentry ##1##2##3##4{%
+        \pdfoutline goto name{\pdfmkpgn{##4}}count-\expnumber{sec##2.##3}{##1}}
+      \def\unnumbsecentry ##1##2{%
+        \pdfoutline goto name{\pdfmkpgn{##2}}{##1}}
+      \def\subsecentry ##1##2##3##4##5{%
+        \pdfoutline goto name{\pdfmkpgn{##5}}count-\expnumber{subsec##2.##3.##4}{##1}}
+      \def\unnumbsubsecentry ##1##2{%
+        \pdfoutline goto name{\pdfmkpgn{##2}}{##1}}
+      \def\subsubsecentry ##1##2##3##4##5##6{%
+        \pdfoutline goto name{\pdfmkpgn{##6}}{##1}}
+      \def\unnumbsubsubsecentry ##1##2{%
+        \pdfoutline goto name{\pdfmkpgn{##2}}{##1}}
+      \input \jobname.toc
+    \egroup\fi
+  }}
+  \def\makelinks #1,{%
+    \def\params{#1}\def\E{END}%
+    \ifx\params\E
+      \let\nextmakelinks=\relax
+    \else
+      \let\nextmakelinks=\makelinks
+      \ifnum\lnkcount>0,\fi
+      \picknum{#1}%
+      \startlink attr{/Border [0 0 0]} 
+        goto name{\pdfmkpgn{\the\pgn}}%
+      \linkcolor #1%
+      \advance\lnkcount by 1%
+      \endlink
+    \fi
+    \nextmakelinks
+  }
+  \def\picknum#1{\expandafter\pn#1}
+  \def\pn#1{%
+    \def\p{#1}%
+    \ifx\p\lbrace
+      \let\nextpn=\ppn
+    \else
+      \let\nextpn=\ppnn
+      \def\first{#1}
+    \fi
+    \nextpn
+  }
+  \def\ppn#1{\pgn=#1\gobble}
+  \def\ppnn{\pgn=\first}
+  \def\pdfmklnk#1{\lnkcount=0\makelinks #1,END,}
+  \def\addtokens#1#2{\edef\addtoks{\noexpand#1={\the#1#2}}\addtoks}
+  \def\skipspaces#1{\def\PP{#1}\def\D{|}%
+    \ifx\PP\D\let\nextsp\relax
+    \else\let\nextsp\skipspaces
+      \ifx\p\space\else\addtokens{\filename}{\PP}%
+        \advance\filenamelength by 1
+      \fi
+    \fi
+    \nextsp}
+  \def\getfilename#1{\filenamelength=0\expandafter\skipspaces#1|\relax}
+  \ifnum\pdftexversion < 14
+    \let \startlink \pdfannotlink
+  \else
+    \let \startlink \pdfstartlink
+  \fi
+  \def\pdfurl#1{%
+    \begingroup
+      \normalturnoffactive\def\@{@}%
+      \leavevmode\Red
+      \startlink attr{/Border [0 0 0]}%
+        user{/Subtype /Link /A << /S /URI /URI (#1) >>}%
+        % #1
+    \endgroup}
+  \def\pdfgettoks#1.{\setbox\boxA=\hbox{\toksA={#1.}\toksB={}\maketoks}}
+  \def\addtokens#1#2{\edef\addtoks{\noexpand#1={\the#1#2}}\addtoks}
+  \def\adn#1{\addtokens{\toksC}{#1}\global\countA=1\let\next=\maketoks}
+  \def\poptoks#1#2|ENDTOKS|{\let\first=#1\toksD={#1}\toksA={#2}}
+  \def\maketoks{%
+    \expandafter\poptoks\the\toksA|ENDTOKS|
+    \ifx\first0\adn0
+    \else\ifx\first1\adn1 \else\ifx\first2\adn2 \else\ifx\first3\adn3
+    \else\ifx\first4\adn4 \else\ifx\first5\adn5 \else\ifx\first6\adn6
+    \else\ifx\first7\adn7 \else\ifx\first8\adn8 \else\ifx\first9\adn9 
+    \else
+      \ifnum0=\countA\else\makelink\fi
+      \ifx\first.\let\next=\done\else
+        \let\next=\maketoks
+        \addtokens{\toksB}{\the\toksD}
+        \ifx\first,\addtokens{\toksB}{\space}\fi
+      \fi
+    \fi\fi\fi\fi\fi\fi\fi\fi\fi\fi
+    \next}
+  \def\makelink{\addtokens{\toksB}%
+    {\noexpand\pdflink{\the\toksC}}\toksC={}\global\countA=0}
+  \def\pdflink#1{%
+    \startlink attr{/Border [0 0 0]} goto name{\mkpgn{#1}}
+    \linkcolor #1\endlink}
+  \def\mkpgn#1{#1@} 
+  \def\done{\edef\st{\global\noexpand\toksA={\the\toksB}}\st}
+\fi % \ifx\pdfoutput
+
+
+\message{fonts,}
+% Font-change commands.
+
+% Texinfo sort of supports the sans serif font style, which plain TeX does not.
+% So we set up a \sf analogous to plain's \rm, etc.
+\newfam\sffam
+\def\sf{\fam=\sffam \tensf}
+\let\li = \sf % Sometimes we call it \li, not \sf.
+
+% We don't need math for this one.
+\def\ttsl{\tenttsl}
+
+% Use Computer Modern fonts at \magstephalf (11pt).
+\newcount\mainmagstep
+\mainmagstep=\magstephalf
+
+% Set the font macro #1 to the font named #2, adding on the
+% specified font prefix (normally `cm').
+% #3 is the font's design size, #4 is a scale factor
+\def\setfont#1#2#3#4{\font#1=\fontprefix#2#3 scaled #4}
+
+% Use cm as the default font prefix.
+% To specify the font prefix, you must define \fontprefix
+% before you read in texinfo.tex.
+\ifx\fontprefix\undefined
+\def\fontprefix{cm}
+\fi
+% Support font families that don't use the same naming scheme as CM.
+\def\rmshape{r}
+\def\rmbshape{bx}               %where the normal face is bold
+\def\bfshape{b}
+\def\bxshape{bx}
+\def\ttshape{tt}
+\def\ttbshape{tt}
+\def\ttslshape{sltt}
+\def\itshape{ti}
+\def\itbshape{bxti}
+\def\slshape{sl}
+\def\slbshape{bxsl}
+\def\sfshape{ss}
+\def\sfbshape{ss}
+\def\scshape{csc}
+\def\scbshape{csc}
+
+\ifx\bigger\relax
+\let\mainmagstep=\magstep1
+\setfont\textrm\rmshape{12}{1000}
+\setfont\texttt\ttshape{12}{1000}
+\else
+\setfont\textrm\rmshape{10}{\mainmagstep}
+\setfont\texttt\ttshape{10}{\mainmagstep}
+\fi
+% Instead of cmb10, you many want to use cmbx10.
+% cmbx10 is a prettier font on its own, but cmb10
+% looks better when embedded in a line with cmr10.
+\setfont\textbf\bfshape{10}{\mainmagstep}
+\setfont\textit\itshape{10}{\mainmagstep}
+\setfont\textsl\slshape{10}{\mainmagstep}
+\setfont\textsf\sfshape{10}{\mainmagstep}
+\setfont\textsc\scshape{10}{\mainmagstep}
+\setfont\textttsl\ttslshape{10}{\mainmagstep}
+\font\texti=cmmi10 scaled \mainmagstep
+\font\textsy=cmsy10 scaled \mainmagstep
+
+% A few fonts for @defun, etc.
+\setfont\defbf\bxshape{10}{\magstep1} %was 1314
+\setfont\deftt\ttshape{10}{\magstep1}
+\def\df{\let\tentt=\deftt \let\tenbf = \defbf \bf}
+
+% Fonts for indices, footnotes, small examples (9pt).
+\setfont\smallrm\rmshape{9}{1000}
+\setfont\smalltt\ttshape{9}{1000}
+\setfont\smallbf\bfshape{10}{900}
+\setfont\smallit\itshape{9}{1000}
+\setfont\smallsl\slshape{9}{1000}
+\setfont\smallsf\sfshape{9}{1000}
+\setfont\smallsc\scshape{10}{900}
+\setfont\smallttsl\ttslshape{10}{900}
+\font\smalli=cmmi9
+\font\smallsy=cmsy9
+
+% Fonts for title page:
+\setfont\titlerm\rmbshape{12}{\magstep3}
+\setfont\titleit\itbshape{10}{\magstep4}
+\setfont\titlesl\slbshape{10}{\magstep4}
+\setfont\titlett\ttbshape{12}{\magstep3}
+\setfont\titlettsl\ttslshape{10}{\magstep4}
+\setfont\titlesf\sfbshape{17}{\magstep1}
+\let\titlebf=\titlerm
+\setfont\titlesc\scbshape{10}{\magstep4}
+\font\titlei=cmmi12 scaled \magstep3
+\font\titlesy=cmsy10 scaled \magstep4
+\def\authorrm{\secrm}
+
+% Chapter (and unnumbered) fonts (17.28pt).
+\setfont\chaprm\rmbshape{12}{\magstep2}
+\setfont\chapit\itbshape{10}{\magstep3}
+\setfont\chapsl\slbshape{10}{\magstep3}
+\setfont\chaptt\ttbshape{12}{\magstep2}
+\setfont\chapttsl\ttslshape{10}{\magstep3}
+\setfont\chapsf\sfbshape{17}{1000}
+\let\chapbf=\chaprm
+\setfont\chapsc\scbshape{10}{\magstep3}
+\font\chapi=cmmi12 scaled \magstep2
+\font\chapsy=cmsy10 scaled \magstep3
+
+% Section fonts (14.4pt).
+\setfont\secrm\rmbshape{12}{\magstep1}
+\setfont\secit\itbshape{10}{\magstep2}
+\setfont\secsl\slbshape{10}{\magstep2}
+\setfont\sectt\ttbshape{12}{\magstep1}
+\setfont\secttsl\ttslshape{10}{\magstep2}
+\setfont\secsf\sfbshape{12}{\magstep1}
+\let\secbf\secrm
+\setfont\secsc\scbshape{10}{\magstep2}
+\font\seci=cmmi12 scaled \magstep1
+\font\secsy=cmsy10 scaled \magstep2
+
+% \setfont\ssecrm\bxshape{10}{\magstep1}    % This size an font looked bad.
+% \setfont\ssecit\itshape{10}{\magstep1}    % The letters were too crowded.
+% \setfont\ssecsl\slshape{10}{\magstep1}
+% \setfont\ssectt\ttshape{10}{\magstep1}
+% \setfont\ssecsf\sfshape{10}{\magstep1}
+
+%\setfont\ssecrm\bfshape{10}{1315}      % Note the use of cmb rather than cmbx.
+%\setfont\ssecit\itshape{10}{1315}      % Also, the size is a little larger than
+%\setfont\ssecsl\slshape{10}{1315}      % being scaled magstep1.
+%\setfont\ssectt\ttshape{10}{1315}
+%\setfont\ssecsf\sfshape{10}{1315}
+
+%\let\ssecbf=\ssecrm
+
+% Subsection fonts (13.15pt).
+\setfont\ssecrm\rmbshape{12}{\magstephalf}
+\setfont\ssecit\itbshape{10}{1315}
+\setfont\ssecsl\slbshape{10}{1315}
+\setfont\ssectt\ttbshape{12}{\magstephalf}
+\setfont\ssecttsl\ttslshape{10}{1315}
+\setfont\ssecsf\sfbshape{12}{\magstephalf}
+\let\ssecbf\ssecrm
+\setfont\ssecsc\scbshape{10}{\magstep1}
+\font\sseci=cmmi12 scaled \magstephalf
+\font\ssecsy=cmsy10 scaled 1315
+% The smallcaps and symbol fonts should actually be scaled \magstep1.5,
+% but that is not a standard magnification.
+
+% In order for the font changes to affect most math symbols and letters,
+% we have to define the \textfont of the standard families.  Since
+% texinfo doesn't allow for producing subscripts and superscripts, we
+% don't bother to reset \scriptfont and \scriptscriptfont (which would
+% also require loading a lot more fonts).
+%
+\def\resetmathfonts{%
+  \textfont0 = \tenrm \textfont1 = \teni \textfont2 = \tensy
+  \textfont\itfam = \tenit \textfont\slfam = \tensl \textfont\bffam = \tenbf
+  \textfont\ttfam = \tentt \textfont\sffam = \tensf
+}
+
+
+% The font-changing commands redefine the meanings of \tenSTYLE, instead
+% of just \STYLE.  We do this so that font changes will continue to work
+% in math mode, where it is the current \fam that is relevant in most
+% cases, not the current font.  Plain TeX does \def\bf{\fam=\bffam
+% \tenbf}, for example.  By redefining \tenbf, we obviate the need to
+% redefine \bf itself.
+\def\textfonts{%
+  \let\tenrm=\textrm \let\tenit=\textit \let\tensl=\textsl
+  \let\tenbf=\textbf \let\tentt=\texttt \let\smallcaps=\textsc
+  \let\tensf=\textsf \let\teni=\texti \let\tensy=\textsy \let\tenttsl=\textttsl
+  \resetmathfonts}
+\def\titlefonts{%
+  \let\tenrm=\titlerm \let\tenit=\titleit \let\tensl=\titlesl
+  \let\tenbf=\titlebf \let\tentt=\titlett \let\smallcaps=\titlesc
+  \let\tensf=\titlesf \let\teni=\titlei \let\tensy=\titlesy
+  \let\tenttsl=\titlettsl
+  \resetmathfonts \setleading{25pt}}
+\def\titlefont#1{{\titlefonts\rm #1}}
+\def\chapfonts{%
+  \let\tenrm=\chaprm \let\tenit=\chapit \let\tensl=\chapsl
+  \let\tenbf=\chapbf \let\tentt=\chaptt \let\smallcaps=\chapsc
+  \let\tensf=\chapsf \let\teni=\chapi \let\tensy=\chapsy \let\tenttsl=\chapttsl
+  \resetmathfonts \setleading{19pt}}
+\def\secfonts{%
+  \let\tenrm=\secrm \let\tenit=\secit \let\tensl=\secsl
+  \let\tenbf=\secbf \let\tentt=\sectt \let\smallcaps=\secsc
+  \let\tensf=\secsf \let\teni=\seci \let\tensy=\secsy \let\tenttsl=\secttsl
+  \resetmathfonts \setleading{16pt}}
+\def\subsecfonts{%
+  \let\tenrm=\ssecrm \let\tenit=\ssecit \let\tensl=\ssecsl
+  \let\tenbf=\ssecbf \let\tentt=\ssectt \let\smallcaps=\ssecsc
+  \let\tensf=\ssecsf \let\teni=\sseci \let\tensy=\ssecsy \let\tenttsl=\ssecttsl
+  \resetmathfonts \setleading{15pt}}
+\let\subsubsecfonts = \subsecfonts % Maybe make sssec fonts scaled magstephalf?
+\def\smallfonts{%
+  \let\tenrm=\smallrm \let\tenit=\smallit \let\tensl=\smallsl
+  \let\tenbf=\smallbf \let\tentt=\smalltt \let\smallcaps=\smallsc
+  \let\tensf=\smallsf \let\teni=\smalli \let\tensy=\smallsy
+  \let\tenttsl=\smallttsl
+  \resetmathfonts \setleading{11pt}}
+
+% Set up the default fonts, so we can use them for creating boxes.
+%
+\textfonts
+
+% Define these so they can be easily changed for other fonts.
+\def\angleleft{$\langle$}
+\def\angleright{$\rangle$}
+
+% Count depth in font-changes, for error checks
+\newcount\fontdepth \fontdepth=0
+
+% Fonts for short table of contents.
+\setfont\shortcontrm\rmshape{12}{1000}
+\setfont\shortcontbf\bxshape{12}{1000}
+\setfont\shortcontsl\slshape{12}{1000}
+
+%% Add scribe-like font environments, plus @l for inline lisp (usually sans
+%% serif) and @ii for TeX italic
+
+% \smartitalic{ARG} outputs arg in italics, followed by an italic correction
+% unless the following character is such as not to need one.
+\def\smartitalicx{\ifx\next,\else\ifx\next-\else\ifx\next.\else\/\fi\fi\fi}
+\def\smartslanted#1{{\sl #1}\futurelet\next\smartitalicx}
+\def\smartitalic#1{{\it #1}\futurelet\next\smartitalicx}
+
+\let\i=\smartitalic
+\let\var=\smartslanted
+\let\dfn=\smartslanted
+\let\emph=\smartitalic
+\let\cite=\smartslanted
+
+\def\b#1{{\bf #1}}
+\let\strong=\b
+
+% We can't just use \exhyphenpenalty, because that only has effect at
+% the end of a paragraph.  Restore normal hyphenation at the end of the
+% group within which \nohyphenation is presumably called.
+%
+\def\nohyphenation{\hyphenchar\font = -1  \aftergroup\restorehyphenation}
+\def\restorehyphenation{\hyphenchar\font = `- }
+
+\def\t#1{%
+  {\tt \rawbackslash \frenchspacing #1}%
+  \null
+}
+\let\ttfont=\t
+\def\samp#1{`\tclose{#1}'\null}
+\setfont\keyrm\rmshape{8}{1000}
+\font\keysy=cmsy9
+\def\key#1{{\keyrm\textfont2=\keysy \leavevmode\hbox{%
+  \raise0.4pt\hbox{\angleleft}\kern-.08em\vtop{%
+    \vbox{\hrule\kern-0.4pt
+     \hbox{\raise0.4pt\hbox{\vphantom{\angleleft}}#1}}%
+    \kern-0.4pt\hrule}%
+  \kern-.06em\raise0.4pt\hbox{\angleright}}}}
+% The old definition, with no lozenge:
+%\def\key #1{{\ttsl \nohyphenation \uppercase{#1}}\null}
+\def\ctrl #1{{\tt \rawbackslash \hat}#1}
+
+% @file, @option are the same as @samp.
+\let\file=\samp
+\let\option=\samp
+
+% @code is a modification of @t,
+% which makes spaces the same size as normal in the surrounding text.
+\def\tclose#1{%
+  {%
+    % Change normal interword space to be same as for the current font.
+    \spaceskip = \fontdimen2\font
+    %
+    % Switch to typewriter.
+    \tt
+    %
+    % But `\ ' produces the large typewriter interword space.
+    \def\ {{\spaceskip = 0pt{} }}%
+    %
+    % Turn off hyphenation.
+    \nohyphenation
+    %
+    \rawbackslash
+    \frenchspacing
+    #1%
+  }%
+  \null
+}
+
+% We *must* turn on hyphenation at `-' and `_' in \code.
+% Otherwise, it is too hard to avoid overfull hboxes
+% in the Emacs manual, the Library manual, etc.
+
+% Unfortunately, TeX uses one parameter (\hyphenchar) to control
+% both hyphenation at - and hyphenation within words.
+% We must therefore turn them both off (\tclose does that)
+% and arrange explicitly to hyphenate at a dash.
+%  -- rms.
+{
+  \catcode`\-=\active
+  \catcode`\_=\active
+  %
+  \global\def\code{\begingroup
+    \catcode`\-=\active \let-\codedash
+    \catcode`\_=\active \let_\codeunder
+    \codex
+  }
+  %
+  % If we end up with any active - characters when handling the index,
+  % just treat them as a normal -.
+  \global\def\indexbreaks{\catcode`\-=\active \let-\realdash}
+}
+
+\def\realdash{-}
+\def\codedash{-\discretionary{}{}{}}
+\def\codeunder{\ifusingtt{\normalunderscore\discretionary{}{}{}}{\_}}
+\def\codex #1{\tclose{#1}\endgroup}
+
+%\let\exp=\tclose  %Was temporary
+
+% @kbd is like @code, except that if the argument is just one @key command,
+% then @kbd has no effect.
+
+% @kbdinputstyle -- arg is `distinct' (@kbd uses slanted tty font always),
+%   `example' (@kbd uses ttsl only inside of @example and friends),
+%   or `code' (@kbd uses normal tty font always).
+\def\kbdinputstyle{\parsearg\kbdinputstylexxx}
+\def\kbdinputstylexxx#1{%
+  \def\arg{#1}%
+  \ifx\arg\worddistinct
+    \gdef\kbdexamplefont{\ttsl}\gdef\kbdfont{\ttsl}%
+  \else\ifx\arg\wordexample
+    \gdef\kbdexamplefont{\ttsl}\gdef\kbdfont{\tt}%
+  \else\ifx\arg\wordcode
+    \gdef\kbdexamplefont{\tt}\gdef\kbdfont{\tt}%
+  \fi\fi\fi
+}
+\def\worddistinct{distinct}
+\def\wordexample{example}
+\def\wordcode{code}
+
+% Default is kbdinputdistinct.  (Too much of a hassle to call the macro,
+% the catcodes are wrong for parsearg to work.)
+\gdef\kbdexamplefont{\ttsl}\gdef\kbdfont{\ttsl}
+
+\def\xkey{\key}
+\def\kbdfoo#1#2#3\par{\def\one{#1}\def\three{#3}\def\threex{??}%
+\ifx\one\xkey\ifx\threex\three \key{#2}%
+\else{\tclose{\kbdfont\look}}\fi
+\else{\tclose{\kbdfont\look}}\fi}
+
+% For @url, @env, @command quotes seem unnecessary, so use \code.
+\let\url=\code
+\let\env=\code
+\let\command=\code
+
+% @uref (abbreviation for `urlref') takes an optional (comma-separated)
+% second argument specifying the text to display and an optional third
+% arg as text to display instead of (rather than in addition to) the url
+% itself.  First (mandatory) arg is the url.  Perhaps eventually put in
+% a hypertex \special here.
+%
+\def\uref#1{\douref #1,,,\finish}
+\def\douref#1,#2,#3,#4\finish{\begingroup
+  \unsepspaces
+  \pdfurl{#1}%
+  \setbox0 = \hbox{\ignorespaces #3}%
+  \ifdim\wd0 > 0pt
+    \unhbox0 % third arg given, show only that
+  \else
+    \setbox0 = \hbox{\ignorespaces #2}%
+    \ifdim\wd0 > 0pt
+      \ifpdf
+        \unhbox0             % PDF: 2nd arg given, show only it
+      \else
+        \unhbox0\ (\code{#1})% DVI: 2nd arg given, show both it and url
+      \fi
+    \else
+      \code{#1}% only url given, so show it
+    \fi
+  \fi
+  \endlink
+\endgroup}
+
+% rms does not like angle brackets --karl, 17may97.
+% So now @email is just like @uref, unless we are pdf.
+% 
+%\def\email#1{\angleleft{\tt #1}\angleright}
+\ifpdf
+  \def\email#1{\doemail#1,,\finish}
+  \def\doemail#1,#2,#3\finish{\begingroup
+    \unsepspaces
+    \pdfurl{mailto:#1}%
+    \setbox0 = \hbox{\ignorespaces #2}%
+    \ifdim\wd0>0pt\unhbox0\else\code{#1}\fi
+    \endlink
+  \endgroup}
+\else
+  \let\email=\uref
+\fi
+
+% Check if we are currently using a typewriter font.  Since all the
+% Computer Modern typewriter fonts have zero interword stretch (and
+% shrink), and it is reasonable to expect all typewriter fonts to have
+% this property, we can check that font parameter.
+%
+\def\ifmonospace{\ifdim\fontdimen3\font=0pt }
+
+% Typeset a dimension, e.g., `in' or `pt'.  The only reason for the
+% argument is to make the input look right: @dmn{pt} instead of @dmn{}pt.
+%
+\def\dmn#1{\thinspace #1}
+
+\def\kbd#1{\def\look{#1}\expandafter\kbdfoo\look??\par}
+
+% @l was never documented to mean ``switch to the Lisp font'',
+% and it is not used as such in any manual I can find.  We need it for
+% Polish suppressed-l.  --karl, 22sep96.
+%\def\l#1{{\li #1}\null}
+
+% Explicit font changes: @r, @sc, undocumented @ii.
+\def\r#1{{\rm #1}}              % roman font
+\def\sc#1{{\smallcaps#1}}       % smallcaps font
+\def\ii#1{{\it #1}}             % italic font
+
+% @acronym downcases the argument and prints in smallcaps.
+\def\acronym#1{{\smallcaps \lowercase{#1}}}
+
+% @pounds{} is a sterling sign.
+\def\pounds{{\it\$}}
+
+
+\message{page headings,}
+
+\newskip\titlepagetopglue \titlepagetopglue = 1.5in
+\newskip\titlepagebottomglue \titlepagebottomglue = 2pc
+
+% First the title page.  Must do @settitle before @titlepage.
+\newif\ifseenauthor
+\newif\iffinishedtitlepage
+
+% Do an implicit @contents or @shortcontents after @end titlepage if the
+% user says @setcontentsaftertitlepage or @setshortcontentsaftertitlepage.
+%
+\newif\ifsetcontentsaftertitlepage
+ \let\setcontentsaftertitlepage = \setcontentsaftertitlepagetrue
+\newif\ifsetshortcontentsaftertitlepage
+ \let\setshortcontentsaftertitlepage = \setshortcontentsaftertitlepagetrue
+
+\def\shorttitlepage{\parsearg\shorttitlepagezzz}
+\def\shorttitlepagezzz #1{\begingroup\hbox{}\vskip 1.5in \chaprm \centerline{#1}%
+        \endgroup\page\hbox{}\page}
+
+\def\titlepage{\begingroup \parindent=0pt \textfonts
+   \let\subtitlerm=\tenrm
+   \def\subtitlefont{\subtitlerm \normalbaselineskip = 13pt \normalbaselines}%
+   %
+   \def\authorfont{\authorrm \normalbaselineskip = 16pt \normalbaselines}%
+   %
+   % Leave some space at the very top of the page.
+   \vglue\titlepagetopglue
+   %
+   % Now you can print the title using @title.
+   \def\title{\parsearg\titlezzz}%
+   \def\titlezzz##1{\leftline{\titlefonts\rm ##1}
+                    % print a rule at the page bottom also.
+                    \finishedtitlepagefalse
+                    \vskip4pt \hrule height 4pt width \hsize \vskip4pt}%
+   % No rule at page bottom unless we print one at the top with @title.
+   \finishedtitlepagetrue
+   %
+   % Now you can put text using @subtitle.
+   \def\subtitle{\parsearg\subtitlezzz}%
+   \def\subtitlezzz##1{{\subtitlefont \rightline{##1}}}%
+   %
+   % @author should come last, but may come many times.
+   \def\author{\parsearg\authorzzz}%
+   \def\authorzzz##1{\ifseenauthor\else\vskip 0pt plus 1filll\seenauthortrue\fi
+      {\authorfont \leftline{##1}}}%
+   %
+   % Most title ``pages'' are actually two pages long, with space
+   % at the top of the second.  We don't want the ragged left on the second.
+   \let\oldpage = \page
+   \def\page{%
+      \iffinishedtitlepage\else
+         \finishtitlepage
+      \fi
+      \oldpage
+      \let\page = \oldpage
+      \hbox{}}%
+%   \def\page{\oldpage \hbox{}}
+}
+
+\def\Etitlepage{%
+   \iffinishedtitlepage\else
+      \finishtitlepage
+   \fi
+   % It is important to do the page break before ending the group,
+   % because the headline and footline are only empty inside the group.
+   % If we use the new definition of \page, we always get a blank page
+   % after the title page, which we certainly don't want.
+   \oldpage
+   \endgroup
+   %
+   % If they want short, they certainly want long too.
+   \ifsetshortcontentsaftertitlepage
+     \shortcontents
+     \contents
+     \global\let\shortcontents = \relax
+     \global\let\contents = \relax
+   \fi
+   %
+   \ifsetcontentsaftertitlepage
+     \contents
+     \global\let\contents = \relax
+     \global\let\shortcontents = \relax
+   \fi
+   %
+   \ifpdf \pdfmakepagedesttrue \fi
+   %
+   \HEADINGSon
+}
+
+\def\finishtitlepage{%
+   \vskip4pt \hrule height 2pt width \hsize
+   \vskip\titlepagebottomglue
+   \finishedtitlepagetrue
+}
+
+%%% Set up page headings and footings.
+
+\let\thispage=\folio
+
+\newtoks\evenheadline    % headline on even pages
+\newtoks\oddheadline     % headline on odd pages
+\newtoks\evenfootline    % footline on even pages
+\newtoks\oddfootline     % footline on odd pages
+
+% Now make Tex use those variables
+\headline={{\textfonts\rm \ifodd\pageno \the\oddheadline
+                            \else \the\evenheadline \fi}}
+\footline={{\textfonts\rm \ifodd\pageno \the\oddfootline
+                            \else \the\evenfootline \fi}\HEADINGShook}
+\let\HEADINGShook=\relax
+
+% Commands to set those variables.
+% For example, this is what  @headings on  does
+% @evenheading @thistitle|@thispage|@thischapter
+% @oddheading @thischapter|@thispage|@thistitle
+% @evenfooting @thisfile||
+% @oddfooting ||@thisfile
+
+\def\evenheading{\parsearg\evenheadingxxx}
+\def\oddheading{\parsearg\oddheadingxxx}
+\def\everyheading{\parsearg\everyheadingxxx}
+
+\def\evenfooting{\parsearg\evenfootingxxx}
+\def\oddfooting{\parsearg\oddfootingxxx}
+\def\everyfooting{\parsearg\everyfootingxxx}
+
+{\catcode`\@=0 %
+
+\gdef\evenheadingxxx #1{\evenheadingyyy #1@|@|@|@|\finish}
+\gdef\evenheadingyyy #1@|#2@|#3@|#4\finish{%
+\global\evenheadline={\rlap{\centerline{#2}}\line{#1\hfil#3}}}
+
+\gdef\oddheadingxxx #1{\oddheadingyyy #1@|@|@|@|\finish}
+\gdef\oddheadingyyy #1@|#2@|#3@|#4\finish{%
+\global\oddheadline={\rlap{\centerline{#2}}\line{#1\hfil#3}}}
+
+\gdef\everyheadingxxx#1{\oddheadingxxx{#1}\evenheadingxxx{#1}}%
+
+\gdef\evenfootingxxx #1{\evenfootingyyy #1@|@|@|@|\finish}
+\gdef\evenfootingyyy #1@|#2@|#3@|#4\finish{%
+\global\evenfootline={\rlap{\centerline{#2}}\line{#1\hfil#3}}}
+
+\gdef\oddfootingxxx #1{\oddfootingyyy #1@|@|@|@|\finish}
+\gdef\oddfootingyyy #1@|#2@|#3@|#4\finish{%
+  \global\oddfootline = {\rlap{\centerline{#2}}\line{#1\hfil#3}}%
+  %
+  % Leave some space for the footline.  Hopefully ok to assume
+  % @evenfooting will not be used by itself.
+  \global\advance\pageheight by -\baselineskip
+  \global\advance\vsize by -\baselineskip
+}
+
+\gdef\everyfootingxxx#1{\oddfootingxxx{#1}\evenfootingxxx{#1}}
+%
+}% unbind the catcode of @.
+
+% @headings double      turns headings on for double-sided printing.
+% @headings single      turns headings on for single-sided printing.
+% @headings off         turns them off.
+% @headings on          same as @headings double, retained for compatibility.
+% @headings after       turns on double-sided headings after this page.
+% @headings doubleafter turns on double-sided headings after this page.
+% @headings singleafter turns on single-sided headings after this page.
+% By default, they are off at the start of a document,
+% and turned `on' after @end titlepage.
+
+\def\headings #1 {\csname HEADINGS#1\endcsname}
+
+\def\HEADINGSoff{
+\global\evenheadline={\hfil} \global\evenfootline={\hfil}
+\global\oddheadline={\hfil} \global\oddfootline={\hfil}}
+\HEADINGSoff
+% When we turn headings on, set the page number to 1.
+% For double-sided printing, put current file name in lower left corner,
+% chapter name on inside top of right hand pages, document
+% title on inside top of left hand pages, and page numbers on outside top
+% edge of all pages.
+\def\HEADINGSdouble{
+\global\pageno=1
+\global\evenfootline={\hfil}
+\global\oddfootline={\hfil}
+\global\evenheadline={\line{\folio\hfil\thistitle}}
+\global\oddheadline={\line{\thischapter\hfil\folio}}
+\global\let\contentsalignmacro = \chapoddpage
+}
+\let\contentsalignmacro = \chappager
+
+% For single-sided printing, chapter title goes across top left of page,
+% page number on top right.
+\def\HEADINGSsingle{
+\global\pageno=1
+\global\evenfootline={\hfil}
+\global\oddfootline={\hfil}
+\global\evenheadline={\line{\thischapter\hfil\folio}}
+\global\oddheadline={\line{\thischapter\hfil\folio}}
+\global\let\contentsalignmacro = \chappager
+}
+\def\HEADINGSon{\HEADINGSdouble}
+
+\def\HEADINGSafter{\let\HEADINGShook=\HEADINGSdoublex}
+\let\HEADINGSdoubleafter=\HEADINGSafter
+\def\HEADINGSdoublex{%
+\global\evenfootline={\hfil}
+\global\oddfootline={\hfil}
+\global\evenheadline={\line{\folio\hfil\thistitle}}
+\global\oddheadline={\line{\thischapter\hfil\folio}}
+\global\let\contentsalignmacro = \chapoddpage
+}
+
+\def\HEADINGSsingleafter{\let\HEADINGShook=\HEADINGSsinglex}
+\def\HEADINGSsinglex{%
+\global\evenfootline={\hfil}
+\global\oddfootline={\hfil}
+\global\evenheadline={\line{\thischapter\hfil\folio}}
+\global\oddheadline={\line{\thischapter\hfil\folio}}
+\global\let\contentsalignmacro = \chappager
+}
+
+% Subroutines used in generating headings
+% This produces Day Month Year style of output.
+% Only define if not already defined, in case a txi-??.tex file has set
+% up a different format (e.g., txi-cs.tex does this).
+\ifx\today\undefined
+\def\today{%
+  \number\day\space
+  \ifcase\month
+  \or\putwordMJan\or\putwordMFeb\or\putwordMMar\or\putwordMApr
+  \or\putwordMMay\or\putwordMJun\or\putwordMJul\or\putwordMAug
+  \or\putwordMSep\or\putwordMOct\or\putwordMNov\or\putwordMDec
+  \fi
+  \space\number\year}
+\fi
+
+% @settitle line...  specifies the title of the document, for headings.
+% It generates no output of its own.
+\def\thistitle{\putwordNoTitle}
+\def\settitle{\parsearg\settitlezzz}
+\def\settitlezzz #1{\gdef\thistitle{#1}}
+
+
+\message{tables,}
+% Tables -- @table, @ftable, @vtable, @item(x), @kitem(x), @xitem(x).
+
+% default indentation of table text
+\newdimen\tableindent \tableindent=.8in
+% default indentation of @itemize and @enumerate text
+\newdimen\itemindent  \itemindent=.3in
+% margin between end of table item and start of table text.
+\newdimen\itemmargin  \itemmargin=.1in
+
+% used internally for \itemindent minus \itemmargin
+\newdimen\itemmax
+
+% Note @table, @vtable, and @vtable define @item, @itemx, etc., with
+% these defs.
+% They also define \itemindex
+% to index the item name in whatever manner is desired (perhaps none).
+
+\newif\ifitemxneedsnegativevskip
+
+\def\itemxpar{\par\ifitemxneedsnegativevskip\nobreak\vskip-\parskip\nobreak\fi}
+
+\def\internalBitem{\smallbreak \parsearg\itemzzz}
+\def\internalBitemx{\itemxpar \parsearg\itemzzz}
+
+\def\internalBxitem "#1"{\def\xitemsubtopix{#1} \smallbreak \parsearg\xitemzzz}
+\def\internalBxitemx "#1"{\def\xitemsubtopix{#1} \itemxpar \parsearg\xitemzzz}
+
+\def\internalBkitem{\smallbreak \parsearg\kitemzzz}
+\def\internalBkitemx{\itemxpar \parsearg\kitemzzz}
+
+\def\kitemzzz #1{\dosubind {kw}{\code{#1}}{for {\bf \lastfunction}}%
+                 \itemzzz {#1}}
+
+\def\xitemzzz #1{\dosubind {kw}{\code{#1}}{for {\bf \xitemsubtopic}}%
+                 \itemzzz {#1}}
+
+\def\itemzzz #1{\begingroup %
+  \advance\hsize by -\rightskip
+  \advance\hsize by -\tableindent
+  \setbox0=\hbox{\itemfont{#1}}%
+  \itemindex{#1}%
+  \nobreak % This prevents a break before @itemx.
+  %
+  % If the item text does not fit in the space we have, put it on a line
+  % by itself, and do not allow a page break either before or after that
+  % line.  We do not start a paragraph here because then if the next
+  % command is, e.g., @kindex, the whatsit would get put into the
+  % horizontal list on a line by itself, resulting in extra blank space.
+  \ifdim \wd0>\itemmax
+    %
+    % Make this a paragraph so we get the \parskip glue and wrapping,
+    % but leave it ragged-right.
+    \begingroup
+      \advance\leftskip by-\tableindent
+      \advance\hsize by\tableindent
+      \advance\rightskip by0pt plus1fil
+      \leavevmode\unhbox0\par
+    \endgroup
+    %
+    % We're going to be starting a paragraph, but we don't want the
+    % \parskip glue -- logically it's part of the @item we just started.
+    \nobreak \vskip-\parskip
+    %
+    % Stop a page break at the \parskip glue coming up.  Unfortunately
+    % we can't prevent a possible page break at the following
+    % \baselineskip glue.
+    \nobreak
+    \endgroup
+    \itemxneedsnegativevskipfalse
+  \else
+    % The item text fits into the space.  Start a paragraph, so that the
+    % following text (if any) will end up on the same line.
+    \noindent
+    % Do this with kerns and \unhbox so that if there is a footnote in
+    % the item text, it can migrate to the main vertical list and
+    % eventually be printed.
+    \nobreak\kern-\tableindent
+    \dimen0 = \itemmax  \advance\dimen0 by \itemmargin \advance\dimen0 by -\wd0
+    \unhbox0
+    \nobreak\kern\dimen0
+    \endgroup
+    \itemxneedsnegativevskiptrue
+  \fi
+}
+
+\def\item{\errmessage{@item while not in a table}}
+\def\itemx{\errmessage{@itemx while not in a table}}
+\def\kitem{\errmessage{@kitem while not in a table}}
+\def\kitemx{\errmessage{@kitemx while not in a table}}
+\def\xitem{\errmessage{@xitem while not in a table}}
+\def\xitemx{\errmessage{@xitemx while not in a table}}
+
+% Contains a kludge to get @end[description] to work.
+\def\description{\tablez{\dontindex}{1}{}{}{}{}}
+
+% @table, @ftable, @vtable.
+\def\table{\begingroup\inENV\obeylines\obeyspaces\tablex}
+{\obeylines\obeyspaces%
+\gdef\tablex #1^^M{%
+\tabley\dontindex#1        \endtabley}}
+
+\def\ftable{\begingroup\inENV\obeylines\obeyspaces\ftablex}
+{\obeylines\obeyspaces%
+\gdef\ftablex #1^^M{%
+\tabley\fnitemindex#1        \endtabley
+\def\Eftable{\endgraf\afterenvbreak\endgroup}%
+\let\Etable=\relax}}
+
+\def\vtable{\begingroup\inENV\obeylines\obeyspaces\vtablex}
+{\obeylines\obeyspaces%
+\gdef\vtablex #1^^M{%
+\tabley\vritemindex#1        \endtabley
+\def\Evtable{\endgraf\afterenvbreak\endgroup}%
+\let\Etable=\relax}}
+
+\def\dontindex #1{}
+\def\fnitemindex #1{\doind {fn}{\code{#1}}}%
+\def\vritemindex #1{\doind {vr}{\code{#1}}}%
+
+{\obeyspaces %
+\gdef\tabley#1#2 #3 #4 #5 #6 #7\endtabley{\endgroup%
+\tablez{#1}{#2}{#3}{#4}{#5}{#6}}}
+
+\def\tablez #1#2#3#4#5#6{%
+\aboveenvbreak %
+\begingroup %
+\def\Edescription{\Etable}% Necessary kludge.
+\let\itemindex=#1%
+\ifnum 0#3>0 \advance \leftskip by #3\mil \fi %
+\ifnum 0#4>0 \tableindent=#4\mil \fi %
+\ifnum 0#5>0 \advance \rightskip by #5\mil \fi %
+\def\itemfont{#2}%
+\itemmax=\tableindent %
+\advance \itemmax by -\itemmargin %
+\advance \leftskip by \tableindent %
+\exdentamount=\tableindent
+\parindent = 0pt
+\parskip = \smallskipamount
+\ifdim \parskip=0pt \parskip=2pt \fi%
+\def\Etable{\endgraf\afterenvbreak\endgroup}%
+\let\item = \internalBitem %
+\let\itemx = \internalBitemx %
+\let\kitem = \internalBkitem %
+\let\kitemx = \internalBkitemx %
+\let\xitem = \internalBxitem %
+\let\xitemx = \internalBxitemx %
+}
+
+% This is the counter used by @enumerate, which is really @itemize
+
+\newcount \itemno
+
+\def\itemize{\parsearg\itemizezzz}
+
+\def\itemizezzz #1{%
+  \begingroup % ended by the @end itemize
+  \itemizey {#1}{\Eitemize}
+}
+
+\def\itemizey #1#2{%
+\aboveenvbreak %
+\itemmax=\itemindent %
+\advance \itemmax by -\itemmargin %
+\advance \leftskip by \itemindent %
+\exdentamount=\itemindent
+\parindent = 0pt %
+\parskip = \smallskipamount %
+\ifdim \parskip=0pt \parskip=2pt \fi%
+\def#2{\endgraf\afterenvbreak\endgroup}%
+\def\itemcontents{#1}%
+\let\item=\itemizeitem}
+
+% Set sfcode to normal for the chars that usually have another value.
+% These are `.?!:;,'
+\def\frenchspacing{\sfcode46=1000 \sfcode63=1000 \sfcode33=1000
+  \sfcode58=1000 \sfcode59=1000 \sfcode44=1000 }
+
+% \splitoff TOKENS\endmark defines \first to be the first token in
+% TOKENS, and \rest to be the remainder.
+%
+\def\splitoff#1#2\endmark{\def\first{#1}\def\rest{#2}}%
+
+% Allow an optional argument of an uppercase letter, lowercase letter,
+% or number, to specify the first label in the enumerated list.  No
+% argument is the same as `1'.
+%
+\def\enumerate{\parsearg\enumeratezzz}
+\def\enumeratezzz #1{\enumeratey #1  \endenumeratey}
+\def\enumeratey #1 #2\endenumeratey{%
+  \begingroup % ended by the @end enumerate
+  %
+  % If we were given no argument, pretend we were given `1'.
+  \def\thearg{#1}%
+  \ifx\thearg\empty \def\thearg{1}\fi
+  %
+  % Detect if the argument is a single token.  If so, it might be a
+  % letter.  Otherwise, the only valid thing it can be is a number.
+  % (We will always have one token, because of the test we just made.
+  % This is a good thing, since \splitoff doesn't work given nothing at
+  % all -- the first parameter is undelimited.)
+  \expandafter\splitoff\thearg\endmark
+  \ifx\rest\empty
+    % Only one token in the argument.  It could still be anything.
+    % A ``lowercase letter'' is one whose \lccode is nonzero.
+    % An ``uppercase letter'' is one whose \lccode is both nonzero, and
+    %   not equal to itself.
+    % Otherwise, we assume it's a number.
+    %
+    % We need the \relax at the end of the \ifnum lines to stop TeX from
+    % continuing to look for a <number>.
+    %
+    \ifnum\lccode\expandafter`\thearg=0\relax
+      \numericenumerate % a number (we hope)
+    \else
+      % It's a letter.
+      \ifnum\lccode\expandafter`\thearg=\expandafter`\thearg\relax
+        \lowercaseenumerate % lowercase letter
+      \else
+        \uppercaseenumerate % uppercase letter
+      \fi
+    \fi
+  \else
+    % Multiple tokens in the argument.  We hope it's a number.
+    \numericenumerate
+  \fi
+}
+
+% An @enumerate whose labels are integers.  The starting integer is
+% given in \thearg.
+%
+\def\numericenumerate{%
+  \itemno = \thearg
+  \startenumeration{\the\itemno}%
+}
+
+% The starting (lowercase) letter is in \thearg.
+\def\lowercaseenumerate{%
+  \itemno = \expandafter`\thearg
+  \startenumeration{%
+    % Be sure we're not beyond the end of the alphabet.
+    \ifnum\itemno=0
+      \errmessage{No more lowercase letters in @enumerate; get a bigger
+                  alphabet}%
+    \fi
+    \char\lccode\itemno
+  }%
+}
+
+% The starting (uppercase) letter is in \thearg.
+\def\uppercaseenumerate{%
+  \itemno = \expandafter`\thearg
+  \startenumeration{%
+    % Be sure we're not beyond the end of the alphabet.
+    \ifnum\itemno=0
+      \errmessage{No more uppercase letters in @enumerate; get a bigger
+                  alphabet}
+    \fi
+    \char\uccode\itemno
+  }%
+}
+
+% Call itemizey, adding a period to the first argument and supplying the
+% common last two arguments.  Also subtract one from the initial value in
+% \itemno, since @item increments \itemno.
+%
+\def\startenumeration#1{%
+  \advance\itemno by -1
+  \itemizey{#1.}\Eenumerate\flushcr
+}
+
+% @alphaenumerate and @capsenumerate are abbreviations for giving an arg
+% to @enumerate.
+%
+\def\alphaenumerate{\enumerate{a}}
+\def\capsenumerate{\enumerate{A}}
+\def\Ealphaenumerate{\Eenumerate}
+\def\Ecapsenumerate{\Eenumerate}
+
+% Definition of @item while inside @itemize.
+
+\def\itemizeitem{%
+\advance\itemno by 1
+{\let\par=\endgraf \smallbreak}%
+\ifhmode \errmessage{In hmode at itemizeitem}\fi
+{\parskip=0in \hskip 0pt
+\hbox to 0pt{\hss \itemcontents\hskip \itemmargin}%
+\vadjust{\penalty 1200}}%
+\flushcr}
+
+% @multitable macros
+% Amy Hendrickson, 8/18/94, 3/6/96
+%
+% @multitable ... @end multitable will make as many columns as desired.
+% Contents of each column will wrap at width given in preamble.  Width
+% can be specified either with sample text given in a template line,
+% or in percent of \hsize, the current width of text on page.
+
+% Table can continue over pages but will only break between lines.
+
+% To make preamble:
+%
+% Either define widths of columns in terms of percent of \hsize:
+%   @multitable @columnfractions .25 .3 .45
+%   @item ...
+%
+%   Numbers following @columnfractions are the percent of the total
+%   current hsize to be used for each column. You may use as many
+%   columns as desired.
+
+
+% Or use a template:
+%   @multitable {Column 1 template} {Column 2 template} {Column 3 template}
+%   @item ...
+%   using the widest term desired in each column.
+%
+% For those who want to use more than one line's worth of words in
+% the preamble, break the line within one argument and it
+% will parse correctly, i.e.,
+%
+%     @multitable {Column 1 template} {Column 2 template} {Column 3
+%      template}
+% Not:
+%     @multitable {Column 1 template} {Column 2 template}
+%      {Column 3 template}
+
+% Each new table line starts with @item, each subsequent new column
+% starts with @tab. Empty columns may be produced by supplying @tab's
+% with nothing between them for as many times as empty columns are needed,
+% ie, @tab@tab@tab will produce two empty columns.
+
+% @item, @tab, @multitable or @end multitable do not need to be on their
+% own lines, but it will not hurt if they are.
+
+% Sample multitable:
+
+%   @multitable {Column 1 template} {Column 2 template} {Column 3 template}
+%   @item first col stuff @tab second col stuff @tab third col
+%   @item
+%   first col stuff
+%   @tab
+%   second col stuff
+%   @tab
+%   third col
+%   @item first col stuff @tab second col stuff
+%   @tab Many paragraphs of text may be used in any column.
+%
+%         They will wrap at the width determined by the template.
+%   @item@tab@tab This will be in third column.
+%   @end multitable
+
+% Default dimensions may be reset by user.
+% @multitableparskip is vertical space between paragraphs in table.
+% @multitableparindent is paragraph indent in table.
+% @multitablecolmargin is horizontal space to be left between columns.
+% @multitablelinespace is space to leave between table items, baseline
+%                                                            to baseline.
+%   0pt means it depends on current normal line spacing.
+%
+\newskip\multitableparskip
+\newskip\multitableparindent
+\newdimen\multitablecolspace
+\newskip\multitablelinespace
+\multitableparskip=0pt
+\multitableparindent=6pt
+\multitablecolspace=12pt
+\multitablelinespace=0pt
+
+% Macros used to set up halign preamble:
+%
+\let\endsetuptable\relax
+\def\xendsetuptable{\endsetuptable}
+\let\columnfractions\relax
+\def\xcolumnfractions{\columnfractions}
+\newif\ifsetpercent
+
+% #1 is the part of the @columnfraction before the decimal point, which
+% is presumably either 0 or the empty string (but we don't check, we
+% just throw it away).  #2 is the decimal part, which we use as the
+% percent of \hsize for this column.
+\def\pickupwholefraction#1.#2 {%
+  \global\advance\colcount by 1
+  \expandafter\xdef\csname col\the\colcount\endcsname{.#2\hsize}%
+  \setuptable
+}
+
+\newcount\colcount
+\def\setuptable#1{%
+  \def\firstarg{#1}%
+  \ifx\firstarg\xendsetuptable
+    \let\go = \relax
+  \else
+    \ifx\firstarg\xcolumnfractions
+      \global\setpercenttrue
+    \else
+      \ifsetpercent
+         \let\go\pickupwholefraction
+      \else
+         \global\advance\colcount by 1
+         \setbox0=\hbox{#1\unskip }% Add a normal word space as a separator;
+                            % typically that is always in the input, anyway.
+         \expandafter\xdef\csname col\the\colcount\endcsname{\the\wd0}%
+      \fi
+    \fi
+    \ifx\go\pickupwholefraction
+      % Put the argument back for the \pickupwholefraction call, so
+      % we'll always have a period there to be parsed.
+      \def\go{\pickupwholefraction#1}%
+    \else
+      \let\go = \setuptable
+    \fi%
+  \fi
+  \go
+}
+
+% This used to have \hskip1sp.  But then the space in a template line is
+% not enough.  That is bad.  So let's go back to just & until we
+% encounter the problem it was intended to solve again.
+% --karl, nathan@acm.org, 20apr99.
+\def\tab{&}
+
+% @multitable ... @end multitable definitions:
+%
+\def\multitable{\parsearg\dotable}
+\def\dotable#1{\bgroup
+  \vskip\parskip
+  \let\item\crcr
+  \tolerance=9500
+  \hbadness=9500
+  \setmultitablespacing
+  \parskip=\multitableparskip
+  \parindent=\multitableparindent
+  \overfullrule=0pt
+  \global\colcount=0
+  \def\Emultitable{\global\setpercentfalse\cr\egroup\egroup}%
+  %
+  % To parse everything between @multitable and @item:
+  \setuptable#1 \endsetuptable
+  %
+  % \everycr will reset column counter, \colcount, at the end of
+  % each line. Every column entry will cause \colcount to advance by one.
+  % The table preamble
+  % looks at the current \colcount to find the correct column width.
+  \everycr{\noalign{%
+  %
+  % \filbreak%% keeps underfull box messages off when table breaks over pages.
+  % Maybe so, but it also creates really weird page breaks when the table
+  % breaks over pages. Wouldn't \vfil be better?  Wait until the problem
+  % manifests itself, so it can be fixed for real --karl.
+    \global\colcount=0\relax}}%
+  %
+  % This preamble sets up a generic column definition, which will
+  % be used as many times as user calls for columns.
+  % \vtop will set a single line and will also let text wrap and
+  % continue for many paragraphs if desired.
+  \halign\bgroup&\global\advance\colcount by 1\relax
+    \multistrut\vtop{\hsize=\expandafter\csname col\the\colcount\endcsname
+  %
+  % In order to keep entries from bumping into each other
+  % we will add a \leftskip of \multitablecolspace to all columns after
+  % the first one.
+  %
+  % If a template has been used, we will add \multitablecolspace
+  % to the width of each template entry.
+  %
+  % If the user has set preamble in terms of percent of \hsize we will
+  % use that dimension as the width of the column, and the \leftskip
+  % will keep entries from bumping into each other.  Table will start at
+  % left margin and final column will justify at right margin.
+  %
+  % Make sure we don't inherit \rightskip from the outer environment.
+  \rightskip=0pt
+  \ifnum\colcount=1
+    % The first column will be indented with the surrounding text.
+    \advance\hsize by\leftskip
+  \else
+    \ifsetpercent \else
+      % If user has not set preamble in terms of percent of \hsize
+      % we will advance \hsize by \multitablecolspace.
+      \advance\hsize by \multitablecolspace
+    \fi
+   % In either case we will make \leftskip=\multitablecolspace:
+  \leftskip=\multitablecolspace
+  \fi
+  % Ignoring space at the beginning and end avoids an occasional spurious
+  % blank line, when TeX decides to break the line at the space before the
+  % box from the multistrut, so the strut ends up on a line by itself.
+  % For example:
+  % @multitable @columnfractions .11 .89
+  % @item @code{#}
+  % @tab Legal holiday which is valid in major parts of the whole country.
+  % Is automatically provided with highlighting sequences respectively marking
+  % characters.
+  \noindent\ignorespaces##\unskip\multistrut}\cr
+}
+
+\def\setmultitablespacing{% test to see if user has set \multitablelinespace.
+% If so, do nothing. If not, give it an appropriate dimension based on
+% current baselineskip.
+\ifdim\multitablelinespace=0pt
+\setbox0=\vbox{X}\global\multitablelinespace=\the\baselineskip
+\global\advance\multitablelinespace by-\ht0
+%% strut to put in table in case some entry doesn't have descenders,
+%% to keep lines equally spaced
+\let\multistrut = \strut
+\else
+%% FIXME: what is \box0 supposed to be?
+\gdef\multistrut{\vrule height\multitablelinespace depth\dp0
+width0pt\relax} \fi
+%% Test to see if parskip is larger than space between lines of
+%% table. If not, do nothing.
+%%        If so, set to same dimension as multitablelinespace.
+\ifdim\multitableparskip>\multitablelinespace
+\global\multitableparskip=\multitablelinespace
+\global\advance\multitableparskip-7pt %% to keep parskip somewhat smaller
+                                      %% than skip between lines in the table.
+\fi%
+\ifdim\multitableparskip=0pt
+\global\multitableparskip=\multitablelinespace
+\global\advance\multitableparskip-7pt %% to keep parskip somewhat smaller
+                                      %% than skip between lines in the table.
+\fi}
+
+
+\message{conditionals,}
+% Prevent errors for section commands.
+% Used in @ignore and in failing conditionals.
+\def\ignoresections{%
+  \let\chapter=\relax
+  \let\unnumbered=\relax
+  \let\top=\relax
+  \let\unnumberedsec=\relax
+  \let\unnumberedsection=\relax
+  \let\unnumberedsubsec=\relax
+  \let\unnumberedsubsection=\relax
+  \let\unnumberedsubsubsec=\relax
+  \let\unnumberedsubsubsection=\relax
+  \let\section=\relax
+  \let\subsec=\relax
+  \let\subsubsec=\relax
+  \let\subsection=\relax
+  \let\subsubsection=\relax
+  \let\appendix=\relax
+  \let\appendixsec=\relax
+  \let\appendixsection=\relax
+  \let\appendixsubsec=\relax
+  \let\appendixsubsection=\relax
+  \let\appendixsubsubsec=\relax
+  \let\appendixsubsubsection=\relax
+  \let\contents=\relax
+  \let\smallbook=\relax
+  \let\titlepage=\relax
+}
+
+% Used in nested conditionals, where we have to parse the Texinfo source
+% and so want to turn off most commands, in case they are used
+% incorrectly.
+%
+\def\ignoremorecommands{%
+  \let\defcodeindex = \relax
+  \let\defcv = \relax
+  \let\deffn = \relax
+  \let\deffnx = \relax
+  \let\defindex = \relax
+  \let\defivar = \relax
+  \let\defmac = \relax
+  \let\defmethod = \relax
+  \let\defop = \relax
+  \let\defopt = \relax
+  \let\defspec = \relax
+  \let\deftp = \relax
+  \let\deftypefn = \relax
+  \let\deftypefun = \relax
+  \let\deftypeivar = \relax
+  \let\deftypeop = \relax
+  \let\deftypevar = \relax
+  \let\deftypevr = \relax
+  \let\defun = \relax
+  \let\defvar = \relax
+  \let\defvr = \relax
+  \let\ref = \relax
+  \let\xref = \relax
+  \let\printindex = \relax
+  \let\pxref = \relax
+  \let\settitle = \relax
+  \let\setchapternewpage = \relax
+  \let\setchapterstyle = \relax
+  \let\everyheading = \relax
+  \let\evenheading = \relax
+  \let\oddheading = \relax
+  \let\everyfooting = \relax
+  \let\evenfooting = \relax
+  \let\oddfooting = \relax
+  \let\headings = \relax
+  \let\include = \relax
+  \let\lowersections = \relax
+  \let\down = \relax
+  \let\raisesections = \relax
+  \let\up = \relax
+  \let\set = \relax
+  \let\clear = \relax
+  \let\item = \relax
+}
+
+% Ignore @ignore ... @end ignore.
+%
+\def\ignore{\doignore{ignore}}
+
+% Ignore @ifinfo, @ifhtml, @ifnottex, @html, @menu, and @direntry text.
+%
+\def\ifinfo{\doignore{ifinfo}}
+\def\ifhtml{\doignore{ifhtml}}
+\def\ifnottex{\doignore{ifnottex}}
+\def\html{\doignore{html}}
+\def\menu{\doignore{menu}}
+\def\direntry{\doignore{direntry}}
+
+% @dircategory CATEGORY  -- specify a category of the dir file
+% which this file should belong to.  Ignore this in TeX.
+\let\dircategory = \comment
+
+% Ignore text until a line `@end #1'.
+%
+\def\doignore#1{\begingroup
+  % Don't complain about control sequences we have declared \outer.
+  \ignoresections
+  %
+  % Define a command to swallow text until we reach `@end #1'.
+  % This @ is a catcode 12 token (that is the normal catcode of @ in
+  % this texinfo.tex file).  We change the catcode of @ below to match.
+  \long\def\doignoretext##1@end #1{\enddoignore}%
+  %
+  % Make sure that spaces turn into tokens that match what \doignoretext wants.
+  \catcode32 = 10
+  %
+  % Ignore braces, too, so mismatched braces don't cause trouble.
+  \catcode`\{ = 9
+  \catcode`\} = 9
+  %
+  % We must not have @c interpreted as a control sequence.
+  \catcode`\@ = 12
+  %
+  % Make the letter c a comment character so that the rest of the line
+  % will be ignored. This way, the document can have (for example)
+  %   @c @end ifinfo
+  % and the @end ifinfo will be properly ignored.
+  % (We've just changed @ to catcode 12.)
+  \catcode`\c = 14
+  %
+  % And now expand that command.
+  \doignoretext
+}
+
+% What we do to finish off ignored text.
+%
+\def\enddoignore{\endgroup\ignorespaces}%
+
+\newif\ifwarnedobs\warnedobsfalse
+\def\obstexwarn{%
+  \ifwarnedobs\relax\else
+  % We need to warn folks that they may have trouble with TeX 3.0.
+  % This uses \immediate\write16 rather than \message to get newlines.
+    \immediate\write16{}
+    \immediate\write16{WARNING: for users of Unix TeX 3.0!}
+    \immediate\write16{This manual trips a bug in TeX version 3.0 (tex hangs).}
+    \immediate\write16{If you are running another version of TeX, relax.}
+    \immediate\write16{If you are running Unix TeX 3.0, kill this TeX process.}
+    \immediate\write16{  Then upgrade your TeX installation if you can.}
+    \immediate\write16{  (See ftp://ftp.gnu.org/pub/gnu/TeX.README.)}
+    \immediate\write16{If you are stuck with version 3.0, run the}
+    \immediate\write16{  script ``tex3patch'' from the Texinfo distribution}
+    \immediate\write16{  to use a workaround.}
+    \immediate\write16{}
+    \global\warnedobstrue
+    \fi
+}
+
+% **In TeX 3.0, setting text in \nullfont hangs tex.  For a
+% workaround (which requires the file ``dummy.tfm'' to be installed),
+% uncomment the following line:
+%%%%%\font\nullfont=dummy\let\obstexwarn=\relax
+
+% Ignore text, except that we keep track of conditional commands for
+% purposes of nesting, up to an `@end #1' command.
+%
+\def\nestedignore#1{%
+  \obstexwarn
+  % We must actually expand the ignored text to look for the @end
+  % command, so that nested ignore constructs work.  Thus, we put the
+  % text into a \vbox and then do nothing with the result.  To minimize
+  % the change of memory overflow, we follow the approach outlined on
+  % page 401 of the TeXbook: make the current font be a dummy font.
+  %
+  \setbox0 = \vbox\bgroup
+    % Don't complain about control sequences we have declared \outer.
+    \ignoresections
+    %
+    % Define `@end #1' to end the box, which will in turn undefine the
+    % @end command again.
+    \expandafter\def\csname E#1\endcsname{\egroup\ignorespaces}%
+    %
+    % We are going to be parsing Texinfo commands.  Most cause no
+    % trouble when they are used incorrectly, but some commands do
+    % complicated argument parsing or otherwise get confused, so we
+    % undefine them.
+    %
+    % We can't do anything about stray @-signs, unfortunately;
+    % they'll produce `undefined control sequence' errors.
+    \ignoremorecommands
+    %
+    % Set the current font to be \nullfont, a TeX primitive, and define
+    % all the font commands to also use \nullfont.  We don't use
+    % dummy.tfm, as suggested in the TeXbook, because not all sites
+    % might have that installed.  Therefore, math mode will still
+    % produce output, but that should be an extremely small amount of
+    % stuff compared to the main input.
+    %
+    \nullfont
+    \let\tenrm=\nullfont \let\tenit=\nullfont \let\tensl=\nullfont
+    \let\tenbf=\nullfont \let\tentt=\nullfont \let\smallcaps=\nullfont
+    \let\tensf=\nullfont
+    % Similarly for index fonts (mostly for their use in smallexample).
+    \let\smallrm=\nullfont \let\smallit=\nullfont \let\smallsl=\nullfont
+    \let\smallbf=\nullfont \let\smalltt=\nullfont \let\smallsc=\nullfont
+    \let\smallsf=\nullfont
+    %
+    % Don't complain when characters are missing from the fonts.
+    \tracinglostchars = 0
+    %
+    % Don't bother to do space factor calculations.
+    \frenchspacing
+    %
+    % Don't report underfull hboxes.
+    \hbadness = 10000
+    %
+    % Do minimal line-breaking.
+    \pretolerance = 10000
+    %
+    % Do not execute instructions in @tex
+    \def\tex{\doignore{tex}}%
+    % Do not execute macro definitions.
+    % `c' is a comment character, so the word `macro' will get cut off.
+    \def\macro{\doignore{ma}}%
+}
+
+% @set VAR sets the variable VAR to an empty value.
+% @set VAR REST-OF-LINE sets VAR to the value REST-OF-LINE.
+%
+% Since we want to separate VAR from REST-OF-LINE (which might be
+% empty), we can't just use \parsearg; we have to insert a space of our
+% own to delimit the rest of the line, and then take it out again if we
+% didn't need it.  Make sure the catcode of space is correct to avoid
+% losing inside @example, for instance.
+%
+\def\set{\begingroup\catcode` =10
+  \catcode`\-=12 \catcode`\_=12 % Allow - and _ in VAR.
+  \parsearg\setxxx}
+\def\setxxx#1{\setyyy#1 \endsetyyy}
+\def\setyyy#1 #2\endsetyyy{%
+  \def\temp{#2}%
+  \ifx\temp\empty \global\expandafter\let\csname SET#1\endcsname = \empty
+  \else \setzzz{#1}#2\endsetzzz % Remove the trailing space \setxxx inserted.
+  \fi
+  \endgroup
+}
+% Can't use \xdef to pre-expand #2 and save some time, since \temp or
+% \next or other control sequences that we've defined might get us into
+% an infinite loop. Consider `@set foo @cite{bar}'.
+\def\setzzz#1#2 \endsetzzz{\expandafter\gdef\csname SET#1\endcsname{#2}}
+
+% @clear VAR clears (i.e., unsets) the variable VAR.
+%
+\def\clear{\parsearg\clearxxx}
+\def\clearxxx#1{\global\expandafter\let\csname SET#1\endcsname=\relax}
+
+% @value{foo} gets the text saved in variable foo.
+{
+  \catcode`\_ = \active
+  %
+  % We might end up with active _ or - characters in the argument if
+  % we're called from @code, as @code{@value{foo-bar_}}.  So \let any
+  % such active characters to their normal equivalents.
+  \gdef\value{\begingroup
+    \catcode`\-=12 \catcode`\_=12
+    \indexbreaks \let_\normalunderscore
+    \valuexxx}
+}
+\def\valuexxx#1{\expandablevalue{#1}\endgroup}
+
+% We have this subroutine so that we can handle at least some @value's
+% properly in indexes (we \let\value to this in \indexdummies).  Ones
+% whose names contain - or _ still won't work, but we can't do anything
+% about that.  The command has to be fully expandable, since the result
+% winds up in the index file.  This means that if the variable's value
+% contains other Texinfo commands, it's almost certain it will fail
+% (although perhaps we could fix that with sufficient work to do a
+% one-level expansion on the result, instead of complete).
+%
+\def\expandablevalue#1{%
+  \expandafter\ifx\csname SET#1\endcsname\relax
+    {[No value for ``#1'']}%
+  \else
+    \csname SET#1\endcsname
+  \fi
+}
+
+% @ifset VAR ... @end ifset reads the `...' iff VAR has been defined
+% with @set.
+%
+\def\ifset{\parsearg\ifsetxxx}
+\def\ifsetxxx #1{%
+  \expandafter\ifx\csname SET#1\endcsname\relax
+    \expandafter\ifsetfail
+  \else
+    \expandafter\ifsetsucceed
+  \fi
+}
+\def\ifsetsucceed{\conditionalsucceed{ifset}}
+\def\ifsetfail{\nestedignore{ifset}}
+\defineunmatchedend{ifset}
+
+% @ifclear VAR ... @end ifclear reads the `...' iff VAR has never been
+% defined with @set, or has been undefined with @clear.
+%
+\def\ifclear{\parsearg\ifclearxxx}
+\def\ifclearxxx #1{%
+  \expandafter\ifx\csname SET#1\endcsname\relax
+    \expandafter\ifclearsucceed
+  \else
+    \expandafter\ifclearfail
+  \fi
+}
+\def\ifclearsucceed{\conditionalsucceed{ifclear}}
+\def\ifclearfail{\nestedignore{ifclear}}
+\defineunmatchedend{ifclear}
+
+% @iftex, @ifnothtml, @ifnotinfo always succeed; we read the text
+% following, through the first @end iftex (etc.).  Make `@end iftex'
+% (etc.) valid only after an @iftex.
+%
+\def\iftex{\conditionalsucceed{iftex}}
+\def\ifnothtml{\conditionalsucceed{ifnothtml}}
+\def\ifnotinfo{\conditionalsucceed{ifnotinfo}}
+\defineunmatchedend{iftex}
+\defineunmatchedend{ifnothtml}
+\defineunmatchedend{ifnotinfo}
+
+% We can't just want to start a group at @iftex (for example) and end it
+% at @end iftex, since then @set commands inside the conditional have no
+% effect (they'd get reverted at the end of the group).  So we must
+% define \Eiftex to redefine itself to be its previous value.  (We can't
+% just define it to fail again with an ``unmatched end'' error, since
+% the @ifset might be nested.)
+%
+\def\conditionalsucceed#1{%
+  \edef\temp{%
+    % Remember the current value of \E#1.
+    \let\nece{prevE#1} = \nece{E#1}%
+    %
+    % At the `@end #1', redefine \E#1 to be its previous value.
+    \def\nece{E#1}{\let\nece{E#1} = \nece{prevE#1}}%
+  }%
+  \temp
+}
+
+% We need to expand lots of \csname's, but we don't want to expand the
+% control sequences after we've constructed them.
+%
+\def\nece#1{\expandafter\noexpand\csname#1\endcsname}
+
+% @defininfoenclose.
+\let\definfoenclose=\comment
+
+
+\message{indexing,}
+% Index generation facilities
+
+% Define \newwrite to be identical to plain tex's \newwrite
+% except not \outer, so it can be used within \newindex.
+{\catcode`\@=11
+\gdef\newwrite{\alloc@7\write\chardef\sixt@@n}}
+
+% \newindex {foo} defines an index named foo.
+% It automatically defines \fooindex such that
+% \fooindex ...rest of line... puts an entry in the index foo.
+% It also defines \fooindfile to be the number of the output channel for
+% the file that accumulates this index.  The file's extension is foo.
+% The name of an index should be no more than 2 characters long
+% for the sake of vms.
+%
+\def\newindex#1{%
+  \iflinks
+    \expandafter\newwrite \csname#1indfile\endcsname
+    \openout \csname#1indfile\endcsname \jobname.#1 % Open the file
+  \fi
+  \expandafter\xdef\csname#1index\endcsname{%     % Define @#1index
+    \noexpand\doindex{#1}}
+}
+
+% @defindex foo  ==  \newindex{foo}
+
+\def\defindex{\parsearg\newindex}
+
+% Define @defcodeindex, like @defindex except put all entries in @code.
+
+\def\newcodeindex#1{%
+  \iflinks
+    \expandafter\newwrite \csname#1indfile\endcsname
+    \openout \csname#1indfile\endcsname \jobname.#1
+  \fi
+  \expandafter\xdef\csname#1index\endcsname{%
+    \noexpand\docodeindex{#1}}
+}
+
+\def\defcodeindex{\parsearg\newcodeindex}
+
+% @synindex foo bar    makes index foo feed into index bar.
+% Do this instead of @defindex foo if you don't want it as a separate index.
+% The \closeout helps reduce unnecessary open files; the limit on the
+% Acorn RISC OS is a mere 16 files.
+\def\synindex#1 #2 {%
+  \expandafter\let\expandafter\synindexfoo\expandafter=\csname#2indfile\endcsname
+  \expandafter\closeout\csname#1indfile\endcsname
+  \expandafter\let\csname#1indfile\endcsname=\synindexfoo
+  \expandafter\xdef\csname#1index\endcsname{% define \xxxindex
+    \noexpand\doindex{#2}}%
+}
+
+% @syncodeindex foo bar   similar, but put all entries made for index foo
+% inside @code.
+\def\syncodeindex#1 #2 {%
+  \expandafter\let\expandafter\synindexfoo\expandafter=\csname#2indfile\endcsname
+  \expandafter\closeout\csname#1indfile\endcsname
+  \expandafter\let\csname#1indfile\endcsname=\synindexfoo
+  \expandafter\xdef\csname#1index\endcsname{% define \xxxindex
+    \noexpand\docodeindex{#2}}%
+}
+
+% Define \doindex, the driver for all \fooindex macros.
+% Argument #1 is generated by the calling \fooindex macro,
+%  and it is "foo", the name of the index.
+
+% \doindex just uses \parsearg; it calls \doind for the actual work.
+% This is because \doind is more useful to call from other macros.
+
+% There is also \dosubind {index}{topic}{subtopic}
+% which makes an entry in a two-level index such as the operation index.
+
+\def\doindex#1{\edef\indexname{#1}\parsearg\singleindexer}
+\def\singleindexer #1{\doind{\indexname}{#1}}
+
+% like the previous two, but they put @code around the argument.
+\def\docodeindex#1{\edef\indexname{#1}\parsearg\singlecodeindexer}
+\def\singlecodeindexer #1{\doind{\indexname}{\code{#1}}}
+
+\def\indexdummies{%
+\def\ { }%
+% Take care of the plain tex accent commands.
+\def\"{\realbackslash "}%
+\def\`{\realbackslash `}%
+\def\'{\realbackslash '}%
+\def\^{\realbackslash ^}%
+\def\~{\realbackslash ~}%
+\def\={\realbackslash =}%
+\def\b{\realbackslash b}%
+\def\c{\realbackslash c}%
+\def\d{\realbackslash d}%
+\def\u{\realbackslash u}%
+\def\v{\realbackslash v}%
+\def\H{\realbackslash H}%
+% Take care of the plain tex special European modified letters.
+\def\oe{\realbackslash oe}%
+\def\ae{\realbackslash ae}%
+\def\aa{\realbackslash aa}%
+\def\OE{\realbackslash OE}%
+\def\AE{\realbackslash AE}%
+\def\AA{\realbackslash AA}%
+\def\o{\realbackslash o}%
+\def\O{\realbackslash O}%
+\def\l{\realbackslash l}%
+\def\L{\realbackslash L}%
+\def\ss{\realbackslash ss}%
+% Take care of texinfo commands likely to appear in an index entry.
+% (Must be a way to avoid doing expansion at all, and thus not have to
+% laboriously list every single command here.)
+\def\@{@}% will be @@ when we switch to @ as escape char.
+% Need these in case \tex is in effect and \{ is a \delimiter again.
+% But can't use \lbracecmd and \rbracecmd because texindex assumes
+% braces and backslashes are used only as delimiters.  
+\let\{ = \mylbrace
+\let\} = \myrbrace
+\def\_{{\realbackslash _}}%
+\def\w{\realbackslash w }%
+\def\bf{\realbackslash bf }%
+%\def\rm{\realbackslash rm }%
+\def\sl{\realbackslash sl }%
+\def\sf{\realbackslash sf}%
+\def\tt{\realbackslash tt}%
+\def\gtr{\realbackslash gtr}%
+\def\less{\realbackslash less}%
+\def\hat{\realbackslash hat}%
+\def\TeX{\realbackslash TeX}%
+\def\dots{\realbackslash dots }%
+\def\result{\realbackslash result}%
+\def\equiv{\realbackslash equiv}%
+\def\expansion{\realbackslash expansion}%
+\def\print{\realbackslash print}%
+\def\error{\realbackslash error}%
+\def\point{\realbackslash point}%
+\def\copyright{\realbackslash copyright}%
+\def\tclose##1{\realbackslash tclose {##1}}%
+\def\code##1{\realbackslash code {##1}}%
+\def\uref##1{\realbackslash uref {##1}}%
+\def\url##1{\realbackslash url {##1}}%
+\def\env##1{\realbackslash env {##1}}%
+\def\command##1{\realbackslash command {##1}}%
+\def\option##1{\realbackslash option {##1}}%
+\def\dotless##1{\realbackslash dotless {##1}}%
+\def\samp##1{\realbackslash samp {##1}}%
+\def\,##1{\realbackslash ,{##1}}%
+\def\t##1{\realbackslash t {##1}}%
+\def\r##1{\realbackslash r {##1}}%
+\def\i##1{\realbackslash i {##1}}%
+\def\b##1{\realbackslash b {##1}}%
+\def\sc##1{\realbackslash sc {##1}}%
+\def\cite##1{\realbackslash cite {##1}}%
+\def\key##1{\realbackslash key {##1}}%
+\def\file##1{\realbackslash file {##1}}%
+\def\var##1{\realbackslash var {##1}}%
+\def\kbd##1{\realbackslash kbd {##1}}%
+\def\dfn##1{\realbackslash dfn {##1}}%
+\def\emph##1{\realbackslash emph {##1}}%
+\def\acronym##1{\realbackslash acronym {##1}}%
+%
+% Handle some cases of @value -- where the variable name does not
+% contain - or _, and the value does not contain any
+% (non-fully-expandable) commands.
+\let\value = \expandablevalue
+%
+\unsepspaces
+% Turn off macro expansion
+\turnoffmacros
+}
+
+% If an index command is used in an @example environment, any spaces
+% therein should become regular spaces in the raw index file, not the
+% expansion of \tie (\\leavevmode \penalty \@M \ ).
+{\obeyspaces
+ \gdef\unsepspaces{\obeyspaces\let =\space}}
+
+% \indexnofonts no-ops all font-change commands.
+% This is used when outputting the strings to sort the index by.
+\def\indexdummyfont#1{#1}
+\def\indexdummytex{TeX}
+\def\indexdummydots{...}
+
+\def\indexnofonts{%
+% Just ignore accents.
+\let\,=\indexdummyfont
+\let\"=\indexdummyfont
+\let\`=\indexdummyfont
+\let\'=\indexdummyfont
+\let\^=\indexdummyfont
+\let\~=\indexdummyfont
+\let\==\indexdummyfont
+\let\b=\indexdummyfont
+\let\c=\indexdummyfont
+\let\d=\indexdummyfont
+\let\u=\indexdummyfont
+\let\v=\indexdummyfont
+\let\H=\indexdummyfont
+\let\dotless=\indexdummyfont
+% Take care of the plain tex special European modified letters.
+\def\oe{oe}%
+\def\ae{ae}%
+\def\aa{aa}%
+\def\OE{OE}%
+\def\AE{AE}%
+\def\AA{AA}%
+\def\o{o}%
+\def\O{O}%
+\def\l{l}%
+\def\L{L}%
+\def\ss{ss}%
+\let\w=\indexdummyfont
+\let\t=\indexdummyfont
+\let\r=\indexdummyfont
+\let\i=\indexdummyfont
+\let\b=\indexdummyfont
+\let\emph=\indexdummyfont
+\let\strong=\indexdummyfont
+\let\cite=\indexdummyfont
+\let\sc=\indexdummyfont
+%Don't no-op \tt, since it isn't a user-level command
+% and is used in the definitions of the active chars like <, >, |...
+%\let\tt=\indexdummyfont
+\let\tclose=\indexdummyfont
+\let\code=\indexdummyfont
+\let\url=\indexdummyfont
+\let\uref=\indexdummyfont
+\let\env=\indexdummyfont
+\let\acronym=\indexdummyfont
+\let\command=\indexdummyfont
+\let\option=\indexdummyfont
+\let\file=\indexdummyfont
+\let\samp=\indexdummyfont
+\let\kbd=\indexdummyfont
+\let\key=\indexdummyfont
+\let\var=\indexdummyfont
+\let\TeX=\indexdummytex
+\let\dots=\indexdummydots
+\def\@{@}%
+}
+
+% To define \realbackslash, we must make \ not be an escape.
+% We must first make another character (@) an escape
+% so we do not become unable to do a definition.
+
+{\catcode`\@=0 \catcode`\\=\other
+ @gdef@realbackslash{\}}
+
+\let\indexbackslash=0  %overridden during \printindex.
+\let\SETmarginindex=\relax % put index entries in margin (undocumented)?
+
+% For \ifx comparisons.
+\def\emptymacro{\empty}
+
+% Most index entries go through here, but \dosubind is the general case.
+%
+\def\doind#1#2{\dosubind{#1}{#2}\empty}
+
+% Workhorse for all \fooindexes.
+% #1 is name of index, #2 is stuff to put there, #3 is subentry --
+% \empty if called from \doind, as we usually are.  The main exception
+% is with defuns, which call us directly.
+%
+\def\dosubind#1#2#3{%
+  % Put the index entry in the margin if desired.
+  \ifx\SETmarginindex\relax\else
+    \insert\margin{\hbox{\vrule height8pt depth3pt width0pt #2}}%
+  \fi
+  {%
+    \count255=\lastpenalty
+    {%
+      \indexdummies % Must do this here, since \bf, etc expand at this stage
+      \escapechar=`\\
+      {%
+        \let\folio = 0% We will expand all macros now EXCEPT \folio.
+        \def\rawbackslashxx{\indexbackslash}% \indexbackslash isn't defined now
+        % so it will be output as is; and it will print as backslash.
+        %
+        \def\thirdarg{#3}%
+        %
+        % If third arg is present, precede it with space in sort key.
+        \ifx\thirdarg\emptymacro
+          \let\subentry = \empty
+        \else
+          \def\subentry{ #3}%
+        \fi
+        %
+        % First process the index entry with all font commands turned
+        % off to get the string to sort by.
+        {\indexnofonts \xdef\indexsorttmp{#2\subentry}}%
+        %
+        % Now the real index entry with the fonts.
+        \toks0 = {#2}%
+        %
+        % If third (subentry) arg is present, add it to the index
+        % string.  And include a space.
+        \ifx\thirdarg\emptymacro \else
+          \toks0 = \expandafter{\the\toks0 \space #3}%
+        \fi
+        %
+        % Set up the complete index entry, with both the sort key
+        % and the original text, including any font commands.  We write
+        % three arguments to \entry to the .?? file, texindex reduces to
+        % two when writing the .??s sorted result.
+        \edef\temp{%
+          \write\csname#1indfile\endcsname{%
+            \realbackslash entry{\indexsorttmp}{\folio}{\the\toks0}}%
+        }%
+        %
+        % If a skip is the last thing on the list now, preserve it
+        % by backing up by \lastskip, doing the \write, then inserting
+        % the skip again.  Otherwise, the whatsit generated by the
+        % \write will make \lastskip zero.  The result is that sequences
+        % like this:
+        % @end defun
+        % @tindex whatever
+        % @defun ...
+        % will have extra space inserted, because the \medbreak in the
+        % start of the @defun won't see the skip inserted by the @end of
+        % the previous defun.
+        %
+        % But don't do any of this if we're not in vertical mode.  We
+        % don't want to do a \vskip and prematurely end a paragraph.
+        %
+        % Avoid page breaks due to these extra skips, too.
+        %
+        \iflinks
+          \ifvmode
+            \skip0 = \lastskip
+            \ifdim\lastskip = 0pt \else \nobreak\vskip-\lastskip \fi
+          \fi
+          %
+          \temp % do the write
+          %
+          %
+          \ifvmode \ifdim\skip0 = 0pt \else \nobreak\vskip\skip0 \fi \fi
+        \fi
+      }%
+    }%
+    \penalty\count255
+  }%
+}
+
+% The index entry written in the file actually looks like
+%  \entry {sortstring}{page}{topic}
+% or
+%  \entry {sortstring}{page}{topic}{subtopic}
+% The texindex program reads in these files and writes files
+% containing these kinds of lines:
+%  \initial {c}
+%     before the first topic whose initial is c
+%  \entry {topic}{pagelist}
+%     for a topic that is used without subtopics
+%  \primary {topic}
+%     for the beginning of a topic that is used with subtopics
+%  \secondary {subtopic}{pagelist}
+%     for each subtopic.
+
+% Define the user-accessible indexing commands
+% @findex, @vindex, @kindex, @cindex.
+
+\def\findex {\fnindex}
+\def\kindex {\kyindex}
+\def\cindex {\cpindex}
+\def\vindex {\vrindex}
+\def\tindex {\tpindex}
+\def\pindex {\pgindex}
+
+\def\cindexsub {\begingroup\obeylines\cindexsub}
+{\obeylines %
+\gdef\cindexsub "#1" #2^^M{\endgroup %
+\dosubind{cp}{#2}{#1}}}
+
+% Define the macros used in formatting output of the sorted index material.
+
+% @printindex causes a particular index (the ??s file) to get printed.
+% It does not print any chapter heading (usually an @unnumbered).
+%
+\def\printindex{\parsearg\doprintindex}
+\def\doprintindex#1{\begingroup
+  \dobreak \chapheadingskip{10000}%
+  %
+  \smallfonts \rm
+  \tolerance = 9500
+  \indexbreaks
+  %
+  % See if the index file exists and is nonempty.
+  % Change catcode of @ here so that if the index file contains
+  % \initial {@}
+  % as its first line, TeX doesn't complain about mismatched braces
+  % (because it thinks @} is a control sequence).
+  \catcode`\@ = 11
+  \openin 1 \jobname.#1s
+  \ifeof 1
+    % \enddoublecolumns gets confused if there is no text in the index,
+    % and it loses the chapter title and the aux file entries for the
+    % index.  The easiest way to prevent this problem is to make sure
+    % there is some text.
+    \putwordIndexNonexistent
+  \else
+    %
+    % If the index file exists but is empty, then \openin leaves \ifeof
+    % false.  We have to make TeX try to read something from the file, so
+    % it can discover if there is anything in it.
+    \read 1 to \temp
+    \ifeof 1
+      \putwordIndexIsEmpty
+    \else
+      % Index files are almost Texinfo source, but we use \ as the escape
+      % character.  It would be better to use @, but that's too big a change
+      % to make right now.
+      \def\indexbackslash{\rawbackslashxx}%
+      \catcode`\\ = 0
+      \escapechar = `\\
+      \begindoublecolumns
+      \input \jobname.#1s
+      \enddoublecolumns
+    \fi
+  \fi
+  \closein 1
+\endgroup}
+
+% These macros are used by the sorted index file itself.
+% Change them to control the appearance of the index.
+
+\def\initial#1{{%
+  % Some minor font changes for the special characters.
+  \let\tentt=\sectt \let\tt=\sectt \let\sf=\sectt
+  %
+  % Remove any glue we may have, we'll be inserting our own.
+  \removelastskip
+  %
+  % We like breaks before the index initials, so insert a bonus.
+  \penalty -300
+  %
+  % Typeset the initial.  Making this add up to a whole number of
+  % baselineskips increases the chance of the dots lining up from column
+  % to column.  It still won't often be perfect, because of the stretch
+  % we need before each entry, but it's better.
+  %
+  % No shrink because it confuses \balancecolumns.
+  \vskip 1.67\baselineskip plus .5\baselineskip
+  \leftline{\secbf #1}%
+  \vskip .33\baselineskip plus .1\baselineskip
+  %
+  % Do our best not to break after the initial.
+  \nobreak
+}}
+
+% This typesets a paragraph consisting of #1, dot leaders, and then #2
+% flush to the right margin.  It is used for index and table of contents
+% entries.  The paragraph is indented by \leftskip.
+%
+\def\entry#1#2{\begingroup
+  %
+  % Start a new paragraph if necessary, so our assignments below can't
+  % affect previous text.
+  \par
+  %
+  % Do not fill out the last line with white space.
+  \parfillskip = 0in
+  %
+  % No extra space above this paragraph.
+  \parskip = 0in
+  %
+  % Do not prefer a separate line ending with a hyphen to fewer lines.
+  \finalhyphendemerits = 0
+  %
+  % \hangindent is only relevant when the entry text and page number
+  % don't both fit on one line.  In that case, bob suggests starting the
+  % dots pretty far over on the line.  Unfortunately, a large
+  % indentation looks wrong when the entry text itself is broken across
+  % lines.  So we use a small indentation and put up with long leaders.
+  %
+  % \hangafter is reset to 1 (which is the value we want) at the start
+  % of each paragraph, so we need not do anything with that.
+  \hangindent = 2em
+  %
+  % When the entry text needs to be broken, just fill out the first line
+  % with blank space.
+  \rightskip = 0pt plus1fil
+  %
+  % A bit of stretch before each entry for the benefit of balancing columns.
+  \vskip 0pt plus1pt
+  %
+  % Start a ``paragraph'' for the index entry so the line breaking
+  % parameters we've set above will have an effect.
+  \noindent
+  %
+  % Insert the text of the index entry.  TeX will do line-breaking on it.
+  #1%
+  % The following is kludged to not output a line of dots in the index if
+  % there are no page numbers.  The next person who breaks this will be
+  % cursed by a Unix daemon.
+  \def\tempa{{\rm }}%
+  \def\tempb{#2}%
+  \edef\tempc{\tempa}%
+  \edef\tempd{\tempb}%
+  \ifx\tempc\tempd\ \else%
+    %
+    % If we must, put the page number on a line of its own, and fill out
+    % this line with blank space.  (The \hfil is overwhelmed with the
+    % fill leaders glue in \indexdotfill if the page number does fit.)
+    \hfil\penalty50
+    \null\nobreak\indexdotfill % Have leaders before the page number.
+    %
+    % The `\ ' here is removed by the implicit \unskip that TeX does as
+    % part of (the primitive) \par.  Without it, a spurious underfull
+    % \hbox ensues.
+    \ifpdf
+      \pdfgettoks#2.\ \the\toksA % The page number ends the paragraph.
+    \else
+      \ #2% The page number ends the paragraph.
+    \fi
+  \fi%
+  \par
+\endgroup}
+
+% Like \dotfill except takes at least 1 em.
+\def\indexdotfill{\cleaders
+  \hbox{$\mathsurround=0pt \mkern1.5mu ${\it .}$ \mkern1.5mu$}\hskip 1em plus 1fill}
+
+\def\primary #1{\line{#1\hfil}}
+
+\newskip\secondaryindent \secondaryindent=0.5cm
+
+\def\secondary #1#2{
+{\parfillskip=0in \parskip=0in
+\hangindent =1in \hangafter=1
+\noindent\hskip\secondaryindent\hbox{#1}\indexdotfill #2\par
+}}
+
+% Define two-column mode, which we use to typeset indexes.
+% Adapted from the TeXbook, page 416, which is to say,
+% the manmac.tex format used to print the TeXbook itself.
+\catcode`\@=11
+
+\newbox\partialpage
+\newdimen\doublecolumnhsize
+
+\def\begindoublecolumns{\begingroup % ended by \enddoublecolumns
+  % Grab any single-column material above us.
+  \output = {%
+    %
+    % Here is a possibility not foreseen in manmac: if we accumulate a
+    % whole lot of material, we might end up calling this \output
+    % routine twice in a row (see the doublecol-lose test, which is
+    % essentially a couple of indexes with @setchapternewpage off).  In
+    % that case we just ship out what is in \partialpage with the normal
+    % output routine.  Generally, \partialpage will be empty when this
+    % runs and this will be a no-op.  See the indexspread.tex test case.
+    \ifvoid\partialpage \else
+      \onepageout{\pagecontents\partialpage}%
+    \fi
+    %
+    \global\setbox\partialpage = \vbox{%
+      % Unvbox the main output page.
+      \unvbox\PAGE
+      \kern-\topskip \kern\baselineskip
+    }%
+  }%
+  \eject % run that output routine to set \partialpage
+  %
+  % Use the double-column output routine for subsequent pages.
+  \output = {\doublecolumnout}%
+  %
+  % Change the page size parameters.  We could do this once outside this
+  % routine, in each of @smallbook, @afourpaper, and the default 8.5x11
+  % format, but then we repeat the same computation.  Repeating a couple
+  % of assignments once per index is clearly meaningless for the
+  % execution time, so we may as well do it in one place.
+  %
+  % First we halve the line length, less a little for the gutter between
+  % the columns.  We compute the gutter based on the line length, so it
+  % changes automatically with the paper format.  The magic constant
+  % below is chosen so that the gutter has the same value (well, +-<1pt)
+  % as it did when we hard-coded it.
+  %
+  % We put the result in a separate register, \doublecolumhsize, so we
+  % can restore it in \pagesofar, after \hsize itself has (potentially)
+  % been clobbered.
+  %
+  \doublecolumnhsize = \hsize
+    \advance\doublecolumnhsize by -.04154\hsize
+    \divide\doublecolumnhsize by 2
+  \hsize = \doublecolumnhsize
+  %
+  % Double the \vsize as well.  (We don't need a separate register here,
+  % since nobody clobbers \vsize.)
+  \advance\vsize by -\ht\partialpage
+  \vsize = 2\vsize
+}
+
+% The double-column output routine for all double-column pages except
+% the last.
+%
+\def\doublecolumnout{%
+  \splittopskip=\topskip \splitmaxdepth=\maxdepth
+  % Get the available space for the double columns -- the normal
+  % (undoubled) page height minus any material left over from the
+  % previous page.
+  \dimen@ = \vsize
+  \divide\dimen@ by 2
+  %
+  % box0 will be the left-hand column, box2 the right.
+  \setbox0=\vsplit255 to\dimen@ \setbox2=\vsplit255 to\dimen@
+  \onepageout\pagesofar
+  \unvbox255
+  \penalty\outputpenalty
+}
+\def\pagesofar{%
+  % Re-output the contents of the output page -- any previous material,
+  % followed by the two boxes we just split, in box0 and box2.
+  \unvbox\partialpage
+  %
+  \hsize = \doublecolumnhsize
+  \wd0=\hsize \wd2=\hsize
+  \hbox to\pagewidth{\box0\hfil\box2}%
+}
+\def\enddoublecolumns{%
+  \output = {%
+    % Split the last of the double-column material.  Leave it on the
+    % current page, no automatic page break.
+    \balancecolumns
+    %
+    % If we end up splitting too much material for the current page,
+    % though, there will be another page break right after this \output
+    % invocation ends.  Having called \balancecolumns once, we do not
+    % want to call it again.  Therefore, reset \output to its normal
+    % definition right away.  (We hope \balancecolumns will never be
+    % called on to balance too much material, but if it is, this makes
+    % the output somewhat more palatable.)
+    \global\output = {\onepageout{\pagecontents\PAGE}}%
+  }%
+  \eject
+  \endgroup % started in \begindoublecolumns
+  %
+  % \pagegoal was set to the doubled \vsize above, since we restarted
+  % the current page.  We're now back to normal single-column
+  % typesetting, so reset \pagegoal to the normal \vsize (after the
+  % \endgroup where \vsize got restored).
+  \pagegoal = \vsize
+}
+\def\balancecolumns{%
+  % Called at the end of the double column material.
+  \setbox0 = \vbox{\unvbox255}% like \box255 but more efficient, see p.120.
+  \dimen@ = \ht0
+  \advance\dimen@ by \topskip
+  \advance\dimen@ by-\baselineskip
+  \divide\dimen@ by 2 % target to split to
+  %debug\message{final 2-column material height=\the\ht0, target=\the\dimen@.}%
+  \splittopskip = \topskip
+  % Loop until we get a decent breakpoint.
+  {%
+    \vbadness = 10000
+    \loop
+      \global\setbox3 = \copy0
+      \global\setbox1 = \vsplit3 to \dimen@
+    \ifdim\ht3>\dimen@
+      \global\advance\dimen@ by 1pt
+    \repeat
+  }%
+  %debug\message{split to \the\dimen@, column heights: \the\ht1, \the\ht3.}%
+  \setbox0=\vbox to\dimen@{\unvbox1}%
+  \setbox2=\vbox to\dimen@{\unvbox3}%
+  %
+  \pagesofar
+}
+\catcode`\@ = \other
+
+
+\message{sectioning,}
+% Chapters, sections, etc.
+
+\newcount\chapno
+\newcount\secno        \secno=0
+\newcount\subsecno     \subsecno=0
+\newcount\subsubsecno  \subsubsecno=0
+
+% This counter is funny since it counts through charcodes of letters A, B, ...
+\newcount\appendixno  \appendixno = `\@
+% \def\appendixletter{\char\the\appendixno}
+% We do the following for the sake of pdftex, which needs the actual
+% letter in the expansion, not just typeset.
+\def\appendixletter{%
+  \ifnum\appendixno=`A A%
+  \else\ifnum\appendixno=`B B%
+  \else\ifnum\appendixno=`C C%
+  \else\ifnum\appendixno=`D D%
+  \else\ifnum\appendixno=`E E%
+  \else\ifnum\appendixno=`F F%
+  \else\ifnum\appendixno=`G G%
+  \else\ifnum\appendixno=`H H%
+  \else\ifnum\appendixno=`I I%
+  \else\ifnum\appendixno=`J J%
+  \else\ifnum\appendixno=`K K%
+  \else\ifnum\appendixno=`L L%
+  \else\ifnum\appendixno=`M M%
+  \else\ifnum\appendixno=`N N%
+  \else\ifnum\appendixno=`O O%
+  \else\ifnum\appendixno=`P P%
+  \else\ifnum\appendixno=`Q Q%
+  \else\ifnum\appendixno=`R R%
+  \else\ifnum\appendixno=`S S%
+  \else\ifnum\appendixno=`T T%
+  \else\ifnum\appendixno=`U U%
+  \else\ifnum\appendixno=`V V%
+  \else\ifnum\appendixno=`W W%
+  \else\ifnum\appendixno=`X X%
+  \else\ifnum\appendixno=`Y Y%
+  \else\ifnum\appendixno=`Z Z%
+  % The \the is necessary, despite appearances, because \appendixletter is
+  % expanded while writing the .toc file.  \char\appendixno is not
+  % expandable, thus it is written literally, thus all appendixes come out
+  % with the same letter (or @) in the toc without it.
+  \else\char\the\appendixno
+  \fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi
+  \fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi}
+
+% Each @chapter defines this as the name of the chapter.
+% page headings and footings can use it.  @section does likewise.
+\def\thischapter{}
+\def\thissection{}
+
+\newcount\absseclevel % used to calculate proper heading level
+\newcount\secbase\secbase=0 % @raise/lowersections modify this count
+
+% @raisesections: treat @section as chapter, @subsection as section, etc.
+\def\raisesections{\global\advance\secbase by -1}
+\let\up=\raisesections % original BFox name
+
+% @lowersections: treat @chapter as section, @section as subsection, etc.
+\def\lowersections{\global\advance\secbase by 1}
+\let\down=\lowersections % original BFox name
+
+% Choose a numbered-heading macro
+% #1 is heading level if unmodified by @raisesections or @lowersections
+% #2 is text for heading
+\def\numhead#1#2{\absseclevel=\secbase\advance\absseclevel by #1
+\ifcase\absseclevel
+  \chapterzzz{#2}
+\or
+  \seczzz{#2}
+\or
+  \numberedsubseczzz{#2}
+\or
+  \numberedsubsubseczzz{#2}
+\else
+  \ifnum \absseclevel<0
+    \chapterzzz{#2}
+  \else
+    \numberedsubsubseczzz{#2}
+  \fi
+\fi
+}
+
+% like \numhead, but chooses appendix heading levels
+\def\apphead#1#2{\absseclevel=\secbase\advance\absseclevel by #1
+\ifcase\absseclevel
+  \appendixzzz{#2}
+\or
+  \appendixsectionzzz{#2}
+\or
+  \appendixsubseczzz{#2}
+\or
+  \appendixsubsubseczzz{#2}
+\else
+  \ifnum \absseclevel<0
+    \appendixzzz{#2}
+  \else
+    \appendixsubsubseczzz{#2}
+  \fi
+\fi
+}
+
+% like \numhead, but chooses numberless heading levels
+\def\unnmhead#1#2{\absseclevel=\secbase\advance\absseclevel by #1
+\ifcase\absseclevel
+  \unnumberedzzz{#2}
+\or
+  \unnumberedseczzz{#2}
+\or
+  \unnumberedsubseczzz{#2}
+\or
+  \unnumberedsubsubseczzz{#2}
+\else
+  \ifnum \absseclevel<0
+    \unnumberedzzz{#2}
+  \else
+    \unnumberedsubsubseczzz{#2}
+  \fi
+\fi
+}
+
+% @chapter, @appendix, @unnumbered.
+\def\thischaptername{No Chapter Title}
+\outer\def\chapter{\parsearg\chapteryyy}
+\def\chapteryyy #1{\numhead0{#1}} % normally numhead0 calls chapterzzz
+\def\chapterzzz #1{%
+\secno=0 \subsecno=0 \subsubsecno=0
+\global\advance \chapno by 1 \message{\putwordChapter\space \the\chapno}%
+\chapmacro {#1}{\the\chapno}%
+\gdef\thissection{#1}%
+\gdef\thischaptername{#1}%
+% We don't substitute the actual chapter name into \thischapter
+% because we don't want its macros evaluated now.
+\xdef\thischapter{\putwordChapter{} \the\chapno: \noexpand\thischaptername}%
+\toks0 = {#1}%
+\edef\temp{\noexpand\writetocentry{\realbackslash chapentry{\the\toks0}%
+                                  {\the\chapno}}}%
+\temp
+\donoderef
+\global\let\section = \numberedsec
+\global\let\subsection = \numberedsubsec
+\global\let\subsubsection = \numberedsubsubsec
+}
+
+\outer\def\appendix{\parsearg\appendixyyy}
+\def\appendixyyy #1{\apphead0{#1}} % normally apphead0 calls appendixzzz
+\def\appendixzzz #1{%
+\secno=0 \subsecno=0 \subsubsecno=0
+\global\advance \appendixno by 1
+\message{\putwordAppendix\space \appendixletter}%
+\chapmacro {#1}{\putwordAppendix{} \appendixletter}%
+\gdef\thissection{#1}%
+\gdef\thischaptername{#1}%
+\xdef\thischapter{\putwordAppendix{} \appendixletter: \noexpand\thischaptername}%
+\toks0 = {#1}%
+\edef\temp{\noexpand\writetocentry{\realbackslash chapentry{\the\toks0}%
+                       {\putwordAppendix{} \appendixletter}}}%
+\temp
+\appendixnoderef
+\global\let\section = \appendixsec
+\global\let\subsection = \appendixsubsec
+\global\let\subsubsection = \appendixsubsubsec
+}
+
+% @centerchap is like @unnumbered, but the heading is centered.
+\outer\def\centerchap{\parsearg\centerchapyyy}
+\def\centerchapyyy #1{{\let\unnumbchapmacro=\centerchapmacro \unnumberedyyy{#1}}}
+
+% @top is like @unnumbered.
+\outer\def\top{\parsearg\unnumberedyyy}
+
+\outer\def\unnumbered{\parsearg\unnumberedyyy}
+\def\unnumberedyyy #1{\unnmhead0{#1}} % normally unnmhead0 calls unnumberedzzz
+\def\unnumberedzzz #1{%
+\secno=0 \subsecno=0 \subsubsecno=0
+%
+% This used to be simply \message{#1}, but TeX fully expands the
+% argument to \message.  Therefore, if #1 contained @-commands, TeX
+% expanded them.  For example, in `@unnumbered The @cite{Book}', TeX
+% expanded @cite (which turns out to cause errors because \cite is meant
+% to be executed, not expanded).
+%
+% Anyway, we don't want the fully-expanded definition of @cite to appear
+% as a result of the \message, we just want `@cite' itself.  We use
+% \the<toks register> to achieve this: TeX expands \the<toks> only once,
+% simply yielding the contents of <toks register>.  (We also do this for
+% the toc entries.)
+\toks0 = {#1}\message{(\the\toks0)}%
+%
+\unnumbchapmacro {#1}%
+\gdef\thischapter{#1}\gdef\thissection{#1}%
+\toks0 = {#1}%
+\edef\temp{\noexpand\writetocentry{\realbackslash unnumbchapentry{\the\toks0}}}%
+\temp
+\unnumbnoderef
+\global\let\section = \unnumberedsec
+\global\let\subsection = \unnumberedsubsec
+\global\let\subsubsection = \unnumberedsubsubsec
+}
+
+% Sections.
+\outer\def\numberedsec{\parsearg\secyyy}
+\def\secyyy #1{\numhead1{#1}} % normally calls seczzz
+\def\seczzz #1{%
+\subsecno=0 \subsubsecno=0 \global\advance \secno by 1 %
+\gdef\thissection{#1}\secheading {#1}{\the\chapno}{\the\secno}%
+\toks0 = {#1}%
+\edef\temp{\noexpand\writetocentry{\realbackslash secentry{\the\toks0}%
+                                  {\the\chapno}{\the\secno}}}%
+\temp
+\donoderef
+\nobreak
+}
+
+\outer\def\appendixsection{\parsearg\appendixsecyyy}
+\outer\def\appendixsec{\parsearg\appendixsecyyy}
+\def\appendixsecyyy #1{\apphead1{#1}} % normally calls appendixsectionzzz
+\def\appendixsectionzzz #1{%
+\subsecno=0 \subsubsecno=0 \global\advance \secno by 1 %
+\gdef\thissection{#1}\secheading {#1}{\appendixletter}{\the\secno}%
+\toks0 = {#1}%
+\edef\temp{\noexpand\writetocentry{\realbackslash secentry{\the\toks0}%
+                                  {\appendixletter}{\the\secno}}}%
+\temp
+\appendixnoderef
+\nobreak
+}
+
+\outer\def\unnumberedsec{\parsearg\unnumberedsecyyy}
+\def\unnumberedsecyyy #1{\unnmhead1{#1}} % normally calls unnumberedseczzz
+\def\unnumberedseczzz #1{%
+\plainsecheading {#1}\gdef\thissection{#1}%
+\toks0 = {#1}%
+\edef\temp{\noexpand\writetocentry{\realbackslash unnumbsecentry{\the\toks0}}}%
+\temp
+\unnumbnoderef
+\nobreak
+}
+
+% Subsections.
+\outer\def\numberedsubsec{\parsearg\numberedsubsecyyy}
+\def\numberedsubsecyyy #1{\numhead2{#1}} % normally calls numberedsubseczzz
+\def\numberedsubseczzz #1{%
+\gdef\thissection{#1}\subsubsecno=0 \global\advance \subsecno by 1 %
+\subsecheading {#1}{\the\chapno}{\the\secno}{\the\subsecno}%
+\toks0 = {#1}%
+\edef\temp{\noexpand\writetocentry{\realbackslash subsecentry{\the\toks0}%
+                                    {\the\chapno}{\the\secno}{\the\subsecno}}}%
+\temp
+\donoderef
+\nobreak
+}
+
+\outer\def\appendixsubsec{\parsearg\appendixsubsecyyy}
+\def\appendixsubsecyyy #1{\apphead2{#1}} % normally calls appendixsubseczzz
+\def\appendixsubseczzz #1{%
+\gdef\thissection{#1}\subsubsecno=0 \global\advance \subsecno by 1 %
+\subsecheading {#1}{\appendixletter}{\the\secno}{\the\subsecno}%
+\toks0 = {#1}%
+\edef\temp{\noexpand\writetocentry{\realbackslash subsecentry{\the\toks0}%
+                                {\appendixletter}{\the\secno}{\the\subsecno}}}%
+\temp
+\appendixnoderef
+\nobreak
+}
+
+\outer\def\unnumberedsubsec{\parsearg\unnumberedsubsecyyy}
+\def\unnumberedsubsecyyy #1{\unnmhead2{#1}} %normally calls unnumberedsubseczzz
+\def\unnumberedsubseczzz #1{%
+\plainsubsecheading {#1}\gdef\thissection{#1}%
+\toks0 = {#1}%
+\edef\temp{\noexpand\writetocentry{\realbackslash unnumbsubsecentry%
+                                    {\the\toks0}}}%
+\temp
+\unnumbnoderef
+\nobreak
+}
+
+% Subsubsections.
+\outer\def\numberedsubsubsec{\parsearg\numberedsubsubsecyyy}
+\def\numberedsubsubsecyyy #1{\numhead3{#1}} % normally numberedsubsubseczzz
+\def\numberedsubsubseczzz #1{%
+\gdef\thissection{#1}\global\advance \subsubsecno by 1 %
+\subsubsecheading {#1}
+  {\the\chapno}{\the\secno}{\the\subsecno}{\the\subsubsecno}%
+\toks0 = {#1}%
+\edef\temp{\noexpand\writetocentry{\realbackslash subsubsecentry{\the\toks0}%
+  {\the\chapno}{\the\secno}{\the\subsecno}{\the\subsubsecno}}}%
+\temp
+\donoderef
+\nobreak
+}
+
+\outer\def\appendixsubsubsec{\parsearg\appendixsubsubsecyyy}
+\def\appendixsubsubsecyyy #1{\apphead3{#1}} % normally appendixsubsubseczzz
+\def\appendixsubsubseczzz #1{%
+\gdef\thissection{#1}\global\advance \subsubsecno by 1 %
+\subsubsecheading {#1}
+  {\appendixletter}{\the\secno}{\the\subsecno}{\the\subsubsecno}%
+\toks0 = {#1}%
+\edef\temp{\noexpand\writetocentry{\realbackslash subsubsecentry{\the\toks0}%
+  {\appendixletter}{\the\secno}{\the\subsecno}{\the\subsubsecno}}}%
+\temp
+\appendixnoderef
+\nobreak
+}
+
+\outer\def\unnumberedsubsubsec{\parsearg\unnumberedsubsubsecyyy}
+\def\unnumberedsubsubsecyyy #1{\unnmhead3{#1}} %normally unnumberedsubsubseczzz
+\def\unnumberedsubsubseczzz #1{%
+\plainsubsubsecheading {#1}\gdef\thissection{#1}%
+\toks0 = {#1}%
+\edef\temp{\noexpand\writetocentry{\realbackslash unnumbsubsubsecentry%
+                                    {\the\toks0}}}%
+\temp
+\unnumbnoderef
+\nobreak
+}
+
+% These are variants which are not "outer", so they can appear in @ifinfo.
+% Actually, they should now be obsolete; ordinary section commands should work.
+\def\infotop{\parsearg\unnumberedzzz}
+\def\infounnumbered{\parsearg\unnumberedzzz}
+\def\infounnumberedsec{\parsearg\unnumberedseczzz}
+\def\infounnumberedsubsec{\parsearg\unnumberedsubseczzz}
+\def\infounnumberedsubsubsec{\parsearg\unnumberedsubsubseczzz}
+
+\def\infoappendix{\parsearg\appendixzzz}
+\def\infoappendixsec{\parsearg\appendixseczzz}
+\def\infoappendixsubsec{\parsearg\appendixsubseczzz}
+\def\infoappendixsubsubsec{\parsearg\appendixsubsubseczzz}
+
+\def\infochapter{\parsearg\chapterzzz}
+\def\infosection{\parsearg\sectionzzz}
+\def\infosubsection{\parsearg\subsectionzzz}
+\def\infosubsubsection{\parsearg\subsubsectionzzz}
+
+% These macros control what the section commands do, according
+% to what kind of chapter we are in (ordinary, appendix, or unnumbered).
+% Define them by default for a numbered chapter.
+\global\let\section = \numberedsec
+\global\let\subsection = \numberedsubsec
+\global\let\subsubsection = \numberedsubsubsec
+
+% Define @majorheading, @heading and @subheading
+
+% NOTE on use of \vbox for chapter headings, section headings, and such:
+%       1) We use \vbox rather than the earlier \line to permit
+%          overlong headings to fold.
+%       2) \hyphenpenalty is set to 10000 because hyphenation in a
+%          heading is obnoxious; this forbids it.
+%       3) Likewise, headings look best if no \parindent is used, and
+%          if justification is not attempted.  Hence \raggedright.
+
+
+\def\majorheading{\parsearg\majorheadingzzz}
+\def\majorheadingzzz #1{%
+{\advance\chapheadingskip by 10pt \chapbreak }%
+{\chapfonts \vbox{\hyphenpenalty=10000\tolerance=5000
+                  \parindent=0pt\raggedright
+                  \rm #1\hfill}}\bigskip \par\penalty 200}
+
+\def\chapheading{\parsearg\chapheadingzzz}
+\def\chapheadingzzz #1{\chapbreak %
+{\chapfonts \vbox{\hyphenpenalty=10000\tolerance=5000
+                  \parindent=0pt\raggedright
+                  \rm #1\hfill}}\bigskip \par\penalty 200}
+
+% @heading, @subheading, @subsubheading.
+\def\heading{\parsearg\plainsecheading}
+\def\subheading{\parsearg\plainsubsecheading}
+\def\subsubheading{\parsearg\plainsubsubsecheading}
+
+% These macros generate a chapter, section, etc. heading only
+% (including whitespace, linebreaking, etc. around it),
+% given all the information in convenient, parsed form.
+
+%%% Args are the skip and penalty (usually negative)
+\def\dobreak#1#2{\par\ifdim\lastskip<#1\removelastskip\penalty#2\vskip#1\fi}
+
+\def\setchapterstyle #1 {\csname CHAPF#1\endcsname}
+
+%%% Define plain chapter starts, and page on/off switching for it
+% Parameter controlling skip before chapter headings (if needed)
+
+\newskip\chapheadingskip
+
+\def\chapbreak{\dobreak \chapheadingskip {-4000}}
+\def\chappager{\par\vfill\supereject}
+\def\chapoddpage{\chappager \ifodd\pageno \else \hbox to 0pt{} \chappager\fi}
+
+\def\setchapternewpage #1 {\csname CHAPPAG#1\endcsname}
+
+\def\CHAPPAGoff{%
+\global\let\contentsalignmacro = \chappager
+\global\let\pchapsepmacro=\chapbreak
+\global\let\pagealignmacro=\chappager}
+
+\def\CHAPPAGon{%
+\global\let\contentsalignmacro = \chappager
+\global\let\pchapsepmacro=\chappager
+\global\let\pagealignmacro=\chappager
+\global\def\HEADINGSon{\HEADINGSsingle}}
+
+\def\CHAPPAGodd{
+\global\let\contentsalignmacro = \chapoddpage
+\global\let\pchapsepmacro=\chapoddpage
+\global\let\pagealignmacro=\chapoddpage
+\global\def\HEADINGSon{\HEADINGSdouble}}
+
+\CHAPPAGon
+
+\def\CHAPFplain{
+\global\let\chapmacro=\chfplain
+\global\let\unnumbchapmacro=\unnchfplain
+\global\let\centerchapmacro=\centerchfplain}
+
+% Plain chapter opening.
+% #1 is the text, #2 the chapter number or empty if unnumbered.
+\def\chfplain#1#2{%
+  \pchapsepmacro
+  {%
+    \chapfonts \rm
+    \def\chapnum{#2}%
+    \setbox0 = \hbox{#2\ifx\chapnum\empty\else\enspace\fi}%
+    \vbox{\hyphenpenalty=10000 \tolerance=5000 \parindent=0pt \raggedright
+          \hangindent = \wd0 \centerparametersmaybe
+          \unhbox0 #1\par}%
+  }%
+  \nobreak\bigskip % no page break after a chapter title
+  \nobreak
+}
+
+% Plain opening for unnumbered.
+\def\unnchfplain#1{\chfplain{#1}{}}
+
+% @centerchap -- centered and unnumbered.
+\let\centerparametersmaybe = \relax
+\def\centerchfplain#1{{%
+  \def\centerparametersmaybe{%
+    \advance\rightskip by 3\rightskip
+    \leftskip = \rightskip
+    \parfillskip = 0pt
+  }%
+  \chfplain{#1}{}%
+}}
+
+\CHAPFplain % The default
+
+\def\unnchfopen #1{%
+\chapoddpage {\chapfonts \vbox{\hyphenpenalty=10000\tolerance=5000
+                       \parindent=0pt\raggedright
+                       \rm #1\hfill}}\bigskip \par\nobreak
+}
+
+\def\chfopen #1#2{\chapoddpage {\chapfonts
+\vbox to 3in{\vfil \hbox to\hsize{\hfil #2} \hbox to\hsize{\hfil #1} \vfil}}%
+\par\penalty 5000 %
+}
+
+\def\centerchfopen #1{%
+\chapoddpage {\chapfonts \vbox{\hyphenpenalty=10000\tolerance=5000
+                       \parindent=0pt
+                       \hfill {\rm #1}\hfill}}\bigskip \par\nobreak
+}
+
+\def\CHAPFopen{
+\global\let\chapmacro=\chfopen
+\global\let\unnumbchapmacro=\unnchfopen
+\global\let\centerchapmacro=\centerchfopen}
+
+
+% Section titles.
+\newskip\secheadingskip
+\def\secheadingbreak{\dobreak \secheadingskip {-1000}}
+\def\secheading#1#2#3{\sectionheading{sec}{#2.#3}{#1}}
+\def\plainsecheading#1{\sectionheading{sec}{}{#1}}
+
+% Subsection titles.
+\newskip \subsecheadingskip
+\def\subsecheadingbreak{\dobreak \subsecheadingskip {-500}}
+\def\subsecheading#1#2#3#4{\sectionheading{subsec}{#2.#3.#4}{#1}}
+\def\plainsubsecheading#1{\sectionheading{subsec}{}{#1}}
+
+% Subsubsection titles.
+\let\subsubsecheadingskip = \subsecheadingskip
+\let\subsubsecheadingbreak = \subsecheadingbreak
+\def\subsubsecheading#1#2#3#4#5{\sectionheading{subsubsec}{#2.#3.#4.#5}{#1}}
+\def\plainsubsubsecheading#1{\sectionheading{subsubsec}{}{#1}}
+
+
+% Print any size section title.
+%
+% #1 is the section type (sec/subsec/subsubsec), #2 is the section
+% number (maybe empty), #3 the text.
+\def\sectionheading#1#2#3{%
+  {%
+    \expandafter\advance\csname #1headingskip\endcsname by \parskip
+    \csname #1headingbreak\endcsname
+  }%
+  {%
+    % Switch to the right set of fonts.
+    \csname #1fonts\endcsname \rm
+    %
+    % Only insert the separating space if we have a section number.
+    \def\secnum{#2}%
+    \setbox0 = \hbox{#2\ifx\secnum\empty\else\enspace\fi}%
+    %
+    \vbox{\hyphenpenalty=10000 \tolerance=5000 \parindent=0pt \raggedright
+          \hangindent = \wd0 % zero if no section number
+          \unhbox0 #3}%
+  }%
+  \ifdim\parskip<10pt \nobreak\kern10pt\nobreak\kern-\parskip\fi \nobreak
+}
+
+
+\message{toc,}
+% Table of contents.
+\newwrite\tocfile
+
+% Write an entry to the toc file, opening it if necessary.
+% Called from @chapter, etc.  We supply {\folio} at the end of the
+% argument, which will end up as the last argument to the \...entry macro.
+%
+% We open the .toc file here instead of at @setfilename or any other
+% given time so that @contents can be put in the document anywhere.
+%
+\newif\iftocfileopened
+\def\writetocentry#1{%
+  \iftocfileopened\else
+    \immediate\openout\tocfile = \jobname.toc
+    \global\tocfileopenedtrue
+  \fi
+  \iflinks \write\tocfile{#1{\folio}}\fi
+}
+
+\newskip\contentsrightmargin \contentsrightmargin=1in
+\newcount\savepageno
+\newcount\lastnegativepageno \lastnegativepageno = -1
+
+% Finish up the main text and prepare to read what we've written
+% to \tocfile.
+%
+\def\startcontents#1{%
+   % If @setchapternewpage on, and @headings double, the contents should
+   % start on an odd page, unlike chapters.  Thus, we maintain
+   % \contentsalignmacro in parallel with \pagealignmacro.
+   % From: Torbjorn Granlund <tege@matematik.su.se>
+   \contentsalignmacro
+   \immediate\closeout\tocfile
+   %
+   % Don't need to put `Contents' or `Short Contents' in the headline.
+   % It is abundantly clear what they are.
+   \unnumbchapmacro{#1}\def\thischapter{}%
+   \savepageno = \pageno
+   \begingroup                  % Set up to handle contents files properly.
+      \catcode`\\=0  \catcode`\{=1  \catcode`\}=2  \catcode`\@=11
+      % We can't do this, because then an actual ^ in a section
+      % title fails, e.g., @chapter ^ -- exponentiation.  --karl, 9jul97.
+      %\catcode`\^=7 % to see ^^e4 as \"a etc. juha@piuha.ydi.vtt.fi
+      \raggedbottom             % Worry more about breakpoints than the bottom.
+      \advance\hsize by -\contentsrightmargin % Don't use the full line length.
+      %
+      % Roman numerals for page numbers.
+      \ifnum \pageno>0 \pageno = \lastnegativepageno \fi
+}
+
+
+% Normal (long) toc.
+\def\contents{%
+   \startcontents{\putwordTOC}%
+     \openin 1 \jobname.toc
+     \ifeof 1 \else
+       \closein 1
+       \input \jobname.toc
+     \fi
+     \vfill \eject
+     \contentsalignmacro % in case @setchapternewpage odd is in effect
+     \pdfmakeoutlines
+   \endgroup
+   \lastnegativepageno = \pageno
+   \pageno = \savepageno
+}
+
+% And just the chapters.
+\def\summarycontents{%
+   \startcontents{\putwordShortTOC}%
+      %
+      \let\chapentry = \shortchapentry
+      \let\unnumbchapentry = \shortunnumberedentry
+      % We want a true roman here for the page numbers.
+      \secfonts
+      \let\rm=\shortcontrm \let\bf=\shortcontbf \let\sl=\shortcontsl
+      \rm
+      \hyphenpenalty = 10000
+      \advance\baselineskip by 1pt % Open it up a little.
+      \def\secentry ##1##2##3##4{}
+      \def\unnumbsecentry ##1##2{}
+      \def\subsecentry ##1##2##3##4##5{}
+      \def\unnumbsubsecentry ##1##2{}
+      \def\subsubsecentry ##1##2##3##4##5##6{}
+      \def\unnumbsubsubsecentry ##1##2{}
+      \openin 1 \jobname.toc
+      \ifeof 1 \else
+        \closein 1
+        \input \jobname.toc
+      \fi
+     \vfill \eject
+     \contentsalignmacro % in case @setchapternewpage odd is in effect
+   \endgroup
+   \lastnegativepageno = \pageno
+   \pageno = \savepageno
+}
+\let\shortcontents = \summarycontents
+
+\ifpdf
+  \pdfcatalog{/PageMode /UseOutlines}%
+\fi
+
+% These macros generate individual entries in the table of contents.
+% The first argument is the chapter or section name.
+% The last argument is the page number.
+% The arguments in between are the chapter number, section number, ...
+
+% Chapter-level things, for both the long and short contents.
+\def\chapentry#1#2#3{\dochapentry{#2\labelspace#1}{#3}}
+
+% See comments in \dochapentry re vbox and related settings
+\def\shortchapentry#1#2#3{%
+  \tocentry{\shortchaplabel{#2}\labelspace #1}{\doshortpageno\bgroup#3\egroup}%
+}
+
+% Typeset the label for a chapter or appendix for the short contents.
+% The arg is, e.g. `Appendix A' for an appendix, or `3' for a chapter.
+% We could simplify the code here by writing out an \appendixentry
+% command in the toc file for appendices, instead of using \chapentry
+% for both, but it doesn't seem worth it.
+%
+\newdimen\shortappendixwidth
+%
+\def\shortchaplabel#1{%
+  % Compute width of word "Appendix", may change with language.
+  \setbox0 = \hbox{\shortcontrm \putwordAppendix}%
+  \shortappendixwidth = \wd0
+  %
+  % We typeset #1 in a box of constant width, regardless of the text of
+  % #1, so the chapter titles will come out aligned.
+  \setbox0 = \hbox{#1}%
+  \dimen0 = \ifdim\wd0 > \shortappendixwidth \shortappendixwidth \else 0pt \fi
+  %
+  % This space should be plenty, since a single number is .5em, and the
+  % widest letter (M) is 1em, at least in the Computer Modern fonts.
+  % (This space doesn't include the extra space that gets added after
+  % the label; that gets put in by \shortchapentry above.)
+  \advance\dimen0 by 1.1em
+  \hbox to \dimen0{#1\hfil}%
+}
+
+\def\unnumbchapentry#1#2{\dochapentry{#1}{#2}}
+\def\shortunnumberedentry#1#2{\tocentry{#1}{\doshortpageno\bgroup#2\egroup}}
+
+% Sections.
+\def\secentry#1#2#3#4{\dosecentry{#2.#3\labelspace#1}{#4}}
+\def\unnumbsecentry#1#2{\dosecentry{#1}{#2}}
+
+% Subsections.
+\def\subsecentry#1#2#3#4#5{\dosubsecentry{#2.#3.#4\labelspace#1}{#5}}
+\def\unnumbsubsecentry#1#2{\dosubsecentry{#1}{#2}}
+
+% And subsubsections.
+\def\subsubsecentry#1#2#3#4#5#6{%
+  \dosubsubsecentry{#2.#3.#4.#5\labelspace#1}{#6}}
+\def\unnumbsubsubsecentry#1#2{\dosubsubsecentry{#1}{#2}}
+
+% This parameter controls the indentation of the various levels.
+\newdimen\tocindent \tocindent = 3pc
+
+% Now for the actual typesetting. In all these, #1 is the text and #2 is the
+% page number.
+%
+% If the toc has to be broken over pages, we want it to be at chapters
+% if at all possible; hence the \penalty.
+\def\dochapentry#1#2{%
+   \penalty-300 \vskip1\baselineskip plus.33\baselineskip minus.25\baselineskip
+   \begingroup
+     \chapentryfonts
+     \tocentry{#1}{\dopageno\bgroup#2\egroup}%
+   \endgroup
+   \nobreak\vskip .25\baselineskip plus.1\baselineskip
+}
+
+\def\dosecentry#1#2{\begingroup
+  \secentryfonts \leftskip=\tocindent
+  \tocentry{#1}{\dopageno\bgroup#2\egroup}%
+\endgroup}
+
+\def\dosubsecentry#1#2{\begingroup
+  \subsecentryfonts \leftskip=2\tocindent
+  \tocentry{#1}{\dopageno\bgroup#2\egroup}%
+\endgroup}
+
+\def\dosubsubsecentry#1#2{\begingroup
+  \subsubsecentryfonts \leftskip=3\tocindent
+  \tocentry{#1}{\dopageno\bgroup#2\egroup}%
+\endgroup}
+
+% Final typesetting of a toc entry; we use the same \entry macro as for
+% the index entries, but we want to suppress hyphenation here.  (We
+% can't do that in the \entry macro, since index entries might consist
+% of hyphenated-identifiers-that-do-not-fit-on-a-line-and-nothing-else.)
+\def\tocentry#1#2{\begingroup
+  \vskip 0pt plus1pt % allow a little stretch for the sake of nice page breaks
+  % Do not use \turnoffactive in these arguments.  Since the toc is
+  % typeset in cmr, so characters such as _ would come out wrong; we
+  % have to do the usual translation tricks.
+  \entry{#1}{#2}%
+\endgroup}
+
+% Space between chapter (or whatever) number and the title.
+\def\labelspace{\hskip1em \relax}
+
+\def\dopageno#1{{\rm #1}}
+\def\doshortpageno#1{{\rm #1}}
+
+\def\chapentryfonts{\secfonts \rm}
+\def\secentryfonts{\textfonts}
+\let\subsecentryfonts = \textfonts
+\let\subsubsecentryfonts = \textfonts
+
+
+\message{environments,}
+% @foo ... @end foo.
+
+% Since these characters are used in examples, it should be an even number of
+% \tt widths. Each \tt character is 1en, so two makes it 1em.
+% Furthermore, these definitions must come after we define our fonts.
+\newbox\dblarrowbox    \newbox\longdblarrowbox
+\newbox\pushcharbox    \newbox\bullbox
+\newbox\equivbox       \newbox\errorbox
+
+%{\tentt
+%\global\setbox\dblarrowbox = \hbox to 1em{\hfil$\Rightarrow$\hfil}
+%\global\setbox\longdblarrowbox = \hbox to 1em{\hfil$\mapsto$\hfil}
+%\global\setbox\pushcharbox = \hbox to 1em{\hfil$\dashv$\hfil}
+%\global\setbox\equivbox = \hbox to 1em{\hfil$\ptexequiv$\hfil}
+% Adapted from the manmac format (p.420 of TeXbook)
+%\global\setbox\bullbox = \hbox to 1em{\kern.15em\vrule height .75ex width .85ex
+%                                      depth .1ex\hfil}
+%}
+
+% @point{}, @result{}, @expansion{}, @print{}, @equiv{}.
+\def\point{$\star$}
+\def\result{\leavevmode\raise.15ex\hbox to 1em{\hfil$\Rightarrow$\hfil}}
+\def\expansion{\leavevmode\raise.1ex\hbox to 1em{\hfil$\mapsto$\hfil}}
+\def\print{\leavevmode\lower.1ex\hbox to 1em{\hfil$\dashv$\hfil}}
+\def\equiv{\leavevmode\lower.1ex\hbox to 1em{\hfil$\ptexequiv$\hfil}}
+
+% Adapted from the TeXbook's \boxit.
+{\tentt \global\dimen0 = 3em}% Width of the box.
+\dimen2 = .55pt % Thickness of rules
+% The text. (`r' is open on the right, `e' somewhat less so on the left.)
+\setbox0 = \hbox{\kern-.75pt \tensf error\kern-1.5pt}
+
+\global\setbox\errorbox=\hbox to \dimen0{\hfil
+   \hsize = \dimen0 \advance\hsize by -5.8pt % Space to left+right.
+   \advance\hsize by -2\dimen2 % Rules.
+   \vbox{
+      \hrule height\dimen2
+      \hbox{\vrule width\dimen2 \kern3pt          % Space to left of text.
+         \vtop{\kern2.4pt \box0 \kern2.4pt}% Space above/below.
+         \kern3pt\vrule width\dimen2}% Space to right.
+      \hrule height\dimen2}
+    \hfil}
+
+% The @error{} command.
+\def\error{\leavevmode\lower.7ex\copy\errorbox}
+
+% @tex ... @end tex    escapes into raw Tex temporarily.
+% One exception: @ is still an escape character, so that @end tex works.
+% But \@ or @@ will get a plain tex @ character.
+
+\def\tex{\begingroup
+  \catcode `\\=0 \catcode `\{=1 \catcode `\}=2
+  \catcode `\$=3 \catcode `\&=4 \catcode `\#=6
+  \catcode `\^=7 \catcode `\_=8 \catcode `\~=13 \let~=\tie
+  \catcode `\%=14
+  \catcode 43=12 % plus
+  \catcode`\"=12
+  \catcode`\==12
+  \catcode`\|=12
+  \catcode`\<=12
+  \catcode`\>=12
+  \escapechar=`\\
+  %
+  \let\b=\ptexb
+  \let\bullet=\ptexbullet
+  \let\c=\ptexc
+  \let\,=\ptexcomma
+  \let\.=\ptexdot
+  \let\dots=\ptexdots
+  \let\equiv=\ptexequiv
+  \let\!=\ptexexclam
+  \let\i=\ptexi
+  \let\{=\ptexlbrace
+  \let\+=\tabalign
+  \let\}=\ptexrbrace
+  \let\*=\ptexstar
+  \let\t=\ptext
+  %
+  \def\endldots{\mathinner{\ldots\ldots\ldots\ldots}}%
+  \def\enddots{\relax\ifmmode\endldots\else$\mathsurround=0pt \endldots\,$\fi}%
+  \def\@{@}%
+\let\Etex=\endgroup}
+
+% Define @lisp ... @endlisp.
+% @lisp does a \begingroup so it can rebind things,
+% including the definition of @endlisp (which normally is erroneous).
+
+% Amount to narrow the margins by for @lisp.
+\newskip\lispnarrowing \lispnarrowing=0.4in
+
+% This is the definition that ^^M gets inside @lisp, @example, and other
+% such environments.  \null is better than a space, since it doesn't
+% have any width.
+\def\lisppar{\null\endgraf}
+
+% Make each space character in the input produce a normal interword
+% space in the output.  Don't allow a line break at this space, as this
+% is used only in environments like @example, where each line of input
+% should produce a line of output anyway.
+%
+{\obeyspaces %
+\gdef\sepspaces{\obeyspaces\let =\tie}}
+
+% Define \obeyedspace to be our active space, whatever it is.  This is
+% for use in \parsearg.
+{\sepspaces%
+\global\let\obeyedspace= }
+
+% This space is always present above and below environments.
+\newskip\envskipamount \envskipamount = 0pt
+
+% Make spacing and below environment symmetrical.  We use \parskip here
+% to help in doing that, since in @example-like environments \parskip
+% is reset to zero; thus the \afterenvbreak inserts no space -- but the
+% start of the next paragraph will insert \parskip
+%
+\def\aboveenvbreak{{\advance\envskipamount by \parskip
+\endgraf \ifdim\lastskip<\envskipamount
+\removelastskip \penalty-50 \vskip\envskipamount \fi}}
+
+\let\afterenvbreak = \aboveenvbreak
+
+% \nonarrowing is a flag.  If "set", @lisp etc don't narrow margins.
+\let\nonarrowing=\relax
+
+% @cartouche ... @end cartouche: draw rectangle w/rounded corners around
+% environment contents.
+\font\circle=lcircle10
+\newdimen\circthick
+\newdimen\cartouter\newdimen\cartinner
+\newskip\normbskip\newskip\normpskip\newskip\normlskip
+\circthick=\fontdimen8\circle
+%
+\def\ctl{{\circle\char'013\hskip -6pt}}% 6pt from pl file: 1/2charwidth
+\def\ctr{{\hskip 6pt\circle\char'010}}
+\def\cbl{{\circle\char'012\hskip -6pt}}
+\def\cbr{{\hskip 6pt\circle\char'011}}
+\def\carttop{\hbox to \cartouter{\hskip\lskip
+        \ctl\leaders\hrule height\circthick\hfil\ctr
+        \hskip\rskip}}
+\def\cartbot{\hbox to \cartouter{\hskip\lskip
+        \cbl\leaders\hrule height\circthick\hfil\cbr
+        \hskip\rskip}}
+%
+\newskip\lskip\newskip\rskip
+
+\long\def\cartouche{%
+\begingroup
+        \lskip=\leftskip \rskip=\rightskip
+        \leftskip=0pt\rightskip=0pt %we want these *outside*.
+        \cartinner=\hsize \advance\cartinner by-\lskip
+                          \advance\cartinner by-\rskip
+        \cartouter=\hsize
+        \advance\cartouter by 18.4pt % allow for 3pt kerns on either
+%                                    side, and for 6pt waste from
+%                                    each corner char, and rule thickness
+        \normbskip=\baselineskip \normpskip=\parskip \normlskip=\lineskip
+        % Flag to tell @lisp, etc., not to narrow margin.
+        \let\nonarrowing=\comment
+        \vbox\bgroup
+                \baselineskip=0pt\parskip=0pt\lineskip=0pt
+                \carttop
+                \hbox\bgroup
+                        \hskip\lskip
+                        \vrule\kern3pt
+                        \vbox\bgroup
+                                \hsize=\cartinner
+                                \kern3pt
+                                \begingroup
+                                        \baselineskip=\normbskip
+                                        \lineskip=\normlskip
+                                        \parskip=\normpskip
+                                        \vskip -\parskip
+\def\Ecartouche{%
+                                \endgroup
+                                \kern3pt
+                        \egroup
+                        \kern3pt\vrule
+                        \hskip\rskip
+                \egroup
+                \cartbot
+        \egroup
+\endgroup
+}}
+
+
+% This macro is called at the beginning of all the @example variants,
+% inside a group.
+\def\nonfillstart{%
+  \aboveenvbreak
+  \inENV % This group ends at the end of the body
+  \hfuzz = 12pt % Don't be fussy
+  \sepspaces % Make spaces be word-separators rather than space tokens.
+  \singlespace
+  \let\par = \lisppar % don't ignore blank lines
+  \obeylines % each line of input is a line of output
+  \parskip = 0pt
+  \parindent = 0pt
+  \emergencystretch = 0pt % don't try to avoid overfull boxes
+  % @cartouche defines \nonarrowing to inhibit narrowing
+  % at next level down.
+  \ifx\nonarrowing\relax
+    \advance \leftskip by \lispnarrowing
+    \exdentamount=\lispnarrowing
+    \let\exdent=\nofillexdent
+    \let\nonarrowing=\relax
+  \fi
+}
+
+% Define the \E... control sequence only if we are inside the particular
+% environment, so the error checking in \end will work.
+%
+% To end an @example-like environment, we first end the paragraph (via
+% \afterenvbreak's vertical glue), and then the group.  That way we keep
+% the zero \parskip that the environments set -- \parskip glue will be
+% inserted at the beginning of the next paragraph in the document, after
+% the environment.
+%
+\def\nonfillfinish{\afterenvbreak\endgroup}
+
+% @lisp: indented, narrowed, typewriter font.
+\def\lisp{\begingroup
+  \nonfillstart
+  \let\Elisp = \nonfillfinish
+  \tt
+  \let\kbdfont = \kbdexamplefont % Allow @kbd to do something special.
+  \gobble       % eat return
+}
+
+% @example: Same as @lisp.
+\def\example{\begingroup \def\Eexample{\nonfillfinish\endgroup}\lisp}
+
+% @small... is usually equivalent to the non-small (@smallbook
+% redefines).  We must call \example (or whatever) last in the
+% definition, since it reads the return following the @example (or
+% whatever) command.
+%
+% This actually allows (for example) @end display inside an
+% @smalldisplay.  Too bad, but makeinfo will catch the error anyway.
+%
+\def\smalldisplay{\begingroup\def\Esmalldisplay{\nonfillfinish\endgroup}\display}
+\def\smallexample{\begingroup\def\Esmallexample{\nonfillfinish\endgroup}\lisp}
+\def\smallformat{\begingroup\def\Esmallformat{\nonfillfinish\endgroup}\format}
+\def\smalllisp{\begingroup\def\Esmalllisp{\nonfillfinish\endgroup}\lisp}
+
+% Real @smallexample and @smalllisp (when @smallbook): use smaller fonts.
+% Originally contributed by Pavel@xerox.
+\def\smalllispx{\begingroup
+  \def\Esmalllisp{\nonfillfinish\endgroup}%
+  \def\Esmallexample{\nonfillfinish\endgroup}%
+  \smallfonts
+  \lisp
+}
+
+% @display: same as @lisp except keep current font.
+%
+\def\display{\begingroup
+  \nonfillstart
+  \let\Edisplay = \nonfillfinish
+  \gobble
+}
+
+% @smalldisplay (when @smallbook): @display plus smaller fonts.
+%
+\def\smalldisplayx{\begingroup
+  \def\Esmalldisplay{\nonfillfinish\endgroup}%
+  \smallfonts \rm
+  \display
+}
+
+% @format: same as @display except don't narrow margins.
+%
+\def\format{\begingroup
+  \let\nonarrowing = t
+  \nonfillstart
+  \let\Eformat = \nonfillfinish
+  \gobble
+}
+
+% @smallformat (when @smallbook): @format plus smaller fonts.
+%
+\def\smallformatx{\begingroup
+  \def\Esmallformat{\nonfillfinish\endgroup}%
+  \smallfonts \rm
+  \format
+}
+
+% @flushleft (same as @format).
+%
+\def\flushleft{\begingroup \def\Eflushleft{\nonfillfinish\endgroup}\format}
+
+% @flushright.
+%
+\def\flushright{\begingroup
+  \let\nonarrowing = t
+  \nonfillstart
+  \let\Eflushright = \nonfillfinish
+  \advance\leftskip by 0pt plus 1fill
+  \gobble
+}
+
+% @quotation does normal linebreaking (hence we can't use \nonfillstart)
+% and narrows the margins.
+%
+\def\quotation{%
+  \begingroup\inENV %This group ends at the end of the @quotation body
+  {\parskip=0pt \aboveenvbreak}% because \aboveenvbreak inserts \parskip
+  \singlespace
+  \parindent=0pt
+  % We have retained a nonzero parskip for the environment, since we're
+  % doing normal filling. So to avoid extra space below the environment...
+  \def\Equotation{\parskip = 0pt \nonfillfinish}%
+  %
+  % @cartouche defines \nonarrowing to inhibit narrowing at next level down.
+  \ifx\nonarrowing\relax
+    \advance\leftskip by \lispnarrowing
+    \advance\rightskip by \lispnarrowing
+    \exdentamount = \lispnarrowing
+    \let\nonarrowing = \relax
+  \fi
+}
+
+
+\message{defuns,}
+% @defun etc.
+
+% Allow user to change definition object font (\df) internally
+\def\setdeffont #1 {\csname DEF#1\endcsname}
+
+\newskip\defbodyindent \defbodyindent=.4in
+\newskip\defargsindent \defargsindent=50pt
+\newskip\deftypemargin \deftypemargin=12pt
+\newskip\deflastargmargin \deflastargmargin=18pt
+
+\newcount\parencount
+% define \functionparens, which makes ( and ) and & do special things.
+% \functionparens affects the group it is contained in.
+\def\activeparens{%
+\catcode`\(=\active \catcode`\)=\active \catcode`\&=\active
+\catcode`\[=\active \catcode`\]=\active}
+
+% Make control sequences which act like normal parenthesis chars.
+\let\lparen = ( \let\rparen = )
+
+{\activeparens % Now, smart parens don't turn on until &foo (see \amprm)
+
+% Be sure that we always have a definition for `(', etc.  For example,
+% if the fn name has parens in it, \boldbrax will not be in effect yet,
+% so TeX would otherwise complain about undefined control sequence.
+\global\let(=\lparen \global\let)=\rparen
+\global\let[=\lbrack \global\let]=\rbrack
+
+\gdef\functionparens{\boldbrax\let&=\amprm\parencount=0 }
+\gdef\boldbrax{\let(=\opnr\let)=\clnr\let[=\lbrb\let]=\rbrb}
+% This is used to turn on special parens
+% but make & act ordinary (given that it's active).
+\gdef\boldbraxnoamp{\let(=\opnr\let)=\clnr\let[=\lbrb\let]=\rbrb\let&=\ampnr}
+
+% Definitions of (, ) and & used in args for functions.
+% This is the definition of ( outside of all parentheses.
+\gdef\oprm#1 {{\rm\char`\(}#1 \bf \let(=\opnested
+  \global\advance\parencount by 1
+}
+%
+% This is the definition of ( when already inside a level of parens.
+\gdef\opnested{\char`\(\global\advance\parencount by 1 }
+%
+\gdef\clrm{% Print a paren in roman if it is taking us back to depth of 0.
+  % also in that case restore the outer-level definition of (.
+  \ifnum \parencount=1 {\rm \char `\)}\sl \let(=\oprm \else \char `\) \fi
+  \global\advance \parencount by -1 }
+% If we encounter &foo, then turn on ()-hacking afterwards
+\gdef\amprm#1 {{\rm\&#1}\let(=\oprm \let)=\clrm\ }
+%
+\gdef\normalparens{\boldbrax\let&=\ampnr}
+} % End of definition inside \activeparens
+%% These parens (in \boldbrax) actually are a little bolder than the
+%% contained text.  This is especially needed for [ and ]
+\def\opnr{{\sf\char`\(}\global\advance\parencount by 1 }
+\def\clnr{{\sf\char`\)}\global\advance\parencount by -1 }
+\let\ampnr = \&
+\def\lbrb{{\bf\char`\[}}
+\def\rbrb{{\bf\char`\]}}
+
+% Active &'s sneak into the index arguments, so make sure it's defined.
+{
+  \catcode`& = 13
+  \global\let& = \ampnr
+}
+
+% First, defname, which formats the header line itself.
+% #1 should be the function name.
+% #2 should be the type of definition, such as "Function".
+
+\def\defname #1#2{%
+% Get the values of \leftskip and \rightskip as they were
+% outside the @def...
+\dimen2=\leftskip
+\advance\dimen2 by -\defbodyindent
+\noindent
+\setbox0=\hbox{\hskip \deflastargmargin{\rm #2}\hskip \deftypemargin}%
+\dimen0=\hsize \advance \dimen0 by -\wd0 % compute size for first line
+\dimen1=\hsize \advance \dimen1 by -\defargsindent %size for continuations
+\parshape 2 0in \dimen0 \defargsindent \dimen1
+% Now output arg 2 ("Function" or some such)
+% ending at \deftypemargin from the right margin,
+% but stuck inside a box of width 0 so it does not interfere with linebreaking
+{% Adjust \hsize to exclude the ambient margins,
+% so that \rightline will obey them.
+\advance \hsize by -\dimen2
+\rlap{\rightline{{\rm #2}\hskip -1.25pc }}}%
+% Make all lines underfull and no complaints:
+\tolerance=10000 \hbadness=10000
+\advance\leftskip by -\defbodyindent
+\exdentamount=\defbodyindent
+{\df #1}\enskip        % Generate function name
+}
+
+% Actually process the body of a definition
+% #1 should be the terminating control sequence, such as \Edefun.
+% #2 should be the "another name" control sequence, such as \defunx.
+% #3 should be the control sequence that actually processes the header,
+%    such as \defunheader.
+
+\def\defparsebody #1#2#3{\begingroup\inENV% Environment for definitionbody
+\medbreak %
+% Define the end token that this defining construct specifies
+% so that it will exit this group.
+\def#1{\endgraf\endgroup\medbreak}%
+\def#2{\begingroup\obeylines\activeparens\spacesplit#3}%
+\parindent=0in
+\advance\leftskip by \defbodyindent
+\exdentamount=\defbodyindent
+\begingroup %
+\catcode 61=\active % 61 is `='
+\obeylines\activeparens\spacesplit#3}
+
+% #1 is the \E... control sequence to end the definition (which we define).
+% #2 is the \...x control sequence for consecutive fns (which we define).
+% #3 is the control sequence to call to resume processing.
+% #4, delimited by the space, is the class name.
+%
+\def\defmethparsebody#1#2#3#4 {\begingroup\inENV %
+\medbreak %
+% Define the end token that this defining construct specifies
+% so that it will exit this group.
+\def#1{\endgraf\endgroup\medbreak}%
+\def#2##1 {\begingroup\obeylines\activeparens\spacesplit{#3{##1}}}%
+\parindent=0in
+\advance\leftskip by \defbodyindent
+\exdentamount=\defbodyindent
+\begingroup\obeylines\activeparens\spacesplit{#3{#4}}}
+
+% Used for @deftypemethod and @deftypeivar.
+% #1 is the \E... control sequence to end the definition (which we define).
+% #2 is the \...x control sequence for consecutive fns (which we define).
+% #3 is the control sequence to call to resume processing.
+% #4, delimited by a space, is the class name.
+% #5 is the method's return type.
+%
+\def\deftypemethparsebody#1#2#3#4 #5 {\begingroup\inENV
+  \medbreak
+  \def#1{\endgraf\endgroup\medbreak}%
+  \def#2##1 ##2 {\begingroup\obeylines\activeparens\spacesplit{#3{##1}{##2}}}%
+  \parindent=0in
+  \advance\leftskip by \defbodyindent
+  \exdentamount=\defbodyindent
+  \begingroup\obeylines\activeparens\spacesplit{#3{#4}{#5}}}
+
+% Used for @deftypeop.  The change from \deftypemethparsebody is an
+% extra argument at the beginning which is the `category', instead of it
+% being the hardwired string `Method' or `Instance Variable'.  We have
+% to account for this both in the \...x definition and in parsing the
+% input at hand.  Thus also need a control sequence (passed as #5) for
+% the \E... definition to assign the category name to.
+% 
+\def\deftypeopparsebody#1#2#3#4#5 #6 {\begingroup\inENV
+  \medbreak
+  \def#1{\endgraf\endgroup\medbreak}%
+  \def#2##1 ##2 ##3 {%
+    \def#4{##1}%
+    \begingroup\obeylines\activeparens\spacesplit{#3{##2}{##3}}}%
+  \parindent=0in
+  \advance\leftskip by \defbodyindent
+  \exdentamount=\defbodyindent
+  \begingroup\obeylines\activeparens\spacesplit{#3{#5}{#6}}}
+
+\def\defopparsebody #1#2#3#4#5 {\begingroup\inENV %
+\medbreak %
+% Define the end token that this defining construct specifies
+% so that it will exit this group.
+\def#1{\endgraf\endgroup\medbreak}%
+\def#2##1 ##2 {\def#4{##1}%
+\begingroup\obeylines\activeparens\spacesplit{#3{##2}}}%
+\parindent=0in
+\advance\leftskip by \defbodyindent
+\exdentamount=\defbodyindent
+\begingroup\obeylines\activeparens\spacesplit{#3{#5}}}
+
+% These parsing functions are similar to the preceding ones
+% except that they do not make parens into active characters.
+% These are used for "variables" since they have no arguments.
+
+\def\defvarparsebody #1#2#3{\begingroup\inENV% Environment for definitionbody
+\medbreak %
+% Define the end token that this defining construct specifies
+% so that it will exit this group.
+\def#1{\endgraf\endgroup\medbreak}%
+\def#2{\begingroup\obeylines\spacesplit#3}%
+\parindent=0in
+\advance\leftskip by \defbodyindent
+\exdentamount=\defbodyindent
+\begingroup %
+\catcode 61=\active %
+\obeylines\spacesplit#3}
+
+% This is used for \def{tp,vr}parsebody.  It could probably be used for
+% some of the others, too, with some judicious conditionals.
+%
+\def\parsebodycommon#1#2#3{%
+  \begingroup\inENV %
+  \medbreak %
+  % Define the end token that this defining construct specifies
+  % so that it will exit this group.
+  \def#1{\endgraf\endgroup\medbreak}%
+  \def#2##1 {\begingroup\obeylines\spacesplit{#3{##1}}}%
+  \parindent=0in
+  \advance\leftskip by \defbodyindent
+  \exdentamount=\defbodyindent
+  \begingroup\obeylines
+}
+
+\def\defvrparsebody#1#2#3#4 {%
+  \parsebodycommon{#1}{#2}{#3}%
+  \spacesplit{#3{#4}}%
+}
+
+% This loses on `@deftp {Data Type} {struct termios}' -- it thinks the
+% type is just `struct', because we lose the braces in `{struct
+% termios}' when \spacesplit reads its undelimited argument.  Sigh.
+% \let\deftpparsebody=\defvrparsebody
+%
+% So, to get around this, we put \empty in with the type name.  That
+% way, TeX won't find exactly `{...}' as an undelimited argument, and
+% won't strip off the braces.
+%
+\def\deftpparsebody #1#2#3#4 {%
+  \parsebodycommon{#1}{#2}{#3}%
+  \spacesplit{\parsetpheaderline{#3{#4}}}\empty
+}
+
+% Fine, but then we have to eventually remove the \empty *and* the
+% braces (if any).  That's what this does.
+%
+\def\removeemptybraces\empty#1\relax{#1}
+
+% After \spacesplit has done its work, this is called -- #1 is the final
+% thing to call, #2 the type name (which starts with \empty), and #3
+% (which might be empty) the arguments.
+%
+\def\parsetpheaderline#1#2#3{%
+  #1{\removeemptybraces#2\relax}{#3}%
+}%
+
+\def\defopvarparsebody #1#2#3#4#5 {\begingroup\inENV %
+\medbreak %
+% Define the end token that this defining construct specifies
+% so that it will exit this group.
+\def#1{\endgraf\endgroup\medbreak}%
+\def#2##1 ##2 {\def#4{##1}%
+\begingroup\obeylines\spacesplit{#3{##2}}}%
+\parindent=0in
+\advance\leftskip by \defbodyindent
+\exdentamount=\defbodyindent
+\begingroup\obeylines\spacesplit{#3{#5}}}
+
+% Split up #2 at the first space token.
+% call #1 with two arguments:
+%  the first is all of #2 before the space token,
+%  the second is all of #2 after that space token.
+% If #2 contains no space token, all of it is passed as the first arg
+% and the second is passed as empty.
+
+{\obeylines
+\gdef\spacesplit#1#2^^M{\endgroup\spacesplitfoo{#1}#2 \relax\spacesplitfoo}%
+\long\gdef\spacesplitfoo#1#2 #3#4\spacesplitfoo{%
+\ifx\relax #3%
+#1{#2}{}\else #1{#2}{#3#4}\fi}}
+
+% So much for the things common to all kinds of definitions.
+
+% Define @defun.
+
+% First, define the processing that is wanted for arguments of \defun
+% Use this to expand the args and terminate the paragraph they make up
+
+\def\defunargs#1{\functionparens \sl
+% Expand, preventing hyphenation at `-' chars.
+% Note that groups don't affect changes in \hyphenchar.
+% Set the font temporarily and use \font in case \setfont made \tensl a macro.
+{\tensl\hyphenchar\font=0}%
+#1%
+{\tensl\hyphenchar\font=45}%
+\ifnum\parencount=0 \else \errmessage{Unbalanced parentheses in @def}\fi%
+\interlinepenalty=10000
+\advance\rightskip by 0pt plus 1fil
+\endgraf\nobreak\vskip -\parskip\nobreak
+}
+
+\def\deftypefunargs #1{%
+% Expand, preventing hyphenation at `-' chars.
+% Note that groups don't affect changes in \hyphenchar.
+% Use \boldbraxnoamp, not \functionparens, so that & is not special.
+\boldbraxnoamp
+\tclose{#1}% avoid \code because of side effects on active chars
+\interlinepenalty=10000
+\advance\rightskip by 0pt plus 1fil
+\endgraf\nobreak\vskip -\parskip\nobreak
+}
+
+% Do complete processing of one @defun or @defunx line already parsed.
+
+% @deffn Command forward-char nchars
+
+\def\deffn{\defmethparsebody\Edeffn\deffnx\deffnheader}
+
+\def\deffnheader #1#2#3{\doind {fn}{\code{#2}}%
+\begingroup\defname {#2}{#1}\defunargs{#3}\endgroup %
+\catcode 61=\other % Turn off change made in \defparsebody
+}
+
+% @defun == @deffn Function
+
+\def\defun{\defparsebody\Edefun\defunx\defunheader}
+
+\def\defunheader #1#2{\doind {fn}{\code{#1}}% Make entry in function index
+\begingroup\defname {#1}{\putwordDeffunc}%
+\defunargs {#2}\endgroup %
+\catcode 61=\other % Turn off change made in \defparsebody
+}
+
+% @deftypefun int foobar (int @var{foo}, float @var{bar})
+
+\def\deftypefun{\defparsebody\Edeftypefun\deftypefunx\deftypefunheader}
+
+% #1 is the data type.  #2 is the name and args.
+\def\deftypefunheader #1#2{\deftypefunheaderx{#1}#2 \relax}
+% #1 is the data type, #2 the name, #3 the args.
+\def\deftypefunheaderx #1#2 #3\relax{%
+\doind {fn}{\code{#2}}% Make entry in function index
+\begingroup\defname {\defheaderxcond#1\relax$$$#2}{\putwordDeftypefun}%
+\deftypefunargs {#3}\endgroup %
+\catcode 61=\other % Turn off change made in \defparsebody
+}
+
+% @deftypefn {Library Function} int foobar (int @var{foo}, float @var{bar})
+
+\def\deftypefn{\defmethparsebody\Edeftypefn\deftypefnx\deftypefnheader}
+
+% \defheaderxcond#1\relax$$$
+% puts #1 in @code, followed by a space, but does nothing if #1 is null.
+\def\defheaderxcond#1#2$$${\ifx#1\relax\else\code{#1#2} \fi}
+
+% #1 is the classification.  #2 is the data type.  #3 is the name and args.
+\def\deftypefnheader #1#2#3{\deftypefnheaderx{#1}{#2}#3 \relax}
+% #1 is the classification, #2 the data type, #3 the name, #4 the args.
+\def\deftypefnheaderx #1#2#3 #4\relax{%
+\doind {fn}{\code{#3}}% Make entry in function index
+\begingroup
+\normalparens % notably, turn off `&' magic, which prevents
+%               at least some C++ text from working
+\defname {\defheaderxcond#2\relax$$$#3}{#1}%
+\deftypefunargs {#4}\endgroup %
+\catcode 61=\other % Turn off change made in \defparsebody
+}
+
+% @defmac == @deffn Macro
+
+\def\defmac{\defparsebody\Edefmac\defmacx\defmacheader}
+
+\def\defmacheader #1#2{\doind {fn}{\code{#1}}% Make entry in function index
+\begingroup\defname {#1}{\putwordDefmac}%
+\defunargs {#2}\endgroup %
+\catcode 61=\other % Turn off change made in \defparsebody
+}
+
+% @defspec == @deffn Special Form
+
+\def\defspec{\defparsebody\Edefspec\defspecx\defspecheader}
+
+\def\defspecheader #1#2{\doind {fn}{\code{#1}}% Make entry in function index
+\begingroup\defname {#1}{\putwordDefspec}%
+\defunargs {#2}\endgroup %
+\catcode 61=\other % Turn off change made in \defparsebody
+}
+
+% @defop CATEGORY CLASS OPERATION ARG...
+%
+\def\defop #1 {\def\defoptype{#1}%
+\defopparsebody\Edefop\defopx\defopheader\defoptype}
+%
+\def\defopheader#1#2#3{%
+\dosubind {fn}{\code{#2}}{\putwordon\ #1}% Make entry in function index
+\begingroup\defname {#2}{\defoptype\ \putwordon\ #1}%
+\defunargs {#3}\endgroup %
+}
+
+% @deftypeop CATEGORY CLASS TYPE OPERATION ARG...
+%
+\def\deftypeop #1 {\def\deftypeopcategory{#1}%
+  \deftypeopparsebody\Edeftypeop\deftypeopx\deftypeopheader
+                       \deftypeopcategory}
+%
+% #1 is the class name, #2 the data type, #3 the operation name, #4 the args.
+\def\deftypeopheader#1#2#3#4{%
+  \dosubind{fn}{\code{#3}}{\putwordon\ \code{#1}}% entry in function index
+  \begingroup
+    \defname{\defheaderxcond#2\relax$$$#3}
+            {\deftypeopcategory\ \putwordon\ \code{#1}}%
+    \deftypefunargs{#4}%
+  \endgroup
+}
+
+% @deftypemethod CLASS TYPE METHOD ARG...
+%
+\def\deftypemethod{%
+  \deftypemethparsebody\Edeftypemethod\deftypemethodx\deftypemethodheader}
+%
+% #1 is the class name, #2 the data type, #3 the method name, #4 the args.
+\def\deftypemethodheader#1#2#3#4{%
+  \dosubind{fn}{\code{#3}}{\putwordon\ \code{#1}}% entry in function index
+  \begingroup
+    \defname{\defheaderxcond#2\relax$$$#3}{\putwordMethodon\ \code{#1}}%
+    \deftypefunargs{#4}%
+  \endgroup
+}
+
+% @deftypeivar CLASS TYPE VARNAME
+%
+\def\deftypeivar{%
+  \deftypemethparsebody\Edeftypeivar\deftypeivarx\deftypeivarheader}
+%
+% #1 is the class name, #2 the data type, #3 the variable name.
+\def\deftypeivarheader#1#2#3{%
+  \dosubind{vr}{\code{#3}}{\putwordof\ \code{#1}}% entry in variable index
+  \begingroup
+    \defname{\defheaderxcond#2\relax$$$#3}
+            {\putwordInstanceVariableof\ \code{#1}}%
+    \defvarargs{#3}%
+  \endgroup
+}
+
+% @defmethod == @defop Method
+%
+\def\defmethod{\defmethparsebody\Edefmethod\defmethodx\defmethodheader}
+%
+% #1 is the class name, #2 the method name, #3 the args.
+\def\defmethodheader#1#2#3{%
+  \dosubind{fn}{\code{#2}}{\putwordon\ \code{#1}}% entry in function index
+  \begingroup
+    \defname{#2}{\putwordMethodon\ \code{#1}}%
+    \defunargs{#3}%
+  \endgroup
+}
+
+% @defcv {Class Option} foo-class foo-flag
+
+\def\defcv #1 {\def\defcvtype{#1}%
+\defopvarparsebody\Edefcv\defcvx\defcvarheader\defcvtype}
+
+\def\defcvarheader #1#2#3{%
+\dosubind {vr}{\code{#2}}{\putwordof\ #1}% Make entry in var index
+\begingroup\defname {#2}{\defcvtype\ \putwordof\ #1}%
+\defvarargs {#3}\endgroup %
+}
+
+% @defivar CLASS VARNAME == @defcv {Instance Variable} CLASS VARNAME
+%
+\def\defivar{\defvrparsebody\Edefivar\defivarx\defivarheader}
+%
+\def\defivarheader#1#2#3{%
+  \dosubind {vr}{\code{#2}}{\putwordof\ #1}% entry in var index
+  \begingroup
+    \defname{#2}{\putwordInstanceVariableof\ #1}%
+    \defvarargs{#3}%
+  \endgroup
+}
+
+% @defvar
+% First, define the processing that is wanted for arguments of @defvar.
+% This is actually simple: just print them in roman.
+% This must expand the args and terminate the paragraph they make up
+\def\defvarargs #1{\normalparens #1%
+\interlinepenalty=10000
+\endgraf\nobreak\vskip -\parskip\nobreak}
+
+% @defvr Counter foo-count
+
+\def\defvr{\defvrparsebody\Edefvr\defvrx\defvrheader}
+
+\def\defvrheader #1#2#3{\doind {vr}{\code{#2}}%
+\begingroup\defname {#2}{#1}\defvarargs{#3}\endgroup}
+
+% @defvar == @defvr Variable
+
+\def\defvar{\defvarparsebody\Edefvar\defvarx\defvarheader}
+
+\def\defvarheader #1#2{\doind {vr}{\code{#1}}% Make entry in var index
+\begingroup\defname {#1}{\putwordDefvar}%
+\defvarargs {#2}\endgroup %
+}
+
+% @defopt == @defvr {User Option}
+
+\def\defopt{\defvarparsebody\Edefopt\defoptx\defoptheader}
+
+\def\defoptheader #1#2{\doind {vr}{\code{#1}}% Make entry in var index
+\begingroup\defname {#1}{\putwordDefopt}%
+\defvarargs {#2}\endgroup %
+}
+
+% @deftypevar int foobar
+
+\def\deftypevar{\defvarparsebody\Edeftypevar\deftypevarx\deftypevarheader}
+
+% #1 is the data type.  #2 is the name, perhaps followed by text that
+% is actually part of the data type, which should not be put into the index.
+\def\deftypevarheader #1#2{%
+\dovarind#2 \relax% Make entry in variables index
+\begingroup\defname {\defheaderxcond#1\relax$$$#2}{\putwordDeftypevar}%
+\interlinepenalty=10000
+\endgraf\nobreak\vskip -\parskip\nobreak
+\endgroup}
+\def\dovarind#1 #2\relax{\doind{vr}{\code{#1}}}
+
+% @deftypevr {Global Flag} int enable
+
+\def\deftypevr{\defvrparsebody\Edeftypevr\deftypevrx\deftypevrheader}
+
+\def\deftypevrheader #1#2#3{\dovarind#3 \relax%
+\begingroup\defname {\defheaderxcond#2\relax$$$#3}{#1}
+\interlinepenalty=10000
+\endgraf\nobreak\vskip -\parskip\nobreak
+\endgroup}
+
+% Now define @deftp
+% Args are printed in bold, a slight difference from @defvar.
+
+\def\deftpargs #1{\bf \defvarargs{#1}}
+
+% @deftp Class window height width ...
+
+\def\deftp{\deftpparsebody\Edeftp\deftpx\deftpheader}
+
+\def\deftpheader #1#2#3{\doind {tp}{\code{#2}}%
+\begingroup\defname {#2}{#1}\deftpargs{#3}\endgroup}
+
+% These definitions are used if you use @defunx (etc.)
+% anywhere other than immediately after a @defun or @defunx.
+% 
+\def\defcvx#1 {\errmessage{@defcvx in invalid context}}
+\def\deffnx#1 {\errmessage{@deffnx in invalid context}}
+\def\defivarx#1 {\errmessage{@defivarx in invalid context}}
+\def\defmacx#1 {\errmessage{@defmacx in invalid context}}
+\def\defmethodx#1 {\errmessage{@defmethodx in invalid context}}
+\def\defoptx #1 {\errmessage{@defoptx in invalid context}}
+\def\defopx#1 {\errmessage{@defopx in invalid context}}
+\def\defspecx#1 {\errmessage{@defspecx in invalid context}}
+\def\deftpx#1 {\errmessage{@deftpx in invalid context}}
+\def\deftypefnx#1 {\errmessage{@deftypefnx in invalid context}}
+\def\deftypefunx#1 {\errmessage{@deftypefunx in invalid context}}
+\def\deftypeivarx#1 {\errmessage{@deftypeivarx in invalid context}}
+\def\deftypemethodx#1 {\errmessage{@deftypemethodx in invalid context}}
+\def\deftypeopx#1 {\errmessage{@deftypeopx in invalid context}}
+\def\deftypevarx#1 {\errmessage{@deftypevarx in invalid context}}
+\def\deftypevrx#1 {\errmessage{@deftypevrx in invalid context}}
+\def\defunx#1 {\errmessage{@defunx in invalid context}}
+\def\defvarx#1 {\errmessage{@defvarx in invalid context}}
+\def\defvrx#1 {\errmessage{@defvrx in invalid context}}
+
+
+\message{macros,}
+% @macro.
+
+% To do this right we need a feature of e-TeX, \scantokens,
+% which we arrange to emulate with a temporary file in ordinary TeX.
+\ifx\eTeXversion\undefined
+ \newwrite\macscribble
+ \def\scanmacro#1{%
+   \begingroup \newlinechar`\^^M
+   % Undo catcode changes of \startcontents and \doprintindex
+   \catcode`\@=0 \catcode`\\=12 \escapechar=`\@
+   % Append \endinput to make sure that TeX does not see the ending newline.
+   \toks0={#1\endinput}%
+   \immediate\openout\macscribble=\jobname.tmp
+   \immediate\write\macscribble{\the\toks0}%
+   \immediate\closeout\macscribble
+   \let\xeatspaces\eatspaces
+   \input \jobname.tmp
+   \endgroup
+}
+\else
+\def\scanmacro#1{%
+\begingroup \newlinechar`\^^M
+% Undo catcode changes of \startcontents and \doprintindex
+\catcode`\@=0 \catcode`\\=12 \escapechar=`\@
+\let\xeatspaces\eatspaces\scantokens{#1\endinput}\endgroup}
+\fi
+
+\newcount\paramno   % Count of parameters
+\newtoks\macname    % Macro name
+\newif\ifrecursive  % Is it recursive?
+\def\macrolist{}    % List of all defined macros in the form
+                    % \do\macro1\do\macro2...
+
+% Utility routines.
+% Thisdoes \let #1 = #2, except with \csnames.
+\def\cslet#1#2{%
+\expandafter\expandafter
+\expandafter\let
+\expandafter\expandafter
+\csname#1\endcsname
+\csname#2\endcsname}
+
+% Trim leading and trailing spaces off a string.
+% Concepts from aro-bend problem 15 (see CTAN).
+{\catcode`\@=11
+\gdef\eatspaces #1{\expandafter\trim@\expandafter{#1 }}
+\gdef\trim@ #1{\trim@@ @#1 @ #1 @ @@}
+\gdef\trim@@ #1@ #2@ #3@@{\trim@@@\empty #2 @}
+\def\unbrace#1{#1}
+\unbrace{\gdef\trim@@@ #1 } #2@{#1}
+}
+
+% Trim a single trailing ^^M off a string.
+{\catcode`\^^M=12\catcode`\Q=3%
+\gdef\eatcr #1{\eatcra #1Q^^MQ}%
+\gdef\eatcra#1^^MQ{\eatcrb#1Q}%
+\gdef\eatcrb#1Q#2Q{#1}%
+}
+
+% Macro bodies are absorbed as an argument in a context where
+% all characters are catcode 10, 11 or 12, except \ which is active
+% (as in normal texinfo). It is necessary to change the definition of \.
+
+% It's necessary to have hard CRs when the macro is executed. This is
+% done by  making ^^M (\endlinechar) catcode 12 when reading the macro
+% body, and then making it the \newlinechar in \scanmacro.
+
+\def\macrobodyctxt{%
+  \catcode`\~=12
+  \catcode`\^=12
+  \catcode`\_=12
+  \catcode`\|=12
+  \catcode`\<=12
+  \catcode`\>=12
+  \catcode`\+=12
+  \catcode`\{=12
+  \catcode`\}=12
+  \catcode`\@=12
+  \catcode`\^^M=12
+  \usembodybackslash}
+
+\def\macroargctxt{%
+  \catcode`\~=12
+  \catcode`\^=12
+  \catcode`\_=12
+  \catcode`\|=12
+  \catcode`\<=12
+  \catcode`\>=12
+  \catcode`\+=12
+  \catcode`\@=12
+  \catcode`\\=12}
+
+% \mbodybackslash is the definition of \ in @macro bodies.
+% It maps \foo\ => \csname macarg.foo\endcsname => #N
+% where N is the macro parameter number.
+% We define \csname macarg.\endcsname to be \realbackslash, so
+% \\ in macro replacement text gets you a backslash.
+
+{\catcode`@=0 @catcode`@\=@active
+ @gdef@usembodybackslash{@let\=@mbodybackslash}
+ @gdef@mbodybackslash#1\{@csname macarg.#1@endcsname}
+}
+\expandafter\def\csname macarg.\endcsname{\realbackslash}
+
+\def\macro{\recursivefalse\parsearg\macroxxx}
+\def\rmacro{\recursivetrue\parsearg\macroxxx}
+
+\def\macroxxx#1{%
+  \getargs{#1}%           now \macname is the macname and \argl the arglist
+  \ifx\argl\empty       % no arguments
+     \paramno=0%
+  \else
+     \expandafter\parsemargdef \argl;%
+  \fi
+  \if1\csname ismacro.\the\macname\endcsname
+     \message{Warning: redefining \the\macname}%
+  \else
+     \expandafter\ifx\csname \the\macname\endcsname \relax
+     \else \errmessage{The name \the\macname\space is reserved}\fi
+     \global\cslet{macsave.\the\macname}{\the\macname}%
+     \global\expandafter\let\csname ismacro.\the\macname\endcsname=1%
+     % Add the macroname to \macrolist
+     \toks0 = \expandafter{\macrolist\do}%
+     \xdef\macrolist{\the\toks0
+       \expandafter\noexpand\csname\the\macname\endcsname}%
+  \fi
+  \begingroup \macrobodyctxt
+  \ifrecursive \expandafter\parsermacbody
+  \else \expandafter\parsemacbody
+  \fi}
+
+\def\unmacro{\parsearg\unmacroxxx}
+\def\unmacroxxx#1{%
+  \if1\csname ismacro.#1\endcsname
+    \global\cslet{#1}{macsave.#1}%
+    \global\expandafter\let \csname ismacro.#1\endcsname=0%
+    % Remove the macro name from \macrolist
+    \begingroup
+      \edef\tempa{\expandafter\noexpand\csname#1\endcsname}%
+      \def\do##1{%
+        \def\tempb{##1}%
+        \ifx\tempa\tempb
+          % remove this
+        \else
+          \toks0 = \expandafter{\newmacrolist\do}%
+          \edef\newmacrolist{\the\toks0\expandafter\noexpand\tempa}%
+        \fi}%
+      \def\newmacrolist{}%
+      % Execute macro list to define \newmacrolist
+      \macrolist
+      \global\let\macrolist\newmacrolist
+    \endgroup
+  \else
+    \errmessage{Macro #1 not defined}%
+  \fi
+}
+
+% This makes use of the obscure feature that if the last token of a
+% <parameter list> is #, then the preceding argument is delimited by
+% an opening brace, and that opening brace is not consumed.
+\def\getargs#1{\getargsxxx#1{}}
+\def\getargsxxx#1#{\getmacname #1 \relax\getmacargs}
+\def\getmacname #1 #2\relax{\macname={#1}}
+\def\getmacargs#1{\def\argl{#1}}
+
+% Parse the optional {params} list.  Set up \paramno and \paramlist
+% so \defmacro knows what to do.  Define \macarg.blah for each blah
+% in the params list, to be ##N where N is the position in that list.
+% That gets used by \mbodybackslash (above).
+
+% We need to get `macro parameter char #' into several definitions.
+% The technique used is stolen from LaTeX:  let \hash be something
+% unexpandable, insert that wherever you need a #, and then redefine
+% it to # just before using the token list produced.
+%
+% The same technique is used to protect \eatspaces till just before
+% the macro is used.
+
+\def\parsemargdef#1;{\paramno=0\def\paramlist{}%
+        \let\hash\relax\let\xeatspaces\relax\parsemargdefxxx#1,;,}
+\def\parsemargdefxxx#1,{%
+  \if#1;\let\next=\relax
+  \else \let\next=\parsemargdefxxx
+    \advance\paramno by 1%
+    \expandafter\edef\csname macarg.\eatspaces{#1}\endcsname
+        {\xeatspaces{\hash\the\paramno}}%
+    \edef\paramlist{\paramlist\hash\the\paramno,}%
+  \fi\next}
+
+% These two commands read recursive and nonrecursive macro bodies.
+% (They're different since rec and nonrec macros end differently.)
+
+\long\def\parsemacbody#1@end macro%
+{\xdef\temp{\eatcr{#1}}\endgroup\defmacro}%
+\long\def\parsermacbody#1@end rmacro%
+{\xdef\temp{\eatcr{#1}}\endgroup\defmacro}%
+
+% This defines the macro itself. There are six cases: recursive and
+% nonrecursive macros of zero, one, and many arguments.
+% Much magic with \expandafter here.
+% \xdef is used so that macro definitions will survive the file
+% they're defined in; @include reads the file inside a group.
+\def\defmacro{%
+  \let\hash=##% convert placeholders to macro parameter chars
+  \ifrecursive
+    \ifcase\paramno
+    % 0
+      \expandafter\xdef\csname\the\macname\endcsname{%
+        \noexpand\scanmacro{\temp}}%
+    \or % 1
+      \expandafter\xdef\csname\the\macname\endcsname{%
+         \bgroup\noexpand\macroargctxt
+         \noexpand\braceorline
+         \expandafter\noexpand\csname\the\macname xxx\endcsname}%
+      \expandafter\xdef\csname\the\macname xxx\endcsname##1{%
+         \egroup\noexpand\scanmacro{\temp}}%
+    \else % many
+      \expandafter\xdef\csname\the\macname\endcsname{%
+         \bgroup\noexpand\macroargctxt
+         \noexpand\csname\the\macname xx\endcsname}%
+      \expandafter\xdef\csname\the\macname xx\endcsname##1{%
+          \expandafter\noexpand\csname\the\macname xxx\endcsname ##1,}%
+      \expandafter\expandafter
+      \expandafter\xdef
+      \expandafter\expandafter
+        \csname\the\macname xxx\endcsname
+          \paramlist{\egroup\noexpand\scanmacro{\temp}}%
+    \fi
+  \else
+    \ifcase\paramno
+    % 0
+      \expandafter\xdef\csname\the\macname\endcsname{%
+        \noexpand\norecurse{\the\macname}%
+        \noexpand\scanmacro{\temp}\egroup}%
+    \or % 1
+      \expandafter\xdef\csname\the\macname\endcsname{%
+         \bgroup\noexpand\macroargctxt
+         \noexpand\braceorline
+         \expandafter\noexpand\csname\the\macname xxx\endcsname}%
+      \expandafter\xdef\csname\the\macname xxx\endcsname##1{%
+        \egroup
+        \noexpand\norecurse{\the\macname}%
+        \noexpand\scanmacro{\temp}\egroup}%
+    \else % many
+      \expandafter\xdef\csname\the\macname\endcsname{%
+         \bgroup\noexpand\macroargctxt
+         \expandafter\noexpand\csname\the\macname xx\endcsname}%
+      \expandafter\xdef\csname\the\macname xx\endcsname##1{%
+          \expandafter\noexpand\csname\the\macname xxx\endcsname ##1,}%
+      \expandafter\expandafter
+      \expandafter\xdef
+      \expandafter\expandafter
+      \csname\the\macname xxx\endcsname
+      \paramlist{%
+          \egroup
+          \noexpand\norecurse{\the\macname}%
+          \noexpand\scanmacro{\temp}\egroup}%
+    \fi
+  \fi}
+
+\def\norecurse#1{\bgroup\cslet{#1}{macsave.#1}}
+
+% \braceorline decides whether the next nonwhitespace character is a
+% {.  If so it reads up to the closing }, if not, it reads the whole
+% line.  Whatever was read is then fed to the next control sequence
+% as an argument (by \parsebrace or \parsearg)
+\def\braceorline#1{\let\next=#1\futurelet\nchar\braceorlinexxx}
+\def\braceorlinexxx{%
+  \ifx\nchar\bgroup\else
+    \expandafter\parsearg
+  \fi \next}
+
+% We mant to disable all macros during \shipout so that they are not
+% expanded by \write.
+\def\turnoffmacros{\begingroup \def\do##1{\let\noexpand##1=\relax}%
+  \edef\next{\macrolist}\expandafter\endgroup\next}
+
+
+% @alias.
+% We need some trickery to remove the optional spaces around the equal
+% sign.  Just make them active and then expand them all to nothing.
+\def\alias{\begingroup\obeyspaces\parsearg\aliasxxx}
+\def\aliasxxx #1{\aliasyyy#1\relax}
+\def\aliasyyy #1=#2\relax{\ignoreactivespaces
+\edef\next{\global\let\expandafter\noexpand\csname#1\endcsname=%
+           \expandafter\noexpand\csname#2\endcsname}%
+\expandafter\endgroup\next}
+
+
+\message{cross references,}
+% @xref etc.
+
+\newwrite\auxfile
+
+\newif\ifhavexrefs    % True if xref values are known.
+\newif\ifwarnedxrefs  % True if we warned once that they aren't known.
+
+% @inforef is relatively simple.
+\def\inforef #1{\inforefzzz #1,,,,**}
+\def\inforefzzz #1,#2,#3,#4**{\putwordSee{} \putwordInfo{} \putwordfile{} \file{\ignorespaces #3{}},
+  node \samp{\ignorespaces#1{}}}
+
+% @node's job is to define \lastnode.
+\def\node{\ENVcheck\parsearg\nodezzz}
+\def\nodezzz#1{\nodexxx [#1,]}
+\def\nodexxx[#1,#2]{\gdef\lastnode{#1}}
+\let\nwnode=\node
+\let\lastnode=\relax
+
+% The sectioning commands (@chapter, etc.) call these.
+\def\donoderef{%
+  \ifx\lastnode\relax\else
+    \expandafter\expandafter\expandafter\setref{\lastnode}%
+      {Ysectionnumberandtype}%
+    \global\let\lastnode=\relax
+  \fi
+}
+\def\unnumbnoderef{%
+  \ifx\lastnode\relax\else
+    \expandafter\expandafter\expandafter\setref{\lastnode}{Ynothing}%
+    \global\let\lastnode=\relax
+  \fi
+}
+\def\appendixnoderef{%
+  \ifx\lastnode\relax\else
+    \expandafter\expandafter\expandafter\setref{\lastnode}%
+      {Yappendixletterandtype}%
+    \global\let\lastnode=\relax
+  \fi
+}
+
+
+% @anchor{NAME} -- define xref target at arbitrary point.
+%
+\newcount\savesfregister
+\gdef\savesf{\relax \ifhmode \savesfregister=\spacefactor \fi}
+\gdef\restoresf{\relax \ifhmode \spacefactor=\savesfregister \fi}
+\gdef\anchor#1{\savesf \setref{#1}{Ynothing}\restoresf \ignorespaces}
+
+% \setref{NAME}{SNT} defines a cross-reference point NAME, namely
+% NAME-title, NAME-pg, and NAME-SNT.  Called from \foonoderef.  We have
+% to set \indexdummies so commands such as @code in a section title
+% aren't expanded.  It would be nicer not to expand the titles in the
+% first place, but there's so many layers that that is hard to do.
+%
+\def\setref#1#2{{%
+  \indexdummies
+  \pdfmkdest{#1}%
+  \dosetq{#1-title}{Ytitle}%
+  \dosetq{#1-pg}{Ypagenumber}%
+  \dosetq{#1-snt}{#2}%
+}}
+
+% @xref, @pxref, and @ref generate cross-references.  For \xrefX, #1 is
+% the node name, #2 the name of the Info cross-reference, #3 the printed
+% node name, #4 the name of the Info file, #5 the name of the printed
+% manual.  All but the node name can be omitted.
+%
+\def\pxref#1{\putwordsee{} \xrefX[#1,,,,,,,]}
+\def\xref#1{\putwordSee{} \xrefX[#1,,,,,,,]}
+\def\ref#1{\xrefX[#1,,,,,,,]}
+\def\xrefX[#1,#2,#3,#4,#5,#6]{\begingroup
+  \unsepspaces
+  \def\printedmanual{\ignorespaces #5}%
+  \def\printednodename{\ignorespaces #3}%
+  \setbox1=\hbox{\printedmanual}%
+  \setbox0=\hbox{\printednodename}%
+  \ifdim \wd0 = 0pt
+    % No printed node name was explicitly given.
+    \expandafter\ifx\csname SETxref-automatic-section-title\endcsname\relax
+      % Use the node name inside the square brackets.
+      \def\printednodename{\ignorespaces #1}%
+    \else
+      % Use the actual chapter/section title appear inside
+      % the square brackets.  Use the real section title if we have it.
+      \ifdim \wd1 > 0pt
+        % It is in another manual, so we don't have it.
+        \def\printednodename{\ignorespaces #1}%
+      \else
+        \ifhavexrefs
+          % We know the real title if we have the xref values.
+          \def\printednodename{\refx{#1-title}{}}%
+        \else
+          % Otherwise just copy the Info node name.
+          \def\printednodename{\ignorespaces #1}%
+        \fi%
+      \fi
+    \fi
+  \fi
+  %
+  % If we use \unhbox0 and \unhbox1 to print the node names, TeX does not
+  % insert empty discretionaries after hyphens, which means that it will
+  % not find a line break at a hyphen in a node names.  Since some manuals
+  % are best written with fairly long node names, containing hyphens, this
+  % is a loss.  Therefore, we give the text of the node name again, so it
+  % is as if TeX is seeing it for the first time.
+  \ifpdf
+    \leavevmode
+    \getfilename{#4}%
+    \ifnum\filenamelength>0
+      \startlink attr{/Border [0 0 0]}%
+        goto file{\the\filename.pdf} name{#1@}%
+    \else
+      \startlink attr{/Border [0 0 0]}%
+        goto name{#1@}%
+    \fi
+    \linkcolor
+  \fi
+  %
+  \ifdim \wd1 > 0pt
+    \putwordsection{} ``\printednodename'' \putwordin{} \cite{\printedmanual}%
+  \else
+    % _ (for example) has to be the character _ for the purposes of the
+    % control sequence corresponding to the node, but it has to expand
+    % into the usual \leavevmode...\vrule stuff for purposes of
+    % printing. So we \turnoffactive for the \refx-snt, back on for the
+    % printing, back off for the \refx-pg.
+    {\normalturnoffactive
+     % Only output a following space if the -snt ref is nonempty; for
+     % @unnumbered and @anchor, it won't be.
+     \setbox2 = \hbox{\ignorespaces \refx{#1-snt}{}}%
+     \ifdim \wd2 > 0pt \refx{#1-snt}\space\fi
+    }%
+    % [mynode],
+    [\printednodename],\space
+    % page 3
+    \turnoffactive \putwordpage\tie\refx{#1-pg}{}%
+  \fi
+  \endlink
+\endgroup}
+
+% \dosetq is the interface for calls from other macros
+
+% Use \normalturnoffactive so that punctuation chars such as underscore
+% and backslash work in node names.  (\turnoffactive doesn't do \.)
+\def\dosetq#1#2{%
+  {\let\folio=0%
+   \normalturnoffactive
+   \edef\next{\write\auxfile{\internalsetq{#1}{#2}}}%
+   \iflinks
+     \next
+   \fi
+  }%
+}
+
+% \internalsetq {foo}{page} expands into
+% CHARACTERS 'xrdef {foo}{...expansion of \Ypage...}
+% When the aux file is read, ' is the escape character
+
+\def\internalsetq #1#2{'xrdef {#1}{\csname #2\endcsname}}
+
+% Things to be expanded by \internalsetq
+
+\def\Ypagenumber{\folio}
+
+\def\Ytitle{\thissection}
+
+\def\Ynothing{}
+
+\def\Ysectionnumberandtype{%
+\ifnum\secno=0 \putwordChapter\xreftie\the\chapno %
+\else \ifnum \subsecno=0 \putwordSection\xreftie\the\chapno.\the\secno %
+\else \ifnum \subsubsecno=0 %
+\putwordSection\xreftie\the\chapno.\the\secno.\the\subsecno %
+\else %
+\putwordSection\xreftie\the\chapno.\the\secno.\the\subsecno.\the\subsubsecno %
+\fi \fi \fi }
+
+\def\Yappendixletterandtype{%
+\ifnum\secno=0 \putwordAppendix\xreftie'char\the\appendixno{}%
+\else \ifnum \subsecno=0 \putwordSection\xreftie'char\the\appendixno.\the\secno %
+\else \ifnum \subsubsecno=0 %
+\putwordSection\xreftie'char\the\appendixno.\the\secno.\the\subsecno %
+\else %
+\putwordSection\xreftie'char\the\appendixno.\the\secno.\the\subsecno.\the\subsubsecno %
+\fi \fi \fi }
+
+\gdef\xreftie{'tie}
+
+% Use TeX 3.0's \inputlineno to get the line number, for better error
+% messages, but if we're using an old version of TeX, don't do anything.
+%
+\ifx\inputlineno\thisisundefined
+  \let\linenumber = \empty % Non-3.0.
+\else
+  \def\linenumber{\the\inputlineno:\space}
+\fi
+
+% Define \refx{NAME}{SUFFIX} to reference a cross-reference string named NAME.
+% If its value is nonempty, SUFFIX is output afterward.
+
+\def\refx#1#2{%
+  \expandafter\ifx\csname X#1\endcsname\relax
+    % If not defined, say something at least.
+    \angleleft un\-de\-fined\angleright
+    \iflinks
+      \ifhavexrefs
+        \message{\linenumber Undefined cross reference `#1'.}%
+      \else
+        \ifwarnedxrefs\else
+          \global\warnedxrefstrue
+          \message{Cross reference values unknown; you must run TeX again.}%
+        \fi
+      \fi
+    \fi
+  \else
+    % It's defined, so just use it.
+    \csname X#1\endcsname
+  \fi
+  #2% Output the suffix in any case.
+}
+
+% This is the macro invoked by entries in the aux file.
+%
+\def\xrdef#1{\begingroup
+  % Reenable \ as an escape while reading the second argument.
+  \catcode`\\ = 0
+  \afterassignment\endgroup
+  \expandafter\gdef\csname X#1\endcsname
+}
+
+% Read the last existing aux file, if any.  No error if none exists.
+\def\readauxfile{\begingroup
+  \catcode`\^^@=\other
+  \catcode`\^^A=\other
+  \catcode`\^^B=\other
+  \catcode`\^^C=\other
+  \catcode`\^^D=\other
+  \catcode`\^^E=\other
+  \catcode`\^^F=\other
+  \catcode`\^^G=\other
+  \catcode`\^^H=\other
+  \catcode`\^^K=\other
+  \catcode`\^^L=\other
+  \catcode`\^^N=\other
+  \catcode`\^^P=\other
+  \catcode`\^^Q=\other
+  \catcode`\^^R=\other
+  \catcode`\^^S=\other
+  \catcode`\^^T=\other
+  \catcode`\^^U=\other
+  \catcode`\^^V=\other
+  \catcode`\^^W=\other
+  \catcode`\^^X=\other
+  \catcode`\^^Z=\other
+  \catcode`\^^[=\other
+  \catcode`\^^\=\other
+  \catcode`\^^]=\other
+  \catcode`\^^^=\other
+  \catcode`\^^_=\other
+  \catcode`\@=\other
+  \catcode`\^=\other
+  % It was suggested to define this as 7, which would allow ^^e4 etc.
+  % in xref tags, i.e., node names.  But since ^^e4 notation isn't
+  % supported in the main text, it doesn't seem desirable.  Furthermore,
+  % that is not enough: for node names that actually contain a ^
+  % character, we would end up writing a line like this: 'xrdef {'hat
+  % b-title}{'hat b} and \xrdef does a \csname...\endcsname on the first
+  % argument, and \hat is not an expandable control sequence.  It could
+  % all be worked out, but why?  Either we support ^^ or we don't.
+  %
+  % The other change necessary for this was to define \auxhat:
+  % \def\auxhat{\def^{'hat }}% extra space so ok if followed by letter
+  % and then to call \auxhat in \setq.
+  %
+  \catcode`\~=\other
+  \catcode`\[=\other
+  \catcode`\]=\other
+  \catcode`\"=\other
+  \catcode`\_=\other
+  \catcode`\|=\other
+  \catcode`\<=\other
+  \catcode`\>=\other
+  \catcode`\$=\other
+  \catcode`\#=\other
+  \catcode`\&=\other
+  \catcode`+=\other % avoid \+ for paranoia even though we've turned it off
+  % Make the characters 128-255 be printing characters
+  {%
+    \count 1=128
+    \def\loop{%
+      \catcode\count 1=\other
+      \advance\count 1 by 1
+      \ifnum \count 1<256 \loop \fi
+    }%
+  }%
+  % The aux file uses ' as the escape (for now).
+  % Turn off \ as an escape so we do not lose on
+  % entries which were dumped with control sequences in their names.
+  % For example, 'xrdef {$\leq $-fun}{page ...} made by @defun ^^
+  % Reference to such entries still does not work the way one would wish,
+  % but at least they do not bomb out when the aux file is read in.
+  \catcode`\{=1
+  \catcode`\}=2
+  \catcode`\%=\other
+  \catcode`\'=0
+  \catcode`\\=\other
+  %
+  \openin 1 \jobname.aux
+  \ifeof 1 \else
+    \closein 1
+    \input \jobname.aux
+    \global\havexrefstrue
+    \global\warnedobstrue
+  \fi
+  % Open the new aux file.  TeX will close it automatically at exit.
+  \openout\auxfile=\jobname.aux
+\endgroup}
+
+
+% Footnotes.
+
+\newcount \footnoteno
+
+% The trailing space in the following definition for supereject is
+% vital for proper filling; pages come out unaligned when you do a
+% pagealignmacro call if that space before the closing brace is
+% removed. (Generally, numeric constants should always be followed by a
+% space to prevent strange expansion errors.)
+\def\supereject{\par\penalty -20000\footnoteno =0 }
+
+% @footnotestyle is meaningful for info output only.
+\let\footnotestyle=\comment
+
+\let\ptexfootnote=\footnote
+
+{\catcode `\@=11
+%
+% Auto-number footnotes.  Otherwise like plain.
+\gdef\footnote{%
+  \global\advance\footnoteno by \@ne
+  \edef\thisfootno{$^{\the\footnoteno}$}%
+  %
+  % In case the footnote comes at the end of a sentence, preserve the
+  % extra spacing after we do the footnote number.
+  \let\@sf\empty
+  \ifhmode\edef\@sf{\spacefactor\the\spacefactor}\/\fi
+  %
+  % Remove inadvertent blank space before typesetting the footnote number.
+  \unskip
+  \thisfootno\@sf
+  \footnotezzz
+}%
+
+% Don't bother with the trickery in plain.tex to not require the
+% footnote text as a parameter.  Our footnotes don't need to be so general.
+%
+% Oh yes, they do; otherwise, @ifset and anything else that uses
+% \parseargline fail inside footnotes because the tokens are fixed when
+% the footnote is read.  --karl, 16nov96.
+%
+\long\gdef\footnotezzz{\insert\footins\bgroup
+  % We want to typeset this text as a normal paragraph, even if the
+  % footnote reference occurs in (for example) a display environment.
+  % So reset some parameters.
+  \interlinepenalty\interfootnotelinepenalty
+  \splittopskip\ht\strutbox % top baseline for broken footnotes
+  \splitmaxdepth\dp\strutbox
+  \floatingpenalty\@MM
+  \leftskip\z@skip
+  \rightskip\z@skip
+  \spaceskip\z@skip
+  \xspaceskip\z@skip
+  \parindent\defaultparindent
+  %
+  \smallfonts \rm
+  %
+  % Hang the footnote text off the number.
+  \hang
+  \textindent{\thisfootno}%
+  %
+  % Don't crash into the line above the footnote text.  Since this
+  % expands into a box, it must come within the paragraph, lest it
+  % provide a place where TeX can split the footnote.
+  \footstrut
+  \futurelet\next\fo@t
+}
+\def\fo@t{\ifcat\bgroup\noexpand\next \let\next\f@@t
+  \else\let\next\f@t\fi \next}
+\def\f@@t{\bgroup\aftergroup\@foot\let\next}
+\def\f@t#1{#1\@foot}
+\def\@foot{\strut\par\egroup}
+
+}%end \catcode `\@=11
+
+% Set the baselineskip to #1, and the lineskip and strut size
+% correspondingly.  There is no deep meaning behind these magic numbers
+% used as factors; they just match (closely enough) what Knuth defined.
+%
+\def\lineskipfactor{.08333}
+\def\strutheightpercent{.70833}
+\def\strutdepthpercent {.29167}
+%
+\def\setleading#1{%
+  \normalbaselineskip = #1\relax
+  \normallineskip = \lineskipfactor\normalbaselineskip
+  \normalbaselines
+  \setbox\strutbox =\hbox{%
+    \vrule width0pt height\strutheightpercent\baselineskip
+                    depth \strutdepthpercent \baselineskip
+  }%
+}
+
+% @| inserts a changebar to the left of the current line.  It should
+% surround any changed text.  This approach does *not* work if the
+% change spans more than two lines of output.  To handle that, we would
+% have adopt a much more difficult approach (putting marks into the main
+% vertical list for the beginning and end of each change).
+%
+\def\|{%
+  % \vadjust can only be used in horizontal mode.
+  \leavevmode
+  %
+  % Append this vertical mode material after the current line in the output.
+  \vadjust{%
+    % We want to insert a rule with the height and depth of the current
+    % leading; that is exactly what \strutbox is supposed to record.
+    \vskip-\baselineskip
+    %
+    % \vadjust-items are inserted at the left edge of the type.  So
+    % the \llap here moves out into the left-hand margin.
+    \llap{%
+      %
+      % For a thicker or thinner bar, change the `1pt'.
+      \vrule height\baselineskip width1pt
+      %
+      % This is the space between the bar and the text.
+      \hskip 12pt
+    }%
+  }%
+}
+
+% For a final copy, take out the rectangles
+% that mark overfull boxes (in case you have decided
+% that the text looks ok even though it passes the margin).
+%
+\def\finalout{\overfullrule=0pt}
+
+% @image.  We use the macros from epsf.tex to support this.
+% If epsf.tex is not installed and @image is used, we complain.
+%
+% Check for and read epsf.tex up front.  If we read it only at @image
+% time, we might be inside a group, and then its definitions would get
+% undone and the next image would fail.
+\openin 1 = epsf.tex
+\ifeof 1 \else
+  \closein 1
+  % Do not bother showing banner with post-v2.7 epsf.tex (available in
+  % doc/epsf.tex until it shows up on ctan).
+  \def\epsfannounce{\toks0 = }%
+  \input epsf.tex
+\fi
+%
+% We will only complain once about lack of epsf.tex.
+\newif\ifwarnednoepsf
+\newhelp\noepsfhelp{epsf.tex must be installed for images to
+  work.  It is also included in the Texinfo distribution, or you can get
+  it from ftp://tug.org/tex/epsf.tex.}
+%
+\def\image#1{%
+  \ifx\epsfbox\undefined
+    \ifwarnednoepsf \else
+      \errhelp = \noepsfhelp
+      \errmessage{epsf.tex not found, images will be ignored}%
+      \global\warnednoepsftrue
+    \fi
+  \else
+    \imagexxx #1,,,\finish
+  \fi
+}
+%
+% Arguments to @image:
+% #1 is (mandatory) image filename; we tack on .eps extension.
+% #2 is (optional) width, #3 is (optional) height.
+% #4 is just the usual extra ignored arg for parsing this stuff.
+\def\imagexxx#1,#2,#3,#4\finish{%
+  \ifpdf
+    \centerline{\dopdfimage{#1}{#2}{#3}}%
+  \else
+    % \epsfbox itself resets \epsf?size at each figure.
+    \setbox0 = \hbox{\ignorespaces #2}\ifdim\wd0 > 0pt \epsfxsize=#2\relax \fi
+    \setbox0 = \hbox{\ignorespaces #3}\ifdim\wd0 > 0pt \epsfysize=#3\relax \fi
+    \begingroup
+      \catcode`\^^M = 5 % in case we're inside an example
+      % If the image is by itself, center it.
+      \ifvmode
+        \nobreak\bigskip
+        % Usually we'll have text after the image which will insert
+        % \parskip glue, so insert it here too to equalize the space
+        % above and below. 
+        \nobreak\vskip\parskip
+        \nobreak
+        \centerline{\epsfbox{#1.eps}}%
+        \bigbreak
+      \else
+        % In the middle of a paragraph, no extra space.
+        \epsfbox{#1.eps}%
+      \fi
+    \endgroup
+  \fi
+}
+
+
+\message{localization,}
+% and i18n.
+
+% @documentlanguage is usually given very early, just after
+% @setfilename.  If done too late, it may not override everything
+% properly.  Single argument is the language abbreviation.
+% It would be nice if we could set up a hyphenation file here.
+%
+\def\documentlanguage{\parsearg\dodocumentlanguage}
+\def\dodocumentlanguage#1{%
+  \tex % read txi-??.tex file in plain TeX.
+  % Read the file if it exists.
+  \openin 1 txi-#1.tex
+  \ifeof1
+    \errhelp = \nolanghelp
+    \errmessage{Cannot read language file txi-#1.tex}%
+    \let\temp = \relax
+  \else
+    \def\temp{\input txi-#1.tex }%
+  \fi
+  \temp
+  \endgroup
+}
+\newhelp\nolanghelp{The given language definition file cannot be found or
+is empty.  Maybe you need to install it?  In the current directory
+should work if nowhere else does.}
+
+
+% @documentencoding should change something in TeX eventually, most
+% likely, but for now just recognize it.
+\let\documentencoding = \comment
+
+
+% Page size parameters.
+%
+\newdimen\defaultparindent \defaultparindent = 15pt
+
+\chapheadingskip = 15pt plus 4pt minus 2pt
+\secheadingskip = 12pt plus 3pt minus 2pt
+\subsecheadingskip = 9pt plus 2pt minus 2pt
+
+% Prevent underfull vbox error messages.
+\vbadness = 10000
+
+% Don't be so finicky about underfull hboxes, either.
+\hbadness = 2000
+
+% Following George Bush, just get rid of widows and orphans.
+\widowpenalty=10000
+\clubpenalty=10000
+
+% Use TeX 3.0's \emergencystretch to help line breaking, but if we're
+% using an old version of TeX, don't do anything.  We want the amount of
+% stretch added to depend on the line length, hence the dependence on
+% \hsize.  We call this whenever the paper size is set.
+%
+\def\setemergencystretch{%
+  \ifx\emergencystretch\thisisundefined
+    % Allow us to assign to \emergencystretch anyway.
+    \def\emergencystretch{\dimen0}%
+  \else
+    \emergencystretch = .15\hsize
+  \fi
+}
+
+% Parameters in order: 1) textheight; 2) textwidth; 3) voffset;
+% 4) hoffset; 5) binding offset; 6) topskip.  Then whoever calls us can
+% set \parskip and call \setleading for \baselineskip.
+%
+\def\internalpagesizes#1#2#3#4#5#6{%
+  \voffset = #3\relax
+  \topskip = #6\relax
+  \splittopskip = \topskip
+  %
+  \vsize = #1\relax
+  \advance\vsize by \topskip
+  \outervsize = \vsize
+  \advance\outervsize by 2\topandbottommargin
+  \pageheight = \vsize
+  %
+  \hsize = #2\relax
+  \outerhsize = \hsize
+  \advance\outerhsize by 0.5in
+  \pagewidth = \hsize
+  %
+  \normaloffset = #4\relax
+  \bindingoffset = #5\relax
+  %
+  \parindent = \defaultparindent
+  \setemergencystretch
+}
+
+% @letterpaper (the default).
+\def\letterpaper{{\globaldefs = 1
+  \parskip = 3pt plus 2pt minus 1pt
+  \setleading{13.2pt}%
+  %
+  % If page is nothing but text, make it come out even.
+  \internalpagesizes{46\baselineskip}{6in}{\voffset}{.25in}{\bindingoffset}{36pt}%
+}}
+
+% Use @smallbook to reset parameters for 7x9.5 (or so) format.
+\def\smallbook{{\globaldefs = 1
+  \parskip = 2pt plus 1pt
+  \setleading{12pt}%
+  %
+  \internalpagesizes{7.5in}{5.in}{\voffset}{.25in}{\bindingoffset}{16pt}%
+  %
+  \lispnarrowing = 0.3in
+  \tolerance = 700
+  \hfuzz = 1pt
+  \contentsrightmargin = 0pt
+  \deftypemargin = 0pt
+  \defbodyindent = .5cm
+  %
+  \let\smalldisplay = \smalldisplayx
+  \let\smallexample = \smalllispx
+  \let\smallformat = \smallformatx
+  \let\smalllisp = \smalllispx
+}}
+
+% Use @afourpaper to print on European A4 paper.
+\def\afourpaper{{\globaldefs = 1
+  \setleading{12pt}%
+  \parskip = 3pt plus 2pt minus 1pt
+  %
+  \internalpagesizes{53\baselineskip}{160mm}{\voffset}{4mm}{\bindingoffset}{44pt}%
+  %
+  \tolerance = 700
+  \hfuzz = 1pt
+}}
+
+% A specific text layout, 24x15cm overall, intended for A4 paper.  Top margin
+% 29mm, hence bottom margin 28mm, nominal side margin 3cm.
+\def\afourlatex{{\globaldefs = 1
+  \setleading{13.6pt}%
+  %
+  \afourpaper
+  \internalpagesizes{237mm}{150mm}{3.6mm}{3.6mm}{3mm}{7mm}%
+  %
+  \globaldefs = 0
+}}
+
+% Use @afourwide to print on European A4 paper in wide format.
+\def\afourwide{%
+  \afourpaper
+  \internalpagesizes{6.5in}{9.5in}{\hoffset}{\normaloffset}{\bindingoffset}{7mm}%
+  %
+  \globaldefs = 0
+}
+
+% @pagesizes TEXTHEIGHT[,TEXTWIDTH]
+% Perhaps we should allow setting the margins, \topskip, \parskip,
+% and/or leading, also. Or perhaps we should compute them somehow.
+%
+\def\pagesizes{\parsearg\pagesizesxxx}
+\def\pagesizesxxx#1{\pagesizesyyy #1,,\finish}
+\def\pagesizesyyy#1,#2,#3\finish{{%
+  \setbox0 = \hbox{\ignorespaces #2}\ifdim\wd0 > 0pt \hsize=#2\relax \fi
+  \globaldefs = 1
+  %
+  \parskip = 3pt plus 2pt minus 1pt
+  \setleading{13.2pt}%
+  %
+  \internalpagesizes{#1}{\hsize}{\voffset}{\normaloffset}{\bindingoffset}{44pt}%
+}}
+
+% Set default to letter.
+%
+\letterpaper
+
+
+\message{and turning on texinfo input format.}
+
+% Define macros to output various characters with catcode for normal text.
+\catcode`\"=\other
+\catcode`\~=\other
+\catcode`\^=\other
+\catcode`\_=\other
+\catcode`\|=\other
+\catcode`\<=\other
+\catcode`\>=\other
+\catcode`\+=\other
+\catcode`\$=\other
+\def\normaldoublequote{"}
+\def\normaltilde{~}
+\def\normalcaret{^}
+\def\normalunderscore{_}
+\def\normalverticalbar{|}
+\def\normalless{<}
+\def\normalgreater{>}
+\def\normalplus{+}
+\def\normaldollar{$}
+
+% This macro is used to make a character print one way in ttfont
+% where it can probably just be output, and another way in other fonts,
+% where something hairier probably needs to be done.
+%
+% #1 is what to print if we are indeed using \tt; #2 is what to print
+% otherwise.  Since all the Computer Modern typewriter fonts have zero
+% interword stretch (and shrink), and it is reasonable to expect all
+% typewriter fonts to have this, we can check that font parameter.
+%
+\def\ifusingtt#1#2{\ifdim \fontdimen3\font=0pt #1\else #2\fi}
+
+% Same as above, but check for italic font.  Actually this also catches
+% non-italic slanted fonts since it is impossible to distinguish them from
+% italic fonts.  But since this is only used by $ and it uses \sl anyway
+% this is not a problem.
+\def\ifusingit#1#2{\ifdim \fontdimen1\font>0pt #1\else #2\fi}
+
+% Turn off all special characters except @
+% (and those which the user can use as if they were ordinary).
+% Most of these we simply print from the \tt font, but for some, we can
+% use math or other variants that look better in normal text.
+
+\catcode`\"=\active
+\def\activedoublequote{{\tt\char34}}
+\let"=\activedoublequote
+\catcode`\~=\active
+\def~{{\tt\char126}}
+\chardef\hat=`\^
+\catcode`\^=\active
+\def^{{\tt \hat}}
+
+\catcode`\_=\active
+\def_{\ifusingtt\normalunderscore\_}
+% Subroutine for the previous macro.
+\def\_{\leavevmode \kern.06em \vbox{\hrule width.3em height.1ex}}
+
+\catcode`\|=\active
+\def|{{\tt\char124}}
+\chardef \less=`\<
+\catcode`\<=\active
+\def<{{\tt \less}}
+\chardef \gtr=`\>
+\catcode`\>=\active
+\def>{{\tt \gtr}}
+\catcode`\+=\active
+\def+{{\tt \char 43}}
+\catcode`\$=\active
+\def${\ifusingit{{\sl\$}}\normaldollar}
+%\catcode 27=\active
+%\def^^[{$\diamondsuit$}
+
+% Set up an active definition for =, but don't enable it most of the time.
+{\catcode`\==\active
+\global\def={{\tt \char 61}}}
+
+\catcode`+=\active
+\catcode`\_=\active
+
+% If a .fmt file is being used, characters that might appear in a file
+% name cannot be active until we have parsed the command line.
+% So turn them off again, and have \everyjob (or @setfilename) turn them on.
+% \otherifyactive is called near the end of this file.
+\def\otherifyactive{\catcode`+=\other \catcode`\_=\other}
+
+\catcode`\@=0
+
+% \rawbackslashxx output one backslash character in current font
+\global\chardef\rawbackslashxx=`\\
+%{\catcode`\\=\other
+%@gdef@rawbackslashxx{\}}
+
+% \rawbackslash redefines \ as input to do \rawbackslashxx.
+{\catcode`\\=\active
+@gdef@rawbackslash{@let\=@rawbackslashxx }}
+
+% \normalbackslash outputs one backslash in fixed width font.
+\def\normalbackslash{{\tt\rawbackslashxx}}
+
+% \catcode 17=0   % Define control-q
+\catcode`\\=\active
+
+% Used sometimes to turn off (effectively) the active characters
+% even after parsing them.
+@def@turnoffactive{@let"=@normaldoublequote
+@let\=@realbackslash
+@let~=@normaltilde
+@let^=@normalcaret
+@let_=@normalunderscore
+@let|=@normalverticalbar
+@let<=@normalless
+@let>=@normalgreater
+@let+=@normalplus
+@let$=@normaldollar}
+
+@def@normalturnoffactive{@let"=@normaldoublequote
+@let\=@normalbackslash
+@let~=@normaltilde
+@let^=@normalcaret
+@let_=@normalunderscore
+@let|=@normalverticalbar
+@let<=@normalless
+@let>=@normalgreater
+@let+=@normalplus
+@let$=@normaldollar}
+
+% Make _ and + \other characters, temporarily.
+% This is canceled by @fixbackslash.
+@otherifyactive
+
+% If a .fmt file is being used, we don't want the `\input texinfo' to show up.
+% That is what \eatinput is for; after that, the `\' should revert to printing
+% a backslash.
+%
+@gdef@eatinput input texinfo{@fixbackslash}
+@global@let\ = @eatinput
+
+% On the other hand, perhaps the file did not have a `\input texinfo'. Then
+% the first `\{ in the file would cause an error. This macro tries to fix
+% that, assuming it is called before the first `\' could plausibly occur.
+% Also back turn on active characters that might appear in the input
+% file name, in case not using a pre-dumped format.
+%
+@gdef@fixbackslash{%
+  @ifx\@eatinput @let\ = @normalbackslash @fi
+  @catcode`+=@active
+  @catcode`@_=@active
+}
+
+% Say @foo, not \foo, in error messages.
+@escapechar = `@@
+
+% These look ok in all fonts, so just make them not special.  
+@catcode`@& = @other
+@catcode`@# = @other
+@catcode`@% = @other
+
+@c Set initial fonts.
+@textfonts
+@rm
+
+
+@c Local variables:
+@c eval: (add-hook 'write-file-hooks 'time-stamp)
+@c page-delimiter: "^\\\\message"
+@c time-stamp-start: "def\\\\texinfoversion{"
+@c time-stamp-format: "%:y-%02m-%02d.%02H"
+@c time-stamp-end: "}"
+@c End:
diff --git a/contrib/gperf/lib/configure b/contrib/gperf/lib/configure
index 791b6f858621..80fa25bb7f9c 100755
--- a/contrib/gperf/lib/configure
+++ b/contrib/gperf/lib/configure
@@ -1,7 +1,7 @@
 #! /bin/sh
 
 # Guess values for system-dependent variables and create Makefiles.
-# Generated automatically using autoconf version 2.12 
+# Generated automatically using autoconf version 2.13 
 # Copyright (C) 1992, 93, 94, 95, 96 Free Software Foundation, Inc.
 #
 # This configure script is free software; the Free Software Foundation
@@ -49,6 +49,7 @@ mandir='${prefix}/man'
 # Initialize some other variables.
 subdirs=
 MFLAGS= MAKEFLAGS=
+SHELL=${CONFIG_SHELL-/bin/sh}
 # Maximum number of lines to put in a shell here document.
 ac_max_here_lines=12
 
@@ -332,7 +333,7 @@ EOF
     verbose=yes ;;
 
   -version | --version | --versio | --versi | --vers)
-    echo "configure generated by autoconf version 2.12"
+    echo "configure generated by autoconf version 2.13"
     exit 0 ;;
 
   -with-* | --with-*)
@@ -502,9 +503,11 @@ ac_ext=c
 # CFLAGS is not in ac_cpp because -g, -O, etc. are not valid cpp options.
 ac_cpp='$CPP $CPPFLAGS'
 ac_compile='${CC-cc} -c $CFLAGS $CPPFLAGS conftest.$ac_ext 1>&5'
-ac_link='${CC-cc} -o conftest $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
+ac_link='${CC-cc} -o conftest${ac_exeext} $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
 cross_compiling=$ac_cv_prog_cc_cross
 
+ac_exeext=
+ac_objext=o
 if (echo "testing\c"; echo 1,2,3) | grep c >/dev/null; then
   # Stardent Vistra SVR4 grep lacks -e, says ghazi@caip.rutgers.edu.
   if (echo -n testing; echo 1,2,3) | sed s/-n/xn/ | grep xn >/dev/null; then
@@ -519,7 +522,7 @@ fi
 
 
 echo $ac_n "checking whether ${MAKE-make} sets \${MAKE}""... $ac_c" 1>&6
-echo "configure:523: checking whether ${MAKE-make} sets \${MAKE}" >&5
+echo "configure:526: checking whether ${MAKE-make} sets \${MAKE}" >&5
 set dummy ${MAKE-make}; ac_make=`echo "$2" | sed 'y%./+-%__p_%'`
 if eval "test \"`echo '$''{'ac_cv_prog_make_${ac_make}_set'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
@@ -548,15 +551,16 @@ fi
 # Extract the first word of "gcc", so it can be a program name with args.
 set dummy gcc; ac_word=$2
 echo $ac_n "checking for $ac_word""... $ac_c" 1>&6
-echo "configure:552: checking for $ac_word" >&5
+echo "configure:555: checking for $ac_word" >&5
 if eval "test \"`echo '$''{'ac_cv_prog_CC'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
 else
   if test -n "$CC"; then
   ac_cv_prog_CC="$CC" # Let the user override the test.
 else
-  IFS="${IFS= 	}"; ac_save_ifs="$IFS"; IFS="${IFS}:"
-  for ac_dir in $PATH; do
+  IFS="${IFS= 	}"; ac_save_ifs="$IFS"; IFS=":"
+  ac_dummy="$PATH"
+  for ac_dir in $ac_dummy; do
     test -z "$ac_dir" && ac_dir=.
     if test -f $ac_dir/$ac_word; then
       ac_cv_prog_CC="gcc"
@@ -577,16 +581,17 @@ if test -z "$CC"; then
   # Extract the first word of "cc", so it can be a program name with args.
 set dummy cc; ac_word=$2
 echo $ac_n "checking for $ac_word""... $ac_c" 1>&6
-echo "configure:581: checking for $ac_word" >&5
+echo "configure:585: checking for $ac_word" >&5
 if eval "test \"`echo '$''{'ac_cv_prog_CC'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
 else
   if test -n "$CC"; then
   ac_cv_prog_CC="$CC" # Let the user override the test.
 else
-  IFS="${IFS= 	}"; ac_save_ifs="$IFS"; IFS="${IFS}:"
+  IFS="${IFS= 	}"; ac_save_ifs="$IFS"; IFS=":"
   ac_prog_rejected=no
-  for ac_dir in $PATH; do
+  ac_dummy="$PATH"
+  for ac_dir in $ac_dummy; do
     test -z "$ac_dir" && ac_dir=.
     if test -f $ac_dir/$ac_word; then
       if test "$ac_dir/$ac_word" = "/usr/ucb/cc"; then
@@ -621,25 +626,61 @@ else
   echo "$ac_t""no" 1>&6
 fi
 
+  if test -z "$CC"; then
+    case "`uname -s`" in
+    *win32* | *WIN32*)
+      # Extract the first word of "cl", so it can be a program name with args.
+set dummy cl; ac_word=$2
+echo $ac_n "checking for $ac_word""... $ac_c" 1>&6
+echo "configure:636: checking for $ac_word" >&5
+if eval "test \"`echo '$''{'ac_cv_prog_CC'+set}'`\" = set"; then
+  echo $ac_n "(cached) $ac_c" 1>&6
+else
+  if test -n "$CC"; then
+  ac_cv_prog_CC="$CC" # Let the user override the test.
+else
+  IFS="${IFS= 	}"; ac_save_ifs="$IFS"; IFS=":"
+  ac_dummy="$PATH"
+  for ac_dir in $ac_dummy; do
+    test -z "$ac_dir" && ac_dir=.
+    if test -f $ac_dir/$ac_word; then
+      ac_cv_prog_CC="cl"
+      break
+    fi
+  done
+  IFS="$ac_save_ifs"
+fi
+fi
+CC="$ac_cv_prog_CC"
+if test -n "$CC"; then
+  echo "$ac_t""$CC" 1>&6
+else
+  echo "$ac_t""no" 1>&6
+fi
+ ;;
+    esac
+  fi
   test -z "$CC" && { echo "configure: error: no acceptable cc found in \$PATH" 1>&2; exit 1; }
 fi
 
 echo $ac_n "checking whether the C compiler ($CC $CFLAGS $LDFLAGS) works""... $ac_c" 1>&6
-echo "configure:629: checking whether the C compiler ($CC $CFLAGS $LDFLAGS) works" >&5
+echo "configure:668: checking whether the C compiler ($CC $CFLAGS $LDFLAGS) works" >&5
 
 ac_ext=c
 # CFLAGS is not in ac_cpp because -g, -O, etc. are not valid cpp options.
 ac_cpp='$CPP $CPPFLAGS'
 ac_compile='${CC-cc} -c $CFLAGS $CPPFLAGS conftest.$ac_ext 1>&5'
-ac_link='${CC-cc} -o conftest $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
+ac_link='${CC-cc} -o conftest${ac_exeext} $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
 cross_compiling=$ac_cv_prog_cc_cross
 
-cat > conftest.$ac_ext <<EOF
-#line 639 "configure"
+cat > conftest.$ac_ext << EOF
+
+#line 679 "configure"
 #include "confdefs.h"
+
 main(){return(0);}
 EOF
-if { (eval echo configure:643: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest; then
+if { (eval echo configure:684: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest${ac_exeext}; then
   ac_cv_prog_cc_works=yes
   # If we can't run a trivial program, we are probably using a cross compiler.
   if (./conftest; exit) 2>/dev/null; then
@@ -653,18 +694,24 @@ else
   ac_cv_prog_cc_works=no
 fi
 rm -fr conftest*
+ac_ext=c
+# CFLAGS is not in ac_cpp because -g, -O, etc. are not valid cpp options.
+ac_cpp='$CPP $CPPFLAGS'
+ac_compile='${CC-cc} -c $CFLAGS $CPPFLAGS conftest.$ac_ext 1>&5'
+ac_link='${CC-cc} -o conftest${ac_exeext} $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
+cross_compiling=$ac_cv_prog_cc_cross
 
 echo "$ac_t""$ac_cv_prog_cc_works" 1>&6
 if test $ac_cv_prog_cc_works = no; then
   { echo "configure: error: installation or configuration problem: C compiler cannot create executables." 1>&2; exit 1; }
 fi
 echo $ac_n "checking whether the C compiler ($CC $CFLAGS $LDFLAGS) is a cross-compiler""... $ac_c" 1>&6
-echo "configure:663: checking whether the C compiler ($CC $CFLAGS $LDFLAGS) is a cross-compiler" >&5
+echo "configure:710: checking whether the C compiler ($CC $CFLAGS $LDFLAGS) is a cross-compiler" >&5
 echo "$ac_t""$ac_cv_prog_cc_cross" 1>&6
 cross_compiling=$ac_cv_prog_cc_cross
 
 echo $ac_n "checking whether we are using GNU C""... $ac_c" 1>&6
-echo "configure:668: checking whether we are using GNU C" >&5
+echo "configure:715: checking whether we are using GNU C" >&5
 if eval "test \"`echo '$''{'ac_cv_prog_gcc'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
 else
@@ -673,7 +720,7 @@ else
   yes;
 #endif
 EOF
-if { ac_try='${CC-cc} -E conftest.c'; { (eval echo configure:677: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }; } | egrep yes >/dev/null 2>&1; then
+if { ac_try='${CC-cc} -E conftest.c'; { (eval echo configure:724: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }; } | egrep yes >/dev/null 2>&1; then
   ac_cv_prog_gcc=yes
 else
   ac_cv_prog_gcc=no
@@ -684,11 +731,15 @@ echo "$ac_t""$ac_cv_prog_gcc" 1>&6
 
 if test $ac_cv_prog_gcc = yes; then
   GCC=yes
-  ac_test_CFLAGS="${CFLAGS+set}"
-  ac_save_CFLAGS="$CFLAGS"
-  CFLAGS=
-  echo $ac_n "checking whether ${CC-cc} accepts -g""... $ac_c" 1>&6
-echo "configure:692: checking whether ${CC-cc} accepts -g" >&5
+else
+  GCC=
+fi
+
+ac_test_CFLAGS="${CFLAGS+set}"
+ac_save_CFLAGS="$CFLAGS"
+CFLAGS=
+echo $ac_n "checking whether ${CC-cc} accepts -g""... $ac_c" 1>&6
+echo "configure:743: checking whether ${CC-cc} accepts -g" >&5
 if eval "test \"`echo '$''{'ac_cv_prog_cc_g'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
 else
@@ -703,20 +754,24 @@ rm -f conftest*
 fi
 
 echo "$ac_t""$ac_cv_prog_cc_g" 1>&6
-  if test "$ac_test_CFLAGS" = set; then
-    CFLAGS="$ac_save_CFLAGS"
-  elif test $ac_cv_prog_cc_g = yes; then
+if test "$ac_test_CFLAGS" = set; then
+  CFLAGS="$ac_save_CFLAGS"
+elif test $ac_cv_prog_cc_g = yes; then
+  if test "$GCC" = yes; then
     CFLAGS="-g -O2"
   else
-    CFLAGS="-O2"
+    CFLAGS="-g"
   fi
 else
-  GCC=
-  test "${CFLAGS+set}" = set || CFLAGS="-g"
+  if test "$GCC" = yes; then
+    CFLAGS="-O2"
+  else
+    CFLAGS=
+  fi
 fi
 
                       echo $ac_n "checking how to run the C preprocessor""... $ac_c" 1>&6
-echo "configure:720: checking how to run the C preprocessor" >&5
+echo "configure:775: checking how to run the C preprocessor" >&5
 # On Suns, sometimes $CPP names a directory.
 if test -n "$CPP" && test -d "$CPP"; then
   CPP=
@@ -731,14 +786,14 @@ else
   # On the NeXT, cc -E runs the code through the compiler's parser,
   # not just through cpp.
   cat > conftest.$ac_ext <<EOF
-#line 735 "configure"
+#line 790 "configure"
 #include "confdefs.h"
 #include <assert.h>
 Syntax Error
 EOF
 ac_try="$ac_cpp conftest.$ac_ext >/dev/null 2>conftest.out"
-{ (eval echo configure:741: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }
-ac_err=`grep -v '^ *+' conftest.out`
+{ (eval echo configure:796: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }
+ac_err=`grep -v '^ *+' conftest.out | grep -v "^conftest.${ac_ext}\$"`
 if test -z "$ac_err"; then
   :
 else
@@ -748,14 +803,31 @@ else
   rm -rf conftest*
   CPP="${CC-cc} -E -traditional-cpp"
   cat > conftest.$ac_ext <<EOF
-#line 752 "configure"
+#line 807 "configure"
 #include "confdefs.h"
 #include <assert.h>
 Syntax Error
 EOF
 ac_try="$ac_cpp conftest.$ac_ext >/dev/null 2>conftest.out"
-{ (eval echo configure:758: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }
-ac_err=`grep -v '^ *+' conftest.out`
+{ (eval echo configure:813: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }
+ac_err=`grep -v '^ *+' conftest.out | grep -v "^conftest.${ac_ext}\$"`
+if test -z "$ac_err"; then
+  :
+else
+  echo "$ac_err" >&5
+  echo "configure: failed program was:" >&5
+  cat conftest.$ac_ext >&5
+  rm -rf conftest*
+  CPP="${CC-cc} -nologo -E"
+  cat > conftest.$ac_ext <<EOF
+#line 824 "configure"
+#include "confdefs.h"
+#include <assert.h>
+Syntax Error
+EOF
+ac_try="$ac_cpp conftest.$ac_ext >/dev/null 2>conftest.out"
+{ (eval echo configure:830: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }
+ac_err=`grep -v '^ *+' conftest.out | grep -v "^conftest.${ac_ext}\$"`
 if test -z "$ac_err"; then
   :
 else
@@ -768,6 +840,8 @@ fi
 rm -f conftest*
 fi
 rm -f conftest*
+fi
+rm -f conftest*
   ac_cv_prog_CPP="$CPP"
 fi
   CPP="$ac_cv_prog_CPP"
@@ -776,20 +850,21 @@ else
 fi
 echo "$ac_t""$CPP" 1>&6
 
-                      for ac_prog in $CCC c++ g++ gcc CC cxx cc++
+                      for ac_prog in $CCC c++ g++ gcc CC cxx cc++ cl
 do
 # Extract the first word of "$ac_prog", so it can be a program name with args.
 set dummy $ac_prog; ac_word=$2
 echo $ac_n "checking for $ac_word""... $ac_c" 1>&6
-echo "configure:785: checking for $ac_word" >&5
+echo "configure:859: checking for $ac_word" >&5
 if eval "test \"`echo '$''{'ac_cv_prog_CXX'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
 else
   if test -n "$CXX"; then
   ac_cv_prog_CXX="$CXX" # Let the user override the test.
 else
-  IFS="${IFS= 	}"; ac_save_ifs="$IFS"; IFS="${IFS}:"
-  for ac_dir in $PATH; do
+  IFS="${IFS= 	}"; ac_save_ifs="$IFS"; IFS=":"
+  ac_dummy="$PATH"
+  for ac_dir in $ac_dummy; do
     test -z "$ac_dir" && ac_dir=.
     if test -f $ac_dir/$ac_word; then
       ac_cv_prog_CXX="$ac_prog"
@@ -812,21 +887,23 @@ test -n "$CXX" || CXX="gcc"
 
 
 echo $ac_n "checking whether the C++ compiler ($CXX $CXXFLAGS $LDFLAGS) works""... $ac_c" 1>&6
-echo "configure:816: checking whether the C++ compiler ($CXX $CXXFLAGS $LDFLAGS) works" >&5
+echo "configure:891: checking whether the C++ compiler ($CXX $CXXFLAGS $LDFLAGS) works" >&5
 
 ac_ext=C
 # CXXFLAGS is not in ac_cpp because -g, -O, etc. are not valid cpp options.
 ac_cpp='$CXXCPP $CPPFLAGS'
 ac_compile='${CXX-g++} -c $CXXFLAGS $CPPFLAGS conftest.$ac_ext 1>&5'
-ac_link='${CXX-g++} -o conftest $CXXFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
+ac_link='${CXX-g++} -o conftest${ac_exeext} $CXXFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
 cross_compiling=$ac_cv_prog_cxx_cross
 
-cat > conftest.$ac_ext <<EOF
-#line 826 "configure"
+cat > conftest.$ac_ext << EOF
+
+#line 902 "configure"
 #include "confdefs.h"
-main(){return(0);}
+
+int main(){return(0);}
 EOF
-if { (eval echo configure:830: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest; then
+if { (eval echo configure:907: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest${ac_exeext}; then
   ac_cv_prog_cxx_works=yes
   # If we can't run a trivial program, we are probably using a cross compiler.
   if (./conftest; exit) 2>/dev/null; then
@@ -844,7 +921,7 @@ ac_ext=c
 # CFLAGS is not in ac_cpp because -g, -O, etc. are not valid cpp options.
 ac_cpp='$CPP $CPPFLAGS'
 ac_compile='${CC-cc} -c $CFLAGS $CPPFLAGS conftest.$ac_ext 1>&5'
-ac_link='${CC-cc} -o conftest $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
+ac_link='${CC-cc} -o conftest${ac_exeext} $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
 cross_compiling=$ac_cv_prog_cc_cross
 
 echo "$ac_t""$ac_cv_prog_cxx_works" 1>&6
@@ -852,12 +929,12 @@ if test $ac_cv_prog_cxx_works = no; then
   { echo "configure: error: installation or configuration problem: C++ compiler cannot create executables." 1>&2; exit 1; }
 fi
 echo $ac_n "checking whether the C++ compiler ($CXX $CXXFLAGS $LDFLAGS) is a cross-compiler""... $ac_c" 1>&6
-echo "configure:856: checking whether the C++ compiler ($CXX $CXXFLAGS $LDFLAGS) is a cross-compiler" >&5
+echo "configure:933: checking whether the C++ compiler ($CXX $CXXFLAGS $LDFLAGS) is a cross-compiler" >&5
 echo "$ac_t""$ac_cv_prog_cxx_cross" 1>&6
 cross_compiling=$ac_cv_prog_cxx_cross
 
 echo $ac_n "checking whether we are using GNU C++""... $ac_c" 1>&6
-echo "configure:861: checking whether we are using GNU C++" >&5
+echo "configure:938: checking whether we are using GNU C++" >&5
 if eval "test \"`echo '$''{'ac_cv_prog_gxx'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
 else
@@ -866,7 +943,7 @@ else
   yes;
 #endif
 EOF
-if { ac_try='${CXX-g++} -E conftest.C'; { (eval echo configure:870: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }; } | egrep yes >/dev/null 2>&1; then
+if { ac_try='${CXX-g++} -E conftest.C'; { (eval echo configure:947: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }; } | egrep yes >/dev/null 2>&1; then
   ac_cv_prog_gxx=yes
 else
   ac_cv_prog_gxx=no
@@ -877,11 +954,15 @@ echo "$ac_t""$ac_cv_prog_gxx" 1>&6
 
 if test $ac_cv_prog_gxx = yes; then
   GXX=yes
-  ac_test_CXXFLAGS="${CXXFLAGS+set}"
-  ac_save_CXXFLAGS="$CXXFLAGS"
-  CXXFLAGS=
-  echo $ac_n "checking whether ${CXX-g++} accepts -g""... $ac_c" 1>&6
-echo "configure:885: checking whether ${CXX-g++} accepts -g" >&5
+else
+  GXX=
+fi
+
+ac_test_CXXFLAGS="${CXXFLAGS+set}"
+ac_save_CXXFLAGS="$CXXFLAGS"
+CXXFLAGS=
+echo $ac_n "checking whether ${CXX-g++} accepts -g""... $ac_c" 1>&6
+echo "configure:966: checking whether ${CXX-g++} accepts -g" >&5
 if eval "test \"`echo '$''{'ac_cv_prog_cxx_g'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
 else
@@ -896,20 +977,24 @@ rm -f conftest*
 fi
 
 echo "$ac_t""$ac_cv_prog_cxx_g" 1>&6
-  if test "$ac_test_CXXFLAGS" = set; then
-    CXXFLAGS="$ac_save_CXXFLAGS"
-  elif test $ac_cv_prog_cxx_g = yes; then
+if test "$ac_test_CXXFLAGS" = set; then
+  CXXFLAGS="$ac_save_CXXFLAGS"
+elif test $ac_cv_prog_cxx_g = yes; then
+  if test "$GXX" = yes; then
     CXXFLAGS="-g -O2"
   else
-    CXXFLAGS="-O2"
+    CXXFLAGS="-g"
   fi
 else
-  GXX=
-  test "${CXXFLAGS+set}" = set || CXXFLAGS="-g"
+  if test "$GXX" = yes; then
+    CXXFLAGS="-O2"
+  else
+    CXXFLAGS=
+  fi
 fi
 
                       echo $ac_n "checking how to run the C++ preprocessor""... $ac_c" 1>&6
-echo "configure:913: checking how to run the C++ preprocessor" >&5
+echo "configure:998: checking how to run the C++ preprocessor" >&5
 if test -z "$CXXCPP"; then
 if eval "test \"`echo '$''{'ac_cv_prog_CXXCPP'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
@@ -918,17 +1003,17 @@ else
 # CXXFLAGS is not in ac_cpp because -g, -O, etc. are not valid cpp options.
 ac_cpp='$CXXCPP $CPPFLAGS'
 ac_compile='${CXX-g++} -c $CXXFLAGS $CPPFLAGS conftest.$ac_ext 1>&5'
-ac_link='${CXX-g++} -o conftest $CXXFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
+ac_link='${CXX-g++} -o conftest${ac_exeext} $CXXFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
 cross_compiling=$ac_cv_prog_cxx_cross
   CXXCPP="${CXX-g++} -E"
   cat > conftest.$ac_ext <<EOF
-#line 926 "configure"
+#line 1011 "configure"
 #include "confdefs.h"
 #include <stdlib.h>
 EOF
 ac_try="$ac_cpp conftest.$ac_ext >/dev/null 2>conftest.out"
-{ (eval echo configure:931: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }
-ac_err=`grep -v '^ *+' conftest.out`
+{ (eval echo configure:1016: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }
+ac_err=`grep -v '^ *+' conftest.out | grep -v "^conftest.${ac_ext}\$"`
 if test -z "$ac_err"; then
   :
 else
@@ -944,7 +1029,7 @@ ac_ext=c
 # CFLAGS is not in ac_cpp because -g, -O, etc. are not valid cpp options.
 ac_cpp='$CPP $CPPFLAGS'
 ac_compile='${CC-cc} -c $CFLAGS $CPPFLAGS conftest.$ac_ext 1>&5'
-ac_link='${CC-cc} -o conftest $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
+ac_link='${CC-cc} -o conftest${ac_exeext} $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
 cross_compiling=$ac_cv_prog_cc_cross
 fi
 fi
@@ -954,15 +1039,16 @@ echo "$ac_t""$CXXCPP" 1>&6
                       # Extract the first word of "ranlib", so it can be a program name with args.
 set dummy ranlib; ac_word=$2
 echo $ac_n "checking for $ac_word""... $ac_c" 1>&6
-echo "configure:958: checking for $ac_word" >&5
+echo "configure:1043: checking for $ac_word" >&5
 if eval "test \"`echo '$''{'ac_cv_prog_RANLIB'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
 else
   if test -n "$RANLIB"; then
   ac_cv_prog_RANLIB="$RANLIB" # Let the user override the test.
 else
-  IFS="${IFS= 	}"; ac_save_ifs="$IFS"; IFS="${IFS}:"
-  for ac_dir in $PATH; do
+  IFS="${IFS= 	}"; ac_save_ifs="$IFS"; IFS=":"
+  ac_dummy="$PATH"
+  for ac_dir in $ac_dummy; do
     test -z "$ac_dir" && ac_dir=.
     if test -f $ac_dir/$ac_word; then
       ac_cv_prog_RANLIB="ranlib"
@@ -991,7 +1077,7 @@ fi
 # SVR4 /usr/ucb/install, which tries to use the nonexistent group "staff"
 # ./install, which can be erroneously created by make from ./install.sh.
 echo $ac_n "checking for a BSD compatible install""... $ac_c" 1>&6
-echo "configure:995: checking for a BSD compatible install" >&5
+echo "configure:1081: checking for a BSD compatible install" >&5
 if test -z "$INSTALL"; then
 if eval "test \"`echo '$''{'cl_cv_path_install'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
@@ -1016,7 +1102,7 @@ else
 	      # AIX installbsd doesn't work without option "-g".
 	      :
 	    else
-	      ac_cv_path_install="$ac_dir/$ac_prog -c"
+	      cl_cv_path_install="$ac_dir/$ac_prog -c"
 	      break 2
 	    fi
 	  fi
@@ -1066,7 +1152,7 @@ EOF
 # Ultrix sh set writes to stderr and can't be redirected directly,
 # and sets the high bit in the cache file unless we assign to the vars.
 (set) 2>&1 |
-  case `(ac_space=' '; set) 2>&1` in
+  case `(ac_space=' '; set | grep ac_space) 2>&1` in
   *ac_space=\ *)
     # `set' does not quote correctly, so add quotes (double-quote substitution
     # turns \\\\ into \\, and sed turns \\ into \).
@@ -1145,7 +1231,7 @@ do
     echo "running \${CONFIG_SHELL-/bin/sh} $0 $ac_configure_args --no-create --no-recursion"
     exec \${CONFIG_SHELL-/bin/sh} $0 $ac_configure_args --no-create --no-recursion ;;
   -version | --version | --versio | --versi | --vers | --ver | --ve | --v)
-    echo "$CONFIG_STATUS generated by autoconf version 2.12"
+    echo "$CONFIG_STATUS generated by autoconf version 2.13"
     exit 0 ;;
   -help | --help | --hel | --he | --h)
     echo "\$ac_cs_usage"; exit 0 ;;
@@ -1164,9 +1250,11 @@ sed 's/%@/@@/; s/@%/@@/; s/%g\$/@g/; /@g\$/s/[\\\\&%]/\\\\&/g;
  s/@@/%@/; s/@@/@%/; s/@g\$/%g/' > conftest.subs <<\\CEOF
 $ac_vpsub
 $extrasub
+s%@SHELL@%$SHELL%g
 s%@CFLAGS@%$CFLAGS%g
 s%@CPPFLAGS@%$CPPFLAGS%g
 s%@CXXFLAGS@%$CXXFLAGS%g
+s%@FFLAGS@%$FFLAGS%g
 s%@DEFS@%$DEFS%g
 s%@LDFLAGS@%$LDFLAGS%g
 s%@LIBS@%$LIBS%g
diff --git a/contrib/gperf/lib/getopt.h b/contrib/gperf/lib/getopt.h
index 5be631934f17..a9d75f22c815 100644
--- a/contrib/gperf/lib/getopt.h
+++ b/contrib/gperf/lib/getopt.h
@@ -1,5 +1,5 @@
 /* Declarations for getopt.
-   Copyright (C) 1989,90,91,92,93,94,96,97 Free Software Foundation, Inc.
+   Copyright (C) 1989-1998, 2000 Free Software Foundation, Inc.
    This file is part of the GNU C Library.
 
    The GNU C Library is free software; you can redistribute it and/or
@@ -78,7 +78,7 @@ extern int optopt;
 
 struct option
 {
-#if defined (__STDC__) && __STDC__
+#if (defined (__STDC__) && __STDC__) || defined (__cplusplus)
   const char *name;
 #else
   char *name;
diff --git a/contrib/gperf/lib/getopt.h.patch b/contrib/gperf/lib/getopt.h.patch
index 7a7c2f152ecb..8094b1a65d40 100644
--- a/contrib/gperf/lib/getopt.h.patch
+++ b/contrib/gperf/lib/getopt.h.patch
@@ -4,7 +4,37 @@ maintainer has been informed of these patches.
 
 diff -c3 getopt.h.orig getopt.h
 *** getopt.h.orig	Sat Jun 21 03:01:53 1997
---- getopt.h	Thu Apr 16 00:17:42 1998
+--- getopt.h	Mon Aug 28 12:36:27 2000
+***************
+*** 1,5 ****
+  /* Declarations for getopt.
+!    Copyright (C) 1989,90,91,92,93,94,96,97 Free Software Foundation, Inc.
+     This file is part of the GNU C Library.
+  
+     The GNU C Library is free software; you can redistribute it and/or
+--- 1,5 ----
+  /* Declarations for getopt.
+!    Copyright (C) 1989-1998, 2000 Free Software Foundation, Inc.
+     This file is part of the GNU C Library.
+  
+     The GNU C Library is free software; you can redistribute it and/or
+***************
+*** 78,84 ****
+  
+  struct option
+  {
+! #if defined (__STDC__) && __STDC__
+    const char *name;
+  #else
+    char *name;
+--- 78,84 ----
+  
+  struct option
+  {
+! #if (defined (__STDC__) && __STDC__) || defined (__cplusplus)
+    const char *name;
+  #else
+    char *name;
 ***************
 *** 96,102 ****
   #define required_argument	1
diff --git a/contrib/gperf/lib/hash.cc b/contrib/gperf/lib/hash.cc
index 72a287f59ef3..b5bb4ad66836 100644
--- a/contrib/gperf/lib/hash.cc
+++ b/contrib/gperf/lib/hash.cc
@@ -1,36 +1,27 @@
 /* 
-Copyright (C) 1990 Free Software Foundation
+Copyright (C) 1990, 2000 Free Software Foundation
     written by Doug Lea (dl@rocky.oswego.edu)
-
-This file is part of the GNU C++ Library.  This library is free
-software; you can redistribute it and/or modify it under the terms of
-the GNU Library General Public License as published by the Free
-Software Foundation; either version 2 of the License, or (at your
-option) any later version.  This library is distributed in the hope
-that it will be useful, but WITHOUT ANY WARRANTY; without even the
-implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR
-PURPOSE.  See the GNU Library General Public License for more details.
-You should have received a copy of the GNU Library General Public
-License along with this library; if not, write to the Free Software
-Foundation, 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
 */
 
 #include <hash.h>
 
 /*
- some useful hash functions
+ Some useful hash function.
+ It's not a particularly good hash function (<< 5 would be better than << 4),
+ but people believe in it because it comes from Dragon book.
 */
 
-unsigned int hashpjw (const char* x) // From Dragon book, p436
+unsigned int
+hashpjw (const char *x, unsigned int len) // From Dragon book, p436
 {
   unsigned int h = 0;
   unsigned int g;
 
-  while (*x != 0)
-  {
-    h = (h << 4) + (unsigned char) *x++;
-    if ((g = h & 0xf0000000) != 0)
-      h = (h ^ (g >> 24)) ^ g;
-  }
+  for (; len > 0; len--)
+    {
+      h = (h << 4) + (unsigned char) *x++;
+      if ((g = h & 0xf0000000) != 0)
+        h = (h ^ (g >> 24)) ^ g;
+    }
   return h;
 }
diff --git a/contrib/gperf/lib/hash.h b/contrib/gperf/lib/hash.h
index a5f324b8f925..5dbc92b7a186 100644
--- a/contrib/gperf/lib/hash.h
+++ b/contrib/gperf/lib/hash.h
@@ -1,27 +1,15 @@
 // This may look like C code, but it is really -*- C++ -*-
 
 /* 
-Copyright (C) 1988, 1992 Free Software Foundation
+Copyright (C) 1988, 1992, 2000 Free Software Foundation
     written by Doug Lea (dl@rocky.oswego.edu)
-
-This file is part of the GNU C++ Library.  This library is free
-software; you can redistribute it and/or modify it under the terms of
-the GNU Library General Public License as published by the Free
-Software Foundation; either version 2 of the License, or (at your
-option) any later version.  This library is distributed in the hope
-that it will be useful, but WITHOUT ANY WARRANTY; without even the
-implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR
-PURPOSE.  See the GNU Library General Public License for more details.
-You should have received a copy of the GNU Library General Public
-License along with this library; if not, write to the Free Software
-Foundation, 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
 */
 
 #ifndef _hash_h
 #define _hash_h 1
 
-/* a hash function for null-terminated char* strings using the
+/* a hash function for char[] arrays using the
    method described in Aho, Sethi, & Ullman, p 436. */
-extern unsigned int hashpjw (const char*);
+extern unsigned int hashpjw (const char *string, unsigned int len);
 
 #endif
diff --git a/contrib/gperf/mkinstalldirs b/contrib/gperf/mkinstalldirs
new file mode 100755
index 000000000000..4f58503ea4e0
--- /dev/null
+++ b/contrib/gperf/mkinstalldirs
@@ -0,0 +1,40 @@
+#! /bin/sh
+# mkinstalldirs --- make directory hierarchy
+# Author: Noah Friedman <friedman@prep.ai.mit.edu>
+# Created: 1993-05-16
+# Public domain
+
+# $Id: mkinstalldirs,v 1.13 1999/01/05 03:18:55 bje Exp $
+
+errstatus=0
+
+for file
+do
+   set fnord `echo ":$file" | sed -ne 's/^:\//#/;s/^://;s/\// /g;s/^#/\//;p'`
+   shift
+
+   pathcomp=
+   for d
+   do
+     pathcomp="$pathcomp$d"
+     case "$pathcomp" in
+       -* ) pathcomp=./$pathcomp ;;
+     esac
+
+     if test ! -d "$pathcomp"; then
+        echo "mkdir $pathcomp"
+
+        mkdir "$pathcomp" || lasterr=$?
+
+        if test ! -d "$pathcomp"; then
+  	  errstatus=$lasterr
+        fi
+     fi
+
+     pathcomp="$pathcomp/"
+   done
+done
+
+exit $errstatus
+
+# mkinstalldirs ends here
diff --git a/contrib/gperf/src/Makefile.in b/contrib/gperf/src/Makefile.in
index ec1e682bd22b..60f73c735a34 100644
--- a/contrib/gperf/src/Makefile.in
+++ b/contrib/gperf/src/Makefile.in
@@ -1,6 +1,6 @@
 # Makefile for gperf/src
 
-# Copyright (C) 1989, 1992, 1993, 1998 Free Software Foundation, Inc.
+# Copyright (C) 1989, 1992, 1993, 1998, 2000 Free Software Foundation, Inc.
 # written by Douglas C. Schmidt (schmidt@ics.uci.edu)
 #
 # This file is part of GNU GPERF.
@@ -39,6 +39,8 @@ CPP = @CPP@
 CXX = @CXX@
 CXXFLAGS = @CXXFLAGS@
 CXXCPP = @CXXCPP@
+# Both C and C++ compiler
+LDFLAGS = @LDFLAGS@
 # Other
 MV = mv
 LN = ln
@@ -49,6 +51,7 @@ RM = rm -f
 INSTALL = @INSTALL@
 INSTALL_PROGRAM = @INSTALL_PROGRAM@
 INSTALL_DATA = @INSTALL_DATA@
+MKINSTALLDIRS = $(SHELL) $(srcdir)/../mkinstalldirs
 
 #### End of system configuration section. ####
 
@@ -119,14 +122,14 @@ version.o : version.cc $(VERSION_H)
 	$(CXX) $(CXXFLAGS) $(CPPFLAGS) -c $(srcdir)/version.cc
 
 install : all force
-	if [ ! -d $(bindir) ] ; then mkdir $(bindir) ; fi
-	$(INSTALL_PROGRAM) $(TARGETPROG) $(bindir)/$(TARGETPROG)
+	$(MKINSTALLDIRS) $(DESTDIR)$(bindir)
+	$(INSTALL_PROGRAM) $(TARGETPROG) $(DESTDIR)$(bindir)/$(TARGETPROG)
 
 installdirs : force
-	if [ ! -d $(bindir) ] ; then mkdir $(bindir) ; fi
+	$(MKINSTALLDIRS) $(DESTDIR)$(bindir)
 
 uninstall : force
-	$(RM) $(bindir)/$(TARGETPROG)
+	$(RM) $(DESTDIR)$(bindir)/$(TARGETPROG)
 
 check : all
 
diff --git a/contrib/gperf/src/configure b/contrib/gperf/src/configure
index d3d109112d62..edd1fd7f0520 100755
--- a/contrib/gperf/src/configure
+++ b/contrib/gperf/src/configure
@@ -1,7 +1,7 @@
 #! /bin/sh
 
 # Guess values for system-dependent variables and create Makefiles.
-# Generated automatically using autoconf version 2.12 
+# Generated automatically using autoconf version 2.13 
 # Copyright (C) 1992, 93, 94, 95, 96 Free Software Foundation, Inc.
 #
 # This configure script is free software; the Free Software Foundation
@@ -49,6 +49,7 @@ mandir='${prefix}/man'
 # Initialize some other variables.
 subdirs=
 MFLAGS= MAKEFLAGS=
+SHELL=${CONFIG_SHELL-/bin/sh}
 # Maximum number of lines to put in a shell here document.
 ac_max_here_lines=12
 
@@ -332,7 +333,7 @@ EOF
     verbose=yes ;;
 
   -version | --version | --versio | --versi | --vers)
-    echo "configure generated by autoconf version 2.12"
+    echo "configure generated by autoconf version 2.13"
     exit 0 ;;
 
   -with-* | --with-*)
@@ -502,9 +503,11 @@ ac_ext=c
 # CFLAGS is not in ac_cpp because -g, -O, etc. are not valid cpp options.
 ac_cpp='$CPP $CPPFLAGS'
 ac_compile='${CC-cc} -c $CFLAGS $CPPFLAGS conftest.$ac_ext 1>&5'
-ac_link='${CC-cc} -o conftest $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
+ac_link='${CC-cc} -o conftest${ac_exeext} $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
 cross_compiling=$ac_cv_prog_cc_cross
 
+ac_exeext=
+ac_objext=o
 if (echo "testing\c"; echo 1,2,3) | grep c >/dev/null; then
   # Stardent Vistra SVR4 grep lacks -e, says ghazi@caip.rutgers.edu.
   if (echo -n testing; echo 1,2,3) | sed s/-n/xn/ | grep xn >/dev/null; then
@@ -520,7 +523,7 @@ fi
 
 
 echo $ac_n "checking whether ${MAKE-make} sets \${MAKE}""... $ac_c" 1>&6
-echo "configure:524: checking whether ${MAKE-make} sets \${MAKE}" >&5
+echo "configure:527: checking whether ${MAKE-make} sets \${MAKE}" >&5
 set dummy ${MAKE-make}; ac_make=`echo "$2" | sed 'y%./+-%__p_%'`
 if eval "test \"`echo '$''{'ac_cv_prog_make_${ac_make}_set'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
@@ -549,15 +552,16 @@ fi
 # Extract the first word of "gcc", so it can be a program name with args.
 set dummy gcc; ac_word=$2
 echo $ac_n "checking for $ac_word""... $ac_c" 1>&6
-echo "configure:553: checking for $ac_word" >&5
+echo "configure:556: checking for $ac_word" >&5
 if eval "test \"`echo '$''{'ac_cv_prog_CC'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
 else
   if test -n "$CC"; then
   ac_cv_prog_CC="$CC" # Let the user override the test.
 else
-  IFS="${IFS= 	}"; ac_save_ifs="$IFS"; IFS="${IFS}:"
-  for ac_dir in $PATH; do
+  IFS="${IFS= 	}"; ac_save_ifs="$IFS"; IFS=":"
+  ac_dummy="$PATH"
+  for ac_dir in $ac_dummy; do
     test -z "$ac_dir" && ac_dir=.
     if test -f $ac_dir/$ac_word; then
       ac_cv_prog_CC="gcc"
@@ -578,16 +582,17 @@ if test -z "$CC"; then
   # Extract the first word of "cc", so it can be a program name with args.
 set dummy cc; ac_word=$2
 echo $ac_n "checking for $ac_word""... $ac_c" 1>&6
-echo "configure:582: checking for $ac_word" >&5
+echo "configure:586: checking for $ac_word" >&5
 if eval "test \"`echo '$''{'ac_cv_prog_CC'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
 else
   if test -n "$CC"; then
   ac_cv_prog_CC="$CC" # Let the user override the test.
 else
-  IFS="${IFS= 	}"; ac_save_ifs="$IFS"; IFS="${IFS}:"
+  IFS="${IFS= 	}"; ac_save_ifs="$IFS"; IFS=":"
   ac_prog_rejected=no
-  for ac_dir in $PATH; do
+  ac_dummy="$PATH"
+  for ac_dir in $ac_dummy; do
     test -z "$ac_dir" && ac_dir=.
     if test -f $ac_dir/$ac_word; then
       if test "$ac_dir/$ac_word" = "/usr/ucb/cc"; then
@@ -622,25 +627,61 @@ else
   echo "$ac_t""no" 1>&6
 fi
 
+  if test -z "$CC"; then
+    case "`uname -s`" in
+    *win32* | *WIN32*)
+      # Extract the first word of "cl", so it can be a program name with args.
+set dummy cl; ac_word=$2
+echo $ac_n "checking for $ac_word""... $ac_c" 1>&6
+echo "configure:637: checking for $ac_word" >&5
+if eval "test \"`echo '$''{'ac_cv_prog_CC'+set}'`\" = set"; then
+  echo $ac_n "(cached) $ac_c" 1>&6
+else
+  if test -n "$CC"; then
+  ac_cv_prog_CC="$CC" # Let the user override the test.
+else
+  IFS="${IFS= 	}"; ac_save_ifs="$IFS"; IFS=":"
+  ac_dummy="$PATH"
+  for ac_dir in $ac_dummy; do
+    test -z "$ac_dir" && ac_dir=.
+    if test -f $ac_dir/$ac_word; then
+      ac_cv_prog_CC="cl"
+      break
+    fi
+  done
+  IFS="$ac_save_ifs"
+fi
+fi
+CC="$ac_cv_prog_CC"
+if test -n "$CC"; then
+  echo "$ac_t""$CC" 1>&6
+else
+  echo "$ac_t""no" 1>&6
+fi
+ ;;
+    esac
+  fi
   test -z "$CC" && { echo "configure: error: no acceptable cc found in \$PATH" 1>&2; exit 1; }
 fi
 
 echo $ac_n "checking whether the C compiler ($CC $CFLAGS $LDFLAGS) works""... $ac_c" 1>&6
-echo "configure:630: checking whether the C compiler ($CC $CFLAGS $LDFLAGS) works" >&5
+echo "configure:669: checking whether the C compiler ($CC $CFLAGS $LDFLAGS) works" >&5
 
 ac_ext=c
 # CFLAGS is not in ac_cpp because -g, -O, etc. are not valid cpp options.
 ac_cpp='$CPP $CPPFLAGS'
 ac_compile='${CC-cc} -c $CFLAGS $CPPFLAGS conftest.$ac_ext 1>&5'
-ac_link='${CC-cc} -o conftest $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
+ac_link='${CC-cc} -o conftest${ac_exeext} $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
 cross_compiling=$ac_cv_prog_cc_cross
 
-cat > conftest.$ac_ext <<EOF
-#line 640 "configure"
+cat > conftest.$ac_ext << EOF
+
+#line 680 "configure"
 #include "confdefs.h"
+
 main(){return(0);}
 EOF
-if { (eval echo configure:644: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest; then
+if { (eval echo configure:685: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest${ac_exeext}; then
   ac_cv_prog_cc_works=yes
   # If we can't run a trivial program, we are probably using a cross compiler.
   if (./conftest; exit) 2>/dev/null; then
@@ -654,18 +695,24 @@ else
   ac_cv_prog_cc_works=no
 fi
 rm -fr conftest*
+ac_ext=c
+# CFLAGS is not in ac_cpp because -g, -O, etc. are not valid cpp options.
+ac_cpp='$CPP $CPPFLAGS'
+ac_compile='${CC-cc} -c $CFLAGS $CPPFLAGS conftest.$ac_ext 1>&5'
+ac_link='${CC-cc} -o conftest${ac_exeext} $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
+cross_compiling=$ac_cv_prog_cc_cross
 
 echo "$ac_t""$ac_cv_prog_cc_works" 1>&6
 if test $ac_cv_prog_cc_works = no; then
   { echo "configure: error: installation or configuration problem: C compiler cannot create executables." 1>&2; exit 1; }
 fi
 echo $ac_n "checking whether the C compiler ($CC $CFLAGS $LDFLAGS) is a cross-compiler""... $ac_c" 1>&6
-echo "configure:664: checking whether the C compiler ($CC $CFLAGS $LDFLAGS) is a cross-compiler" >&5
+echo "configure:711: checking whether the C compiler ($CC $CFLAGS $LDFLAGS) is a cross-compiler" >&5
 echo "$ac_t""$ac_cv_prog_cc_cross" 1>&6
 cross_compiling=$ac_cv_prog_cc_cross
 
 echo $ac_n "checking whether we are using GNU C""... $ac_c" 1>&6
-echo "configure:669: checking whether we are using GNU C" >&5
+echo "configure:716: checking whether we are using GNU C" >&5
 if eval "test \"`echo '$''{'ac_cv_prog_gcc'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
 else
@@ -674,7 +721,7 @@ else
   yes;
 #endif
 EOF
-if { ac_try='${CC-cc} -E conftest.c'; { (eval echo configure:678: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }; } | egrep yes >/dev/null 2>&1; then
+if { ac_try='${CC-cc} -E conftest.c'; { (eval echo configure:725: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }; } | egrep yes >/dev/null 2>&1; then
   ac_cv_prog_gcc=yes
 else
   ac_cv_prog_gcc=no
@@ -685,11 +732,15 @@ echo "$ac_t""$ac_cv_prog_gcc" 1>&6
 
 if test $ac_cv_prog_gcc = yes; then
   GCC=yes
-  ac_test_CFLAGS="${CFLAGS+set}"
-  ac_save_CFLAGS="$CFLAGS"
-  CFLAGS=
-  echo $ac_n "checking whether ${CC-cc} accepts -g""... $ac_c" 1>&6
-echo "configure:693: checking whether ${CC-cc} accepts -g" >&5
+else
+  GCC=
+fi
+
+ac_test_CFLAGS="${CFLAGS+set}"
+ac_save_CFLAGS="$CFLAGS"
+CFLAGS=
+echo $ac_n "checking whether ${CC-cc} accepts -g""... $ac_c" 1>&6
+echo "configure:744: checking whether ${CC-cc} accepts -g" >&5
 if eval "test \"`echo '$''{'ac_cv_prog_cc_g'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
 else
@@ -704,20 +755,24 @@ rm -f conftest*
 fi
 
 echo "$ac_t""$ac_cv_prog_cc_g" 1>&6
-  if test "$ac_test_CFLAGS" = set; then
-    CFLAGS="$ac_save_CFLAGS"
-  elif test $ac_cv_prog_cc_g = yes; then
+if test "$ac_test_CFLAGS" = set; then
+  CFLAGS="$ac_save_CFLAGS"
+elif test $ac_cv_prog_cc_g = yes; then
+  if test "$GCC" = yes; then
     CFLAGS="-g -O2"
   else
-    CFLAGS="-O2"
+    CFLAGS="-g"
   fi
 else
-  GCC=
-  test "${CFLAGS+set}" = set || CFLAGS="-g"
+  if test "$GCC" = yes; then
+    CFLAGS="-O2"
+  else
+    CFLAGS=
+  fi
 fi
 
                       echo $ac_n "checking how to run the C preprocessor""... $ac_c" 1>&6
-echo "configure:721: checking how to run the C preprocessor" >&5
+echo "configure:776: checking how to run the C preprocessor" >&5
 # On Suns, sometimes $CPP names a directory.
 if test -n "$CPP" && test -d "$CPP"; then
   CPP=
@@ -732,14 +787,14 @@ else
   # On the NeXT, cc -E runs the code through the compiler's parser,
   # not just through cpp.
   cat > conftest.$ac_ext <<EOF
-#line 736 "configure"
+#line 791 "configure"
 #include "confdefs.h"
 #include <assert.h>
 Syntax Error
 EOF
 ac_try="$ac_cpp conftest.$ac_ext >/dev/null 2>conftest.out"
-{ (eval echo configure:742: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }
-ac_err=`grep -v '^ *+' conftest.out`
+{ (eval echo configure:797: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }
+ac_err=`grep -v '^ *+' conftest.out | grep -v "^conftest.${ac_ext}\$"`
 if test -z "$ac_err"; then
   :
 else
@@ -749,14 +804,31 @@ else
   rm -rf conftest*
   CPP="${CC-cc} -E -traditional-cpp"
   cat > conftest.$ac_ext <<EOF
-#line 753 "configure"
+#line 808 "configure"
 #include "confdefs.h"
 #include <assert.h>
 Syntax Error
 EOF
 ac_try="$ac_cpp conftest.$ac_ext >/dev/null 2>conftest.out"
-{ (eval echo configure:759: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }
-ac_err=`grep -v '^ *+' conftest.out`
+{ (eval echo configure:814: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }
+ac_err=`grep -v '^ *+' conftest.out | grep -v "^conftest.${ac_ext}\$"`
+if test -z "$ac_err"; then
+  :
+else
+  echo "$ac_err" >&5
+  echo "configure: failed program was:" >&5
+  cat conftest.$ac_ext >&5
+  rm -rf conftest*
+  CPP="${CC-cc} -nologo -E"
+  cat > conftest.$ac_ext <<EOF
+#line 825 "configure"
+#include "confdefs.h"
+#include <assert.h>
+Syntax Error
+EOF
+ac_try="$ac_cpp conftest.$ac_ext >/dev/null 2>conftest.out"
+{ (eval echo configure:831: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }
+ac_err=`grep -v '^ *+' conftest.out | grep -v "^conftest.${ac_ext}\$"`
 if test -z "$ac_err"; then
   :
 else
@@ -769,6 +841,8 @@ fi
 rm -f conftest*
 fi
 rm -f conftest*
+fi
+rm -f conftest*
   ac_cv_prog_CPP="$CPP"
 fi
   CPP="$ac_cv_prog_CPP"
@@ -777,20 +851,21 @@ else
 fi
 echo "$ac_t""$CPP" 1>&6
 
-                      for ac_prog in $CCC c++ g++ gcc CC cxx cc++
+                      for ac_prog in $CCC c++ g++ gcc CC cxx cc++ cl
 do
 # Extract the first word of "$ac_prog", so it can be a program name with args.
 set dummy $ac_prog; ac_word=$2
 echo $ac_n "checking for $ac_word""... $ac_c" 1>&6
-echo "configure:786: checking for $ac_word" >&5
+echo "configure:860: checking for $ac_word" >&5
 if eval "test \"`echo '$''{'ac_cv_prog_CXX'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
 else
   if test -n "$CXX"; then
   ac_cv_prog_CXX="$CXX" # Let the user override the test.
 else
-  IFS="${IFS= 	}"; ac_save_ifs="$IFS"; IFS="${IFS}:"
-  for ac_dir in $PATH; do
+  IFS="${IFS= 	}"; ac_save_ifs="$IFS"; IFS=":"
+  ac_dummy="$PATH"
+  for ac_dir in $ac_dummy; do
     test -z "$ac_dir" && ac_dir=.
     if test -f $ac_dir/$ac_word; then
       ac_cv_prog_CXX="$ac_prog"
@@ -813,21 +888,23 @@ test -n "$CXX" || CXX="gcc"
 
 
 echo $ac_n "checking whether the C++ compiler ($CXX $CXXFLAGS $LDFLAGS) works""... $ac_c" 1>&6
-echo "configure:817: checking whether the C++ compiler ($CXX $CXXFLAGS $LDFLAGS) works" >&5
+echo "configure:892: checking whether the C++ compiler ($CXX $CXXFLAGS $LDFLAGS) works" >&5
 
 ac_ext=C
 # CXXFLAGS is not in ac_cpp because -g, -O, etc. are not valid cpp options.
 ac_cpp='$CXXCPP $CPPFLAGS'
 ac_compile='${CXX-g++} -c $CXXFLAGS $CPPFLAGS conftest.$ac_ext 1>&5'
-ac_link='${CXX-g++} -o conftest $CXXFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
+ac_link='${CXX-g++} -o conftest${ac_exeext} $CXXFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
 cross_compiling=$ac_cv_prog_cxx_cross
 
-cat > conftest.$ac_ext <<EOF
-#line 827 "configure"
+cat > conftest.$ac_ext << EOF
+
+#line 903 "configure"
 #include "confdefs.h"
-main(){return(0);}
+
+int main(){return(0);}
 EOF
-if { (eval echo configure:831: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest; then
+if { (eval echo configure:908: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest${ac_exeext}; then
   ac_cv_prog_cxx_works=yes
   # If we can't run a trivial program, we are probably using a cross compiler.
   if (./conftest; exit) 2>/dev/null; then
@@ -845,7 +922,7 @@ ac_ext=c
 # CFLAGS is not in ac_cpp because -g, -O, etc. are not valid cpp options.
 ac_cpp='$CPP $CPPFLAGS'
 ac_compile='${CC-cc} -c $CFLAGS $CPPFLAGS conftest.$ac_ext 1>&5'
-ac_link='${CC-cc} -o conftest $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
+ac_link='${CC-cc} -o conftest${ac_exeext} $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
 cross_compiling=$ac_cv_prog_cc_cross
 
 echo "$ac_t""$ac_cv_prog_cxx_works" 1>&6
@@ -853,12 +930,12 @@ if test $ac_cv_prog_cxx_works = no; then
   { echo "configure: error: installation or configuration problem: C++ compiler cannot create executables." 1>&2; exit 1; }
 fi
 echo $ac_n "checking whether the C++ compiler ($CXX $CXXFLAGS $LDFLAGS) is a cross-compiler""... $ac_c" 1>&6
-echo "configure:857: checking whether the C++ compiler ($CXX $CXXFLAGS $LDFLAGS) is a cross-compiler" >&5
+echo "configure:934: checking whether the C++ compiler ($CXX $CXXFLAGS $LDFLAGS) is a cross-compiler" >&5
 echo "$ac_t""$ac_cv_prog_cxx_cross" 1>&6
 cross_compiling=$ac_cv_prog_cxx_cross
 
 echo $ac_n "checking whether we are using GNU C++""... $ac_c" 1>&6
-echo "configure:862: checking whether we are using GNU C++" >&5
+echo "configure:939: checking whether we are using GNU C++" >&5
 if eval "test \"`echo '$''{'ac_cv_prog_gxx'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
 else
@@ -867,7 +944,7 @@ else
   yes;
 #endif
 EOF
-if { ac_try='${CXX-g++} -E conftest.C'; { (eval echo configure:871: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }; } | egrep yes >/dev/null 2>&1; then
+if { ac_try='${CXX-g++} -E conftest.C'; { (eval echo configure:948: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }; } | egrep yes >/dev/null 2>&1; then
   ac_cv_prog_gxx=yes
 else
   ac_cv_prog_gxx=no
@@ -878,11 +955,15 @@ echo "$ac_t""$ac_cv_prog_gxx" 1>&6
 
 if test $ac_cv_prog_gxx = yes; then
   GXX=yes
-  ac_test_CXXFLAGS="${CXXFLAGS+set}"
-  ac_save_CXXFLAGS="$CXXFLAGS"
-  CXXFLAGS=
-  echo $ac_n "checking whether ${CXX-g++} accepts -g""... $ac_c" 1>&6
-echo "configure:886: checking whether ${CXX-g++} accepts -g" >&5
+else
+  GXX=
+fi
+
+ac_test_CXXFLAGS="${CXXFLAGS+set}"
+ac_save_CXXFLAGS="$CXXFLAGS"
+CXXFLAGS=
+echo $ac_n "checking whether ${CXX-g++} accepts -g""... $ac_c" 1>&6
+echo "configure:967: checking whether ${CXX-g++} accepts -g" >&5
 if eval "test \"`echo '$''{'ac_cv_prog_cxx_g'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
 else
@@ -897,20 +978,24 @@ rm -f conftest*
 fi
 
 echo "$ac_t""$ac_cv_prog_cxx_g" 1>&6
-  if test "$ac_test_CXXFLAGS" = set; then
-    CXXFLAGS="$ac_save_CXXFLAGS"
-  elif test $ac_cv_prog_cxx_g = yes; then
+if test "$ac_test_CXXFLAGS" = set; then
+  CXXFLAGS="$ac_save_CXXFLAGS"
+elif test $ac_cv_prog_cxx_g = yes; then
+  if test "$GXX" = yes; then
     CXXFLAGS="-g -O2"
   else
-    CXXFLAGS="-O2"
+    CXXFLAGS="-g"
   fi
 else
-  GXX=
-  test "${CXXFLAGS+set}" = set || CXXFLAGS="-g"
+  if test "$GXX" = yes; then
+    CXXFLAGS="-O2"
+  else
+    CXXFLAGS=
+  fi
 fi
 
                       echo $ac_n "checking how to run the C++ preprocessor""... $ac_c" 1>&6
-echo "configure:914: checking how to run the C++ preprocessor" >&5
+echo "configure:999: checking how to run the C++ preprocessor" >&5
 if test -z "$CXXCPP"; then
 if eval "test \"`echo '$''{'ac_cv_prog_CXXCPP'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
@@ -919,17 +1004,17 @@ else
 # CXXFLAGS is not in ac_cpp because -g, -O, etc. are not valid cpp options.
 ac_cpp='$CXXCPP $CPPFLAGS'
 ac_compile='${CXX-g++} -c $CXXFLAGS $CPPFLAGS conftest.$ac_ext 1>&5'
-ac_link='${CXX-g++} -o conftest $CXXFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
+ac_link='${CXX-g++} -o conftest${ac_exeext} $CXXFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
 cross_compiling=$ac_cv_prog_cxx_cross
   CXXCPP="${CXX-g++} -E"
   cat > conftest.$ac_ext <<EOF
-#line 927 "configure"
+#line 1012 "configure"
 #include "confdefs.h"
 #include <stdlib.h>
 EOF
 ac_try="$ac_cpp conftest.$ac_ext >/dev/null 2>conftest.out"
-{ (eval echo configure:932: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }
-ac_err=`grep -v '^ *+' conftest.out`
+{ (eval echo configure:1017: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }
+ac_err=`grep -v '^ *+' conftest.out | grep -v "^conftest.${ac_ext}\$"`
 if test -z "$ac_err"; then
   :
 else
@@ -945,7 +1030,7 @@ ac_ext=c
 # CFLAGS is not in ac_cpp because -g, -O, etc. are not valid cpp options.
 ac_cpp='$CPP $CPPFLAGS'
 ac_compile='${CC-cc} -c $CFLAGS $CPPFLAGS conftest.$ac_ext 1>&5'
-ac_link='${CC-cc} -o conftest $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
+ac_link='${CC-cc} -o conftest${ac_exeext} $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
 cross_compiling=$ac_cv_prog_cc_cross
 fi
 fi
@@ -963,7 +1048,7 @@ echo "$ac_t""$CXXCPP" 1>&6
 # SVR4 /usr/ucb/install, which tries to use the nonexistent group "staff"
 # ./install, which can be erroneously created by make from ./install.sh.
 echo $ac_n "checking for a BSD compatible install""... $ac_c" 1>&6
-echo "configure:967: checking for a BSD compatible install" >&5
+echo "configure:1052: checking for a BSD compatible install" >&5
 if test -z "$INSTALL"; then
 if eval "test \"`echo '$''{'cl_cv_path_install'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
@@ -988,7 +1073,7 @@ else
 	      # AIX installbsd doesn't work without option "-g".
 	      :
 	    else
-	      ac_cv_path_install="$ac_dir/$ac_prog -c"
+	      cl_cv_path_install="$ac_dir/$ac_prog -c"
 	      break 2
 	    fi
 	  fi
@@ -1016,8 +1101,8 @@ if test -z "$INSTALL_DATA"; then
 fi
 
                       echo $ac_n "checking for working throw()""... $ac_c" 1>&6
-echo "configure:1020: checking for working throw()" >&5
-if eval "test \"`echo '$''{'gp_cxx_throw_decl'+set}'`\" = set"; then
+echo "configure:1105: checking for working throw()" >&5
+if eval "test \"`echo '$''{'gp_cv_cxx_throw_decl'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
 else
   
@@ -1026,11 +1111,11 @@ ac_ext=C
 # CXXFLAGS is not in ac_cpp because -g, -O, etc. are not valid cpp options.
 ac_cpp='$CXXCPP $CPPFLAGS'
 ac_compile='${CXX-g++} -c $CXXFLAGS $CPPFLAGS conftest.$ac_ext 1>&5'
-ac_link='${CXX-g++} -o conftest $CXXFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
+ac_link='${CXX-g++} -o conftest${ac_exeext} $CXXFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
 cross_compiling=$ac_cv_prog_cxx_cross
 
 cat > conftest.$ac_ext <<EOF
-#line 1034 "configure"
+#line 1119 "configure"
 #include "confdefs.h"
 #include <stdlib.h>
 void operator delete (void* ptr) throw() {}
@@ -1038,28 +1123,28 @@ int main() {
 
 ; return 0; }
 EOF
-if { (eval echo configure:1042: \"$ac_compile\") 1>&5; (eval $ac_compile) 2>&5; }; then
+if { (eval echo configure:1127: \"$ac_compile\") 1>&5; (eval $ac_compile) 2>&5; }; then
   rm -rf conftest*
-  gp_cxx_throw_decl=yes
+  gp_cv_cxx_throw_decl=yes
 else
   echo "configure: failed program was:" >&5
   cat conftest.$ac_ext >&5
   rm -rf conftest*
-  gp_cxx_throw_decl=no
+  gp_cv_cxx_throw_decl=no
 fi
 rm -f conftest*
 ac_ext=c
 # CFLAGS is not in ac_cpp because -g, -O, etc. are not valid cpp options.
 ac_cpp='$CPP $CPPFLAGS'
 ac_compile='${CC-cc} -c $CFLAGS $CPPFLAGS conftest.$ac_ext 1>&5'
-ac_link='${CC-cc} -o conftest $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
+ac_link='${CC-cc} -o conftest${ac_exeext} $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS 1>&5'
 cross_compiling=$ac_cv_prog_cc_cross
 
 
 fi
 
-echo "$ac_t""$gp_cxx_throw_decl" 1>&6
-if test $gp_cxx_throw_decl = yes; then
+echo "$ac_t""$gp_cv_cxx_throw_decl" 1>&6
+if test $gp_cv_cxx_throw_decl = yes; then
   cat >> confdefs.h <<\EOF
 #define HAVE_THROW_DECL 1
 EOF
@@ -1069,18 +1154,18 @@ for ac_hdr in unistd.h sys/time.h sys/resource.h
 do
 ac_safe=`echo "$ac_hdr" | sed 'y%./+-%__p_%'`
 echo $ac_n "checking for $ac_hdr""... $ac_c" 1>&6
-echo "configure:1073: checking for $ac_hdr" >&5
+echo "configure:1158: checking for $ac_hdr" >&5
 if eval "test \"`echo '$''{'ac_cv_header_$ac_safe'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
 else
   cat > conftest.$ac_ext <<EOF
-#line 1078 "configure"
+#line 1163 "configure"
 #include "confdefs.h"
 #include <$ac_hdr>
 EOF
 ac_try="$ac_cpp conftest.$ac_ext >/dev/null 2>conftest.out"
-{ (eval echo configure:1083: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }
-ac_err=`grep -v '^ *+' conftest.out`
+{ (eval echo configure:1168: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }
+ac_err=`grep -v '^ *+' conftest.out | grep -v "^conftest.${ac_ext}\$"`
 if test -z "$ac_err"; then
   rm -rf conftest*
   eval "ac_cv_header_$ac_safe=yes"
@@ -1109,12 +1194,12 @@ done
 for ac_func in getrlimit
 do
 echo $ac_n "checking for $ac_func""... $ac_c" 1>&6
-echo "configure:1113: checking for $ac_func" >&5
+echo "configure:1198: checking for $ac_func" >&5
 if eval "test \"`echo '$''{'ac_cv_func_$ac_func'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
 else
   cat > conftest.$ac_ext <<EOF
-#line 1118 "configure"
+#line 1203 "configure"
 #include "confdefs.h"
 /* System header to define __stub macros and hopefully few prototypes,
     which can conflict with char $ac_func(); below.  */
@@ -1137,7 +1222,7 @@ $ac_func();
 
 ; return 0; }
 EOF
-if { (eval echo configure:1141: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest; then
+if { (eval echo configure:1226: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest${ac_exeext}; then
   rm -rf conftest*
   eval "ac_cv_func_$ac_func=yes"
 else
@@ -1165,12 +1250,12 @@ done
 for ac_func in setrlimit
 do
 echo $ac_n "checking for $ac_func""... $ac_c" 1>&6
-echo "configure:1169: checking for $ac_func" >&5
+echo "configure:1254: checking for $ac_func" >&5
 if eval "test \"`echo '$''{'ac_cv_func_$ac_func'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
 else
   cat > conftest.$ac_ext <<EOF
-#line 1174 "configure"
+#line 1259 "configure"
 #include "confdefs.h"
 /* System header to define __stub macros and hopefully few prototypes,
     which can conflict with char $ac_func(); below.  */
@@ -1193,7 +1278,7 @@ $ac_func();
 
 ; return 0; }
 EOF
-if { (eval echo configure:1197: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest; then
+if { (eval echo configure:1282: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest${ac_exeext}; then
   rm -rf conftest*
   eval "ac_cv_func_$ac_func=yes"
 else
@@ -1242,7 +1327,7 @@ EOF
 # Ultrix sh set writes to stderr and can't be redirected directly,
 # and sets the high bit in the cache file unless we assign to the vars.
 (set) 2>&1 |
-  case `(ac_space=' '; set) 2>&1` in
+  case `(ac_space=' '; set | grep ac_space) 2>&1` in
   *ac_space=\ *)
     # `set' does not quote correctly, so add quotes (double-quote substitution
     # turns \\\\ into \\, and sed turns \\ into \).
@@ -1309,7 +1394,7 @@ do
     echo "running \${CONFIG_SHELL-/bin/sh} $0 $ac_configure_args --no-create --no-recursion"
     exec \${CONFIG_SHELL-/bin/sh} $0 $ac_configure_args --no-create --no-recursion ;;
   -version | --version | --versio | --versi | --vers | --ver | --ve | --v)
-    echo "$CONFIG_STATUS generated by autoconf version 2.12"
+    echo "$CONFIG_STATUS generated by autoconf version 2.13"
     exit 0 ;;
   -help | --help | --hel | --he | --h)
     echo "\$ac_cs_usage"; exit 0 ;;
@@ -1328,9 +1413,11 @@ sed 's/%@/@@/; s/@%/@@/; s/%g\$/@g/; /@g\$/s/[\\\\&%]/\\\\&/g;
  s/@@/%@/; s/@@/@%/; s/@g\$/%g/' > conftest.subs <<\\CEOF
 $ac_vpsub
 $extrasub
+s%@SHELL@%$SHELL%g
 s%@CFLAGS@%$CFLAGS%g
 s%@CPPFLAGS@%$CPPFLAGS%g
 s%@CXXFLAGS@%$CXXFLAGS%g
+s%@FFLAGS@%$FFLAGS%g
 s%@DEFS@%$DEFS%g
 s%@LDFLAGS@%$LDFLAGS%g
 s%@LIBS@%$LIBS%g
diff --git a/contrib/gperf/src/configure.in b/contrib/gperf/src/configure.in
index 68d267c22db2..e8880d507f85 100644
--- a/contrib/gperf/src/configure.in
+++ b/contrib/gperf/src/configure.in
@@ -1,6 +1,6 @@
 dnl autoconf configuration for gperf/src
 
-dnl Copyright (C) 1998 Free Software Foundation, Inc.
+dnl Copyright (C) 1998, 2000 Free Software Foundation, Inc.
 dnl written by Douglas C. Schmidt (schmidt@ics.uci.edu)
 dnl
 dnl This file is part of GNU GPERF.
@@ -40,16 +40,16 @@ dnl
 dnl           checks for compiler characteristics
 dnl
 AC_MSG_CHECKING([for working throw()])
-AC_CACHE_VAL(gp_cxx_throw_decl,[
+AC_CACHE_VAL(gp_cv_cxx_throw_decl,[
 AC_LANG_SAVE()
 AC_LANG_CPLUSPLUS()
 AC_TRY_COMPILE([#include <stdlib.h>
 void operator delete (void* ptr) throw() {}], [],
-gp_cxx_throw_decl=yes, gp_cxx_throw_decl=no)
+gp_cv_cxx_throw_decl=yes, gp_cv_cxx_throw_decl=no)
 AC_LANG_RESTORE()
 ])
-AC_MSG_RESULT([$]gp_cxx_throw_decl)
-if test [$]gp_cxx_throw_decl = yes; then
+AC_MSG_RESULT([$]gp_cv_cxx_throw_decl)
+if test [$]gp_cv_cxx_throw_decl = yes; then
   AC_DEFINE(HAVE_THROW_DECL)
 fi
 dnl
diff --git a/contrib/gperf/src/gen-perf.cc b/contrib/gperf/src/gen-perf.cc
index 0d0ad4140dbb..0b5109d4ff4d 100644
--- a/contrib/gperf/src/gen-perf.cc
+++ b/contrib/gperf/src/gen-perf.cc
@@ -1,6 +1,6 @@
 /* Provides high-level routines to manipulate the keywork list
    structures the code generation output.
-   Copyright (C) 1989-1998 Free Software Foundation, Inc.
+   Copyright (C) 1989-1998, 2000 Free Software Foundation, Inc.
    written by Douglas C. Schmidt (schmidt@ics.uci.edu)
 
 This file is part of GNU GPERF.
@@ -81,37 +81,45 @@ Gen_Perf::Gen_Perf (void)
 }
 
 /* Merge two disjoint hash key multisets to form the ordered disjoint union of the sets.
-   (In a multiset, an element can occur multiple times).
+   (In a multiset, an element can occur multiple times.)
    Precondition: both set_1 and set_2 must be ordered. Returns the length
    of the combined set. */
 
 inline int
-Gen_Perf::compute_disjoint_union (const char *set_1, const char *set_2, char *set_3)
+Gen_Perf::compute_disjoint_union  (const char *set_1, int size_1, const char *set_2, int size_2, char *set_3)
 {
   T (Trace t ("Gen_Perf::compute_disjoint_union");)
   char *base = set_3;
 
-  while (*set_1 && *set_2)
+  while (size_1 > 0 && size_2 > 0)
     if (*set_1 == *set_2)
-      set_1++, set_2++;
+      set_1++, size_1--, set_2++, size_2--;
     else
       {
-        *set_3 = *set_1 < *set_2 ? *set_1++ : *set_2++;
-        if (set_3 == base || *set_3 != *(set_3-1)) set_3++;
+        char next;
+        if (*set_1 < *set_2)
+          next = *set_1++, size_1--;
+        else
+          next = *set_2++, size_2--;
+        if (set_3 == base || next != set_3[-1])
+          *set_3++ = next;
       }
 
-  while (*set_1)
+  while (size_1 > 0)
     {
-      *set_3 = *set_1++;
-      if (set_3 == base || *set_3 != *(set_3-1)) set_3++;
+      char next;
+      next = *set_1++, size_1--;
+      if (set_3 == base || next != set_3[-1])
+        *set_3++ = next;
     }
 
-  while (*set_2)
+  while (size_2 > 0)
     {
-      *set_3 = *set_2++;
-      if (set_3 == base || *set_3 != *(set_3-1)) set_3++;
+      char next;
+      next = *set_2++, size_2--;
+      if (set_3 == base || next != set_3[-1])
+        *set_3++ = next;
     }
-  *set_3 = '\0';
   return set_3 - base;
 }
 
@@ -146,10 +154,12 @@ inline int
 Gen_Perf::hash (List_Node *key_node)
 {
   T (Trace t ("Gen_Perf::hash");)
-  int   sum = option[NOLENGTH] ? 0 : key_node->length;
+  int sum = option[NOLENGTH] ? 0 : key_node->key_length;
 
-  for (const char *ptr = key_node->char_set; *ptr; ptr++)
-      sum += asso_values[(unsigned char)(*ptr)];
+  const char *p = key_node->char_set;
+  int i = key_node->char_set_length;
+  for (; i > 0; p++, i--)
+      sum += asso_values[(unsigned char)(*p)];
 
   return key_node->hash_value = sum;
 }
@@ -209,28 +219,35 @@ Gen_Perf::change (List_Node *prior, List_Node *curr)
 {
   T (Trace t ("Gen_Perf::change");)
   static char *union_set;
+  int union_set_length;
 
   if (!union_set)
-    union_set = new char [2 * option.get_max_keysig_size () + 1];
+    union_set = new char [2 * option.get_max_keysig_size ()];
 
   if (option[DEBUG])
     {
-      fprintf (stderr, "collision on keyword #%d, prior = \"%s\", curr = \"%s\" hash = %d\n",
-               num_done, prior->key, curr->key, curr->hash_value);
+      fprintf (stderr, "collision on keyword #%d, prior = \"%.*s\", curr = \"%.*s\" hash = %d\n",
+               num_done,
+               prior->key_length, prior->key,
+               curr->key_length, curr->key,
+               curr->hash_value);
       fflush (stderr);
     }
-  sort_set (union_set, compute_disjoint_union (prior->char_set, curr->char_set, union_set));
+  union_set_length = compute_disjoint_union (prior->char_set, prior->char_set_length, curr->char_set, curr->char_set_length, union_set);
+  sort_set (union_set, union_set_length);
 
   /* Try changing some values, if change doesn't alter other values continue normal action. */
   fewest_collisions++;
 
-  for (char *temp = union_set; *temp; temp++)
-    if (!affects_prev (*temp, curr))
+  const char *p = union_set;
+  int i = union_set_length;
+  for (; i > 0; p++, i--)
+    if (!affects_prev (*p, curr))
       {
         if (option[DEBUG])
           {
             fprintf (stderr, " by changing asso_value['%c'] (char #%d) to %d\n",
-                     *temp, temp - union_set + 1, asso_values[(unsigned char)(*temp)]);
+                     *p, p - union_set + 1, asso_values[(unsigned char)(*p)]);
             fflush (stderr);
           }
         return; /* Good, doesn't affect previous hash values, we'll take it. */
diff --git a/contrib/gperf/src/gen-perf.h b/contrib/gperf/src/gen-perf.h
index 09b8727dbb2e..602d160131d3 100644
--- a/contrib/gperf/src/gen-perf.h
+++ b/contrib/gperf/src/gen-perf.h
@@ -3,7 +3,7 @@
 /* Provides high-level routines to manipulate the keyword list
    structures the code generation output.
 
-   Copyright (C) 1989-1998 Free Software Foundation, Inc.
+   Copyright (C) 1989-1998, 2000 Free Software Foundation, Inc.
    written by Douglas C. Schmidt (schmidt@ics.uci.edu)
 
 This file is part of GNU GPERF.
@@ -38,7 +38,7 @@ private:
   void        change (List_Node *prior, List_Node *curr);
   int         affects_prev (char c, List_Node *curr);
   static int  hash (List_Node *key_node);
-  static int  compute_disjoint_union (const char *set_1, const char *set_2, char *set_3);
+  static int  compute_disjoint_union (const char *set_1, int size_1, const char *set_2, int size_2, char *set_3);
   static void sort_set (char *union_set, int len);
 
 public:
diff --git a/contrib/gperf/src/hash-table.cc b/contrib/gperf/src/hash-table.cc
index db8d6ddcf03f..a147674b3074 100644
--- a/contrib/gperf/src/hash-table.cc
+++ b/contrib/gperf/src/hash-table.cc
@@ -1,5 +1,5 @@
 /* Hash table for checking keyword links.  Implemented using double hashing.
-   Copyright (C) 1989-1998 Free Software Foundation, Inc.
+   Copyright (C) 1989-1998, 2000 Free Software Foundation, Inc.
    written by Douglas C. Schmidt (schmidt@ics.uci.edu)
 
 This file is part of GNU GPERF.
@@ -26,8 +26,6 @@ Software Foundation, 59 Temple Place - Suite 330, Boston, MA 02111, USA.  */
 #include "options.h"
 #include "trace.h"
 
-#define NIL(TYPE) (TYPE *)0
-
 /* The size of the hash table is always the smallest power of 2 >= the size
    indicated by the user.  This allows several optimizations, including
    the use of double hashing and elimination of the mod instruction.
@@ -37,8 +35,8 @@ Software Foundation, 59 Temple Place - Suite 330, Boston, MA 02111, USA.  */
    This compromises information hiding somewhat, but greatly reduces
    memory fragmentation, since we can now use alloca! */
 
-Hash_Table::Hash_Table (List_Node **table_ptr, int s):
-     table (table_ptr), size (s), collisions (0)
+Hash_Table::Hash_Table (List_Node **table_ptr, int s, int ignore_len):
+     table (table_ptr), size (s), collisions (0), ignore_length (ignore_len)
 {
   T (Trace t ("Hash_Table::Hash_Table");)
   memset ((char *) table, 0, size * sizeof (*table));
@@ -60,8 +58,10 @@ Hash_Table::~Hash_Table (void)
 
       for (int i = size - 1; i >= 0; i--)
         if (table[i])
-          fprintf (stderr, "%8d, %*s, %s\n",
-                   i, field_width, table[i]->char_set, table[i]->key);
+          fprintf (stderr, "%8d, %*.*s, %.*s\n",
+                   i,
+                   field_width, table[i]->char_set_length, table[i]->char_set,
+                   table[i]->key_length, table[i]->key);
 
       fprintf (stderr, "\nend dumping hash table\n\n");
     }
@@ -72,20 +72,24 @@ Hash_Table::~Hash_Table (void)
    Uses double hashing. */
 
 List_Node *
-Hash_Table::operator() (List_Node *item, int ignore_length)
+Hash_Table::insert (List_Node *item)
 {
   T (Trace t ("Hash_Table::operator()");)
-  unsigned hash_val  = hashpjw (item->char_set);
-  int      probe     = hash_val & size - 1;
-  int      increment = (hash_val ^ item->length | 1) & size - 1;
+  unsigned hash_val  = hashpjw (item->char_set, item->char_set_length);
+  int      probe     = hash_val & (size - 1);
+  int      increment = ((hash_val ^ item->key_length) | 1) & (size - 1);
 
-  while (table[probe]
-         && (strcmp (table[probe]->char_set, item->char_set)
-             || (!ignore_length && table[probe]->length != item->length)))
+  while (table[probe])
     {
+      if (table[probe]->char_set_length == item->char_set_length
+          && memcmp (table[probe]->char_set, item->char_set, item->char_set_length) == 0
+          && (ignore_length || table[probe]->key_length == item->key_length))
+        return table[probe];
+
       collisions++;
-      probe = probe + increment & size - 1;
+      probe = (probe + increment) & (size - 1);
     }
 
-  return table[probe] ? table[probe] : (table[probe] = item, NIL (List_Node));
+  table[probe] = item;
+  return (List_Node *) 0;
 }
diff --git a/contrib/gperf/src/hash-table.h b/contrib/gperf/src/hash-table.h
index 7d23999f969b..86438d00f7d6 100644
--- a/contrib/gperf/src/hash-table.h
+++ b/contrib/gperf/src/hash-table.h
@@ -2,7 +2,7 @@
 
 /* Hash table used to check for duplicate keyword entries.
 
-   Copyright (C) 1989-1998 Free Software Foundation, Inc.
+   Copyright (C) 1989-1998, 2000 Free Software Foundation, Inc.
    written by Douglas C. Schmidt (schmidt@ics.uci.edu)
 
 This file is part of GNU GPERF.
@@ -32,11 +32,12 @@ private:
   List_Node     **table;      /* Vector of pointers to linked lists of List_Node's. */
   int             size;       /* Size of the vector. */
   int             collisions; /* Find out how well our double hashing is working! */
+  int             ignore_length;
 
 public:
-                  Hash_Table (List_Node **t, int s);
+                  Hash_Table (List_Node **t, int s, int ignore_len);
                  ~Hash_Table (void);
-  List_Node      *operator () (List_Node *item, int ignore_length);
+  List_Node      *insert (List_Node *item);
 };
 
 #endif
diff --git a/contrib/gperf/src/iterator.cc b/contrib/gperf/src/iterator.cc
index dd9d27e2159d..ca66bbb8aca4 100644
--- a/contrib/gperf/src/iterator.cc
+++ b/contrib/gperf/src/iterator.cc
@@ -1,5 +1,5 @@
 /* Provides an Iterator for keyword characters.
-   Copyright (C) 1989-1998 Free Software Foundation, Inc.
+   Copyright (C) 1989-1998, 2000 Free Software Foundation, Inc.
    written by Douglas C. Schmidt (schmidt@ics.uci.edu)
 
 This file is part of GNU GPERF.
@@ -20,7 +20,6 @@ Software Foundation, 59 Temple Place - Suite 330, Boston, MA 02111, USA.  */
 
 #include "iterator.h"
 
-#include <stream.h>
 #include <ctype.h>
 #include "trace.h"
 
diff --git a/contrib/gperf/src/key-list.cc b/contrib/gperf/src/key-list.cc
index 38341f3c6314..1c941a453579 100644
--- a/contrib/gperf/src/key-list.cc
+++ b/contrib/gperf/src/key-list.cc
@@ -1,5 +1,5 @@
 /* Routines for building, ordering, and printing the keyword list.
-   Copyright (C) 1989-1998 Free Software Foundation, Inc.
+   Copyright (C) 1989-1998, 2000 Free Software Foundation, Inc.
    written by Douglas C. Schmidt (schmidt@ics.uci.edu)
 
 This file is part of GNU GPERF.
@@ -21,6 +21,7 @@ Software Foundation, 59 Temple Place - Suite 330, Boston, MA 02111, USA.  */
 #include <stdio.h>
 #include <string.h> /* declares strncpy(), strchr() */
 #include <stdlib.h> /* declares malloc(), free(), abs(), exit(), abort() */
+#include <ctype.h>  /* declares isprint() */
 #include <assert.h> /* defines assert() */
 #include <limits.h> /* defines SCHAR_MAX etc. */
 #include "options.h"
@@ -209,6 +210,144 @@ Key_List::set_output_types (void)
     }
 }
 
+/* Extracts a key from an input line and creates a new List_Node for it. */
+
+static List_Node *
+parse_line (const char *line, const char *delimiters)
+{
+  if (*line == '"')
+    {
+      /* Parse a string in ANSI C syntax. */
+      char *key = new char[strlen(line)];
+      char *kp = key;
+      const char *lp = line + 1;
+
+      for (; *lp;)
+        {
+          char c = *lp;
+
+          if (c == '\0')
+            {
+              fprintf (stderr, "unterminated string: %s\n", line);
+              exit (1);
+            }
+          else if (c == '\\')
+            {
+              c = *++lp;
+              switch (c)
+                {
+                case '0': case '1': case '2': case '3':
+                case '4': case '5': case '6': case '7':
+                  {
+                    int code = 0;
+                    int count = 0;
+                    while (count < 3 && *lp >= '0' && *lp <= '7')
+                      {
+                        code = (code << 3) + (*lp - '0');
+                        lp++;
+                        count++;
+                      }
+                    if (code > UCHAR_MAX)
+                      fprintf (stderr, "octal escape out of range: %s\n", line);
+                    *kp = (char) code;
+                    break;
+                  }
+                case 'x':
+                  {
+                    int code = 0;
+                    int count = 0;
+                    lp++;
+                    while ((*lp >= '0' && *lp <= '9')
+                           || (*lp >= 'A' && *lp <= 'F')
+                           || (*lp >= 'a' && *lp <= 'f'))
+                      {
+                        code = (code << 4)
+                               + (*lp >= 'A' && *lp <= 'F' ? *lp - 'A' + 10 :
+                                  *lp >= 'a' && *lp <= 'f' ? *lp - 'a' + 10 :
+                                  *lp - '0');
+                        lp++;
+                        count++;
+                      }
+                    if (count == 0)
+                      fprintf (stderr, "hexadecimal escape without any hex digits: %s\n", line);
+                    if (code > UCHAR_MAX)
+                      fprintf (stderr, "hexadecimal escape out of range: %s\n", line);
+                    *kp = (char) code;
+                    break;
+                  }
+                case '\\': case '\'': case '"':
+                  *kp = c;
+                  lp++;
+                  break;
+                case 'n':
+                  *kp = '\n';
+                  lp++;
+                  break;
+                case 't':
+                  *kp = '\t';
+                  lp++;
+                  break;
+                case 'r':
+                  *kp = '\r';
+                  lp++;
+                  break;
+                case 'f':
+                  *kp = '\f';
+                  lp++;
+                  break;
+                case 'b':
+                  *kp = '\b';
+                  lp++;
+                  break;
+                case 'a':
+                  *kp = '\a';
+                  lp++;
+                  break;
+                case 'v':
+                  *kp = '\v';
+                  lp++;
+                  break;
+                default:
+                  fprintf (stderr, "invalid escape sequence in string: %s\n", line);
+                  exit (1);
+                }
+            }
+          else if (c == '"')
+            break;
+          else
+            {
+              *kp = c;
+              lp++;
+            }
+          kp++;
+        }
+      lp++;
+      if (*lp != '\0')
+        {
+          if (strchr (delimiters, *lp) == NULL)
+            {
+              fprintf (stderr, "string not followed by delimiter: %s\n", line);
+              exit (1);
+            }
+          lp++;
+        }
+      return new List_Node (key, kp - key, option[TYPE] ? lp : "");
+    }
+  else
+    {
+      /* Not a string. Look for the delimiter. */
+      int len = strcspn (line, delimiters);
+      const char *rest;
+
+      if (line[len] == '\0')
+        rest = "";
+      else
+        /* Skip the first delimiter. */
+        rest = &line[len + 1];
+      return new List_Node (line, len, option[TYPE] ? rest : "");
+    }
+}
+
 /* Reads in all keys from standard input and creates a linked list pointed
    to by Head.  This list is then quickly checked for ``links,'' i.e.,
    unhashable elements possessing identical key sets and lengths. */
@@ -235,13 +374,13 @@ Key_List::read_keys (void)
       const char *delimiter = option.get_delimiter ();
       List_Node  *temp, *trail = 0;
 
-      head = new List_Node (ptr, strcspn (ptr, delimiter));
+      head = parse_line (ptr, delimiter);
 
       for (temp = head;
            (ptr = Read_Line::get_line ()) && strcmp (ptr, "%%");
            temp = temp->next)
         {
-          temp->next = new List_Node (ptr, strcspn (ptr, delimiter));
+          temp->next = parse_line (ptr, delimiter);
           total_keys++;
         }
 
@@ -266,14 +405,14 @@ Key_List::read_keys (void)
 #endif
 
       /* Make large hash table for efficiency. */
-      Hash_Table found_link (table, table_size);
+      Hash_Table found_link (table, table_size, option[NOLENGTH]);
 
       /* Test whether there are any links and also set the maximum length of
         an identifier in the keyword list. */
 
       for (temp = head; temp; temp = temp->next)
         {
-          List_Node *ptr = found_link (temp, option[NOLENGTH]);
+          List_Node *ptr = found_link.insert (temp);
 
           /* Check for links.  We deal with these by building an equivalence class
              of all duplicate values (i.e., links) so that only 1 keyword is
@@ -290,17 +429,19 @@ Key_List::read_keys (void)
 
               /* Complain if user hasn't enabled the duplicate option. */
               if (!option[DUP] || option[DEBUG])
-                fprintf (stderr, "Key link: \"%s\" = \"%s\", with key set \"%s\".\n",
-                                 temp->key, ptr->key, temp->char_set);
+                fprintf (stderr, "Key link: \"%.*s\" = \"%.*s\", with key set \"%.*s\".\n",
+                                 temp->key_length, temp->key,
+                                 ptr->key_length, ptr->key,
+                                 temp->char_set_length, temp->char_set);
             }
           else
             trail = temp;
 
           /* Update minimum and maximum keyword length, if needed. */
-          if (max_key_len < temp->length)
-            max_key_len = temp->length;
-          if (min_key_len > temp->length)
-            min_key_len = temp->length;
+          if (max_key_len < temp->key_length)
+            max_key_len = temp->key_length;
+          if (min_key_len > temp->key_length)
+            min_key_len = temp->key_length;
         }
 
 #if !LARGE_STACK_ARRAYS
@@ -320,6 +461,13 @@ Key_List::read_keys (void)
               exit (1);
             }
         }
+      /* Exit program if an empty string is used as key, since the comparison
+         expressions don't work correctly for looking up an empty string. */
+      if (min_key_len == 0)
+        {
+          fprintf (stderr, "Empty input key is not allowed.\nTo recognize an empty input key, your code should check for\nlen == 0 before calling the gperf generated lookup function.\n");
+          exit (1);
+        }
       if (option[ALLCHARS])
         option.set_keysig_size (max_key_len);
     }
@@ -400,8 +548,10 @@ Key_List::get_occurrence (List_Node *ptr)
   T (Trace t ("Key_List::get_occurrence");)
   int value = 0;
 
-  for (const char *temp = ptr->char_set; *temp; temp++)
-    value += occurrences[(unsigned char)(*temp)];
+  const char *p = ptr->char_set;
+  unsigned int i = ptr->char_set_length;
+  for (; i > 0; p++, i--)
+    value += occurrences[(unsigned char)(*p)];
 
   return value;
 }
@@ -413,8 +563,11 @@ inline void
 Key_List::set_determined (List_Node *ptr)
 {
   T (Trace t ("Key_List::set_determined");)
-  for (const char *temp = ptr->char_set; *temp; temp++)
-    determined[(unsigned char)(*temp)] = 1;
+
+  const char *p = ptr->char_set;
+  unsigned int i = ptr->char_set_length;
+  for (; i > 0; p++, i--)
+    determined[(unsigned char)(*p)] = 1;
 }
 
 /* Returns TRUE if PTR's key set is already completely determined. */
@@ -425,8 +578,10 @@ Key_List::already_determined (List_Node *ptr)
   T (Trace t ("Key_List::already_determined");)
   int is_determined = 1;
 
-  for (const char *temp = ptr->char_set; is_determined && *temp; temp++)
-    is_determined = determined[(unsigned char)(*temp)];
+  const char *p = ptr->char_set;
+  unsigned int i = ptr->char_set_length;
+  for (; is_determined && i > 0; p++, i--)
+    is_determined = determined[(unsigned char)(*p)];
 
   return is_determined;
 }
@@ -653,20 +808,34 @@ Key_List::output_constants (struct Output_Constants& style)
 /* ------------------------------------------------------------------------- */
 
 /* Outputs a keyword, as a string: enclosed in double quotes, escaping
-   backslashes and double quote characters. */
+   backslashes, double quote and unprintable characters. */
 
 static void
-output_string (const char *key)
+output_string (const char *key, int len)
 {
   T (Trace t ("output_string");)
-  char      c;
 
   putchar ('"');
-  while (c = *key++, c != '\0')
+  for (; len > 0; len--)
     {
-      if (c == '"' || c == '\\')
-        putchar ('\\');
-      putchar (c);
+      unsigned char c = (unsigned char) *key++;
+      if (isprint (c))
+        {
+          if (c == '"' || c == '\\')
+            putchar ('\\');
+          putchar (c);
+        }
+      else
+        {
+          /* Use octal escapes, not hexadecimal escapes, because some old
+             C compilers didn't understand hexadecimal escapes, and because
+             hexadecimal escapes are not limited to 2 digits, thus needing
+             special care if the following character happens to be a digit. */
+          putchar ('\\');
+          putchar ('0' + ((c >> 6) & 7));
+          putchar ('0' + ((c >> 3) & 7));
+          putchar ('0' + (c & 7));
+        }
     }
   putchar ('"');
 }
@@ -798,6 +967,36 @@ void Output_Compare_Strncmp::output_comparison (const Output_Expr& expr1,
   expr1.output_expr ();
   printf (" + 1, ");
   expr2.output_expr ();
+  printf (" + 1, len - 1) && ");
+  expr2.output_expr ();
+  printf ("[len] == '\\0'");
+}
+
+/* This class outputs a comparison using memcmp.
+   Note that the length of expr1 (available through the local variable `len')
+   must be verified to be equal to the length of expr2 prior to this
+   comparison. */
+
+struct Output_Compare_Memcmp : public Output_Compare
+{
+  virtual void output_comparison (const Output_Expr& expr1,
+                                  const Output_Expr& expr2) const;
+  Output_Compare_Memcmp () {}
+  virtual ~Output_Compare_Memcmp () {}
+};
+
+void Output_Compare_Memcmp::output_comparison (const Output_Expr& expr1,
+                                               const Output_Expr& expr2) const
+{
+  T (Trace t ("Output_Compare_Memcmp::output_comparison");)
+  printf ("*");
+  expr1.output_expr ();
+  printf (" == *");
+  expr2.output_expr ();
+  printf (" && !memcmp (");
+  expr1.output_expr ();
+  printf (" + 1, ");
+  expr2.output_expr ();
   printf (" + 1, len - 1)");
 }
 
@@ -824,6 +1023,10 @@ Key_List::output_hash_function (void)
   else if (option[KRC] | option[C] | option[ANSIC])
     printf ("#ifdef __GNUC__\n"
             "__inline\n"
+            "#else\n"
+            "#ifdef __cplusplus\n"
+            "inline\n"
+            "#endif\n"
             "#endif\n");
 
   if (option[KRC] | option[C] | option[ANSIC])
@@ -1013,7 +1216,7 @@ Key_List::output_keylength_table (void)
         printf (",");
       if ((column++ % columns) == 0)
         printf("\n%s   ", indent);
-      printf ("%3d", temp->length);
+      printf ("%3d", temp->key_length);
 
       /* Deal with links specially. */
       if (temp->link) // implies option[DUP]
@@ -1023,7 +1226,7 @@ Key_List::output_keylength_table (void)
             printf (",");
             if ((column++ % columns) == 0)
               printf("\n%s   ", indent);
-            printf ("%3d", links->length);
+            printf ("%3d", links->key_length);
           }
 
       index++;
@@ -1042,9 +1245,13 @@ output_keyword_entry (List_Node *temp, const char *indent)
   printf ("%s    ", indent);
   if (option[TYPE])
     printf ("{");
-  output_string (temp->key);
+  output_string (temp->key, temp->key_length);
   if (option[TYPE])
-    printf (",%s}", temp->rest);
+    {
+      if (strlen (temp->rest) > 0)
+        printf (",%s", temp->rest);
+      printf ("}");
+    }
   if (option[DEBUG])
     printf (" /* hash value = %d, index = %d */",
             temp->hash_value, temp->index);
@@ -1053,7 +1260,17 @@ output_keyword_entry (List_Node *temp, const char *indent)
 static void
 output_keyword_blank_entries (int count, const char *indent)
 {
-  const int columns = 9;
+  int columns;
+  if (option[TYPE])
+    {
+      columns = 58 / (6 + strlen (option.get_initializer_suffix()));
+      if (columns == 0)
+        columns = 1;
+    }
+  else
+    {
+      columns = 9;
+    }
   int column = 0;
   for (int i = 0; i < count; i++)
     {
@@ -1069,7 +1286,7 @@ output_keyword_blank_entries (int count, const char *indent)
             printf (", ");
         }
       if (option[TYPE])
-        printf ("{\"\"}");
+        printf ("{\"\"%s}", option.get_initializer_suffix());
       else
         printf ("\"\"");
       column++;
@@ -1183,8 +1400,8 @@ Key_List::output_lookup_array (void)
           int hash_value = temp->hash_value;
           lookup_array[hash_value] = temp->index;
           if (option[DEBUG])
-            fprintf (stderr, "keyword = %s, index = %d\n",
-                     temp->key, temp->index);
+            fprintf (stderr, "keyword = %.*s, index = %d\n",
+                     temp->key_length, temp->key, temp->index);
           if (temp->link
               || (temp->next && hash_value == temp->next->hash_value))
             {
@@ -1200,8 +1417,8 @@ Key_List::output_lookup_array (void)
                       dup_ptr->count++;
                       if (option[DEBUG])
                         fprintf (stderr,
-                                 "static linked keyword = %s, index = %d\n",
-                                 ptr->key, ptr->index);
+                                 "static linked keyword = %.*s, index = %d\n",
+                                 ptr->key_length, ptr->key, ptr->index);
                     }
 
                   if (!(temp->next && hash_value == temp->next->hash_value))
@@ -1211,8 +1428,8 @@ Key_List::output_lookup_array (void)
 
                   dup_ptr->count++;
                   if (option[DEBUG])
-                    fprintf (stderr, "dynamic linked keyword = %s, index = %d\n",
-                             temp->key, temp->index);
+                    fprintf (stderr, "dynamic linked keyword = %.*s, index = %d\n",
+                             temp->key_length, temp->key, temp->index);
                 }
               assert (dup_ptr->count >= 2);
               dup_ptr++;
@@ -1349,15 +1566,15 @@ output_switch_case (List_Node *list, int indent, int *jumps_away)
   T (Trace t ("output_switch_case");)
 
   if (option[DEBUG])
-    printf ("%*s/* hash value = %4d, keyword = \"%s\" */\n",
-            indent, "", list->hash_value, list->key);
+    printf ("%*s/* hash value = %4d, keyword = \"%.*s\" */\n",
+            indent, "", list->hash_value, list->key_length, list->key);
 
   if (option[DUP]
       && (list->link
           || (list->next && list->hash_value == list->next->hash_value)))
     {
       if (option[LENTABLE])
-        printf ("%*slengthptr = &lengthtable[%d]\n",
+        printf ("%*slengthptr = &lengthtable[%d];\n",
                 indent, "", list->index);
       printf ("%*swordptr = &%s[%d];\n",
               indent, "", option.get_wordlist_name (), list->index);
@@ -1383,7 +1600,7 @@ output_switch_case (List_Node *list, int indent, int *jumps_away)
         {
           printf ("%*sif (len == %d)\n"
                   "%*s  {\n",
-                  indent, "", list->length,
+                  indent, "", list->key_length,
                   indent, "");
           indent += 4;
         }
@@ -1392,7 +1609,7 @@ output_switch_case (List_Node *list, int indent, int *jumps_away)
       if (option[TYPE])
         printf ("&%s[%d]", option.get_wordlist_name (), list->index);
       else
-        output_string (list->key);
+        output_string (list->key, list->key_length);
       printf (";\n");
       printf ("%*sgoto compare;\n",
               indent, "");
@@ -1654,60 +1871,64 @@ Key_List::output_lookup_function_body (const Output_Compare& comparison)
               indent -= 4;
               printf ("%*s    }\n", indent, "");
             }
-          printf ("%*s  else if (index < -TOTAL_KEYWORDS)\n"
-                  "%*s    {\n"
-                  "%*s      register int offset = - 1 - TOTAL_KEYWORDS - index;\n",
-                  indent, "", indent, "", indent, "");
-          if (option[LENTABLE])
-            printf ("%*s      register %s%s *lengthptr = &lengthtable[TOTAL_KEYWORDS + lookup[offset]];\n",
-                    indent, "", const_always, smallest_integral_type (max_key_len));
-          printf ("%*s      register ",
-                  indent, "");
-          output_const_type (const_readonly_array, struct_tag);
-          printf ("*wordptr = &%s[TOTAL_KEYWORDS + lookup[offset]];\n",
-                  option.get_wordlist_name ());
-          printf ("%*s      register ",
-                  indent, "");
-          output_const_type (const_readonly_array, struct_tag);
-          printf ("*wordendptr = wordptr + -lookup[offset + 1];\n\n");
-          printf ("%*s      while (wordptr < wordendptr)\n"
-                  "%*s        {\n",
-                  indent, "", indent, "");
-          if (option[LENTABLE])
+          if (total_duplicates > 0)
             {
-              printf ("%*s          if (len == *lengthptr)\n"
-                      "%*s            {\n",
+              printf ("%*s  else if (index < -TOTAL_KEYWORDS)\n"
+                      "%*s    {\n"
+                      "%*s      register int offset = - 1 - TOTAL_KEYWORDS - index;\n",
+                      indent, "", indent, "", indent, "");
+              if (option[LENTABLE])
+                printf ("%*s      register %s%s *lengthptr = &lengthtable[TOTAL_KEYWORDS + lookup[offset]];\n",
+                        indent, "", const_always, smallest_integral_type (max_key_len));
+              printf ("%*s      register ",
+                      indent, "");
+              output_const_type (const_readonly_array, struct_tag);
+              printf ("*wordptr = &%s[TOTAL_KEYWORDS + lookup[offset]];\n",
+                      option.get_wordlist_name ());
+              printf ("%*s      register ",
+                      indent, "");
+              output_const_type (const_readonly_array, struct_tag);
+              printf ("*wordendptr = wordptr + -lookup[offset + 1];\n\n");
+              printf ("%*s      while (wordptr < wordendptr)\n"
+                      "%*s        {\n",
                       indent, "", indent, "");
-              indent += 4;
-            }
-          printf ("%*s          register %schar *s = ",
-                  indent, "", const_always);
-          if (option[TYPE])
-            printf ("wordptr->%s", option.get_key_name ());
-          else
-            printf ("*wordptr");
-          printf (";\n\n"
-                  "%*s          if (",
-                  indent, "");
-          comparison.output_comparison (Output_Expr1 ("str"), Output_Expr1 ("s"));
-          printf (")\n"
-                  "%*s            return %s;\n",
-                  indent, "",
-                  option[TYPE] ? "wordptr" : "s");
-          if (option[LENTABLE])
-            {
-              indent -= 4;
-              printf ("%*s            }\n",
+              if (option[LENTABLE])
+                {
+                  printf ("%*s          if (len == *lengthptr)\n"
+                          "%*s            {\n",
+                          indent, "", indent, "");
+                  indent += 4;
+                }
+              printf ("%*s          register %schar *s = ",
+                      indent, "", const_always);
+              if (option[TYPE])
+                printf ("wordptr->%s", option.get_key_name ());
+              else
+                printf ("*wordptr");
+              printf (";\n\n"
+                      "%*s          if (",
                       indent, "");
+              comparison.output_comparison (Output_Expr1 ("str"), Output_Expr1 ("s"));
+              printf (")\n"
+                      "%*s            return %s;\n",
+                      indent, "",
+                      option[TYPE] ? "wordptr" : "s");
+              if (option[LENTABLE])
+                {
+                  indent -= 4;
+                  printf ("%*s            }\n",
+                          indent, "");
+                }
+              if (option[LENTABLE])
+                printf ("%*s          lengthptr++;\n",
+                        indent, "");
+              printf ("%*s          wordptr++;\n"
+                      "%*s        }\n"
+                      "%*s    }\n",
+                      indent, "", indent, "", indent, "");
             }
-          if (option[LENTABLE])
-            printf ("%*s          lengthptr++;\n",
-                    indent, "");
-          printf ("%*s          wordptr++;\n"
-                  "%*s        }\n"
-                  "%*s    }\n"
-                  "%*s}\n",
-                  indent, "", indent, "", indent, "", indent, "");
+          printf ("%*s}\n",
+                  indent, "");
         }
       else
         {
@@ -1789,10 +2010,15 @@ Key_List::output_lookup_function (void)
   if (!option[GLOBAL])
     output_lookup_tables ();
 
-  if (option[COMP])
-    output_lookup_function_body (Output_Compare_Strncmp ());
+  if (option[LENTABLE])
+    output_lookup_function_body (Output_Compare_Memcmp ());
   else
-    output_lookup_function_body (Output_Compare_Strcmp ());
+    {
+      if (option[COMP])
+        output_lookup_function_body (Output_Compare_Strncmp ());
+      else
+        output_lookup_function_body (Output_Compare_Strcmp ());
+    }
 
   printf ("}\n");
 }
@@ -1916,9 +2142,10 @@ Key_List::dump ()
            field_width, "char_set");
 
   for (List_Node *ptr = head; ptr; ptr = ptr->next)
-    fprintf (stderr, "%11d,%11d,%6d, %*s, %s\n",
-             ptr->hash_value, ptr->length, ptr->index,
-             field_width, ptr->char_set, ptr->key);
+    fprintf (stderr, "%11d,%11d,%6d, %*.*s, %.*s\n",
+             ptr->hash_value, ptr->key_length, ptr->index,
+             field_width, ptr->char_set_length, ptr->char_set,
+             ptr->key_length, ptr->key);
 }
 
 /* Simple-minded constructor action here... */
diff --git a/contrib/gperf/src/list-node.cc b/contrib/gperf/src/list-node.cc
index 6c7888976e13..57a04a071fe4 100644
--- a/contrib/gperf/src/list-node.cc
+++ b/contrib/gperf/src/list-node.cc
@@ -1,5 +1,5 @@
 /* Creates and initializes a new list node.
-   Copyright (C) 1989-1998 Free Software Foundation, Inc.
+   Copyright (C) 1989-1998, 2000 Free Software Foundation, Inc.
    written by Douglas C. Schmidt (schmidt@ics.uci.edu)
 
 This file is part of GNU GPERF.
@@ -57,23 +57,21 @@ List_Node::set_sort (char *base, int len)
    of the total number of keys seen so far.  This is used to initialize
    the INDEX field to some useful value. */
 
-List_Node::List_Node (char *k, int len): link (0), next (0),
-     key (k), rest (option[TYPE] ? k + len + 1 : ""), length (len), index (0)
+List_Node::List_Node (const char *k, int len, const char *r):
+     link (0), next (0), key (k), key_length (len), rest (r), index (0)
 {
   T (Trace t ("List_Node::List_Node");)
-  char *ptr = new char[(option[ALLCHARS] ? len : option.get_max_keysig_size ()) + 1];
-  char *key_set = ptr;
-  k[len]    = '\0';             /* Null terminate KEY to separate it from REST. */
+  char *key_set = new char[(option[ALLCHARS] ? len : option.get_max_keysig_size ())];
+  char *ptr = key_set;
+  int i;
 
-  if (option[ALLCHARS])         /* Use all the character position in the KEY. */
-    for (; *k; k++, ptr++)
+  if (option[ALLCHARS])         /* Use all the character positions in the KEY. */
+    for (i = len; i > 0; k++, ptr++, i--)
       ++occurrences[(unsigned char)(*ptr = *k)];
   else                          /* Only use those character positions specified by the user. */
     {
-      int i;
-
-      /* Iterate thru the list of key_positions, initializing occurrences table
-        and char_set (via char * pointer ptr). */
+      /* Iterate through the list of key_positions, initializing occurrences table
+         and char_set (via char * pointer ptr). */
 
       for (option.reset (); (i = option.get ()) != EOS; )
         {
@@ -90,12 +88,15 @@ List_Node::List_Node (char *k, int len): link (0), next (0),
         keylength, so there are essentially no usable hash positions! */
       if (ptr == char_set && option[NOLENGTH])
         {
-          fprintf (stderr, "Can't hash keyword %s with chosen key positions.\n", key);
+          fprintf (stderr, "Can't hash keyword %.*s with chosen key positions.\n",
+                   key_length, key);
           exit (1);
         }
     }
-  *ptr = '\0';                  /* Terminate this bastard.... */
+
   /* Sort the KEY_SET items alphabetically. */
   set_sort (key_set, ptr - key_set);
+
   char_set = key_set;
+  char_set_length = ptr - key_set;
 }
diff --git a/contrib/gperf/src/list-node.h b/contrib/gperf/src/list-node.h
index 630d91a304ee..3bd21b3359df 100644
--- a/contrib/gperf/src/list-node.h
+++ b/contrib/gperf/src/list-node.h
@@ -2,7 +2,7 @@
 
 /* Data and function members for defining values and operations of a list node.
 
-   Copyright (C) 1989-1998 Free Software Foundation, Inc.
+   Copyright (C) 1989-1998, 2000 Free Software Foundation, Inc.
    written by Douglas C. Schmidt (schmidt@ics.uci.edu)
 
 This file is part of GNU GPERF.
@@ -31,14 +31,15 @@ struct List_Node : private Vectors
   List_Node  *link;              /* TRUE if key has an identical KEY_SET as another key. */
   List_Node  *next;              /* Points to next element on the list. */
   const char *key;               /* Each keyword string stored here. */
+  int         key_length;        /* Length of the key. */
   const char *rest;              /* Additional information for building hash function. */
   const char *char_set;          /* Set of characters to hash, specified by user. */
-  int         length;            /* Length of the key. */
+  int         char_set_length;   /* Length of char_set. */
   int         hash_value;        /* Hash value for the key. */
   int         occurrence;        /* A metric for frequency of key set occurrences. */
   int         index;             /* Position of this node relative to other nodes. */
 
-              List_Node (char *key, int len);
+              List_Node (const char *key, int len, const char *rest);
   static void set_sort (char *base, int len);
 };
 
diff --git a/contrib/gperf/src/main.cc b/contrib/gperf/src/main.cc
index 4e08633a87d3..03b6c7ea242b 100644
--- a/contrib/gperf/src/main.cc
+++ b/contrib/gperf/src/main.cc
@@ -1,5 +1,5 @@
 /* Driver program for the Gen_Perf hash function generator
-   Copyright (C) 1989-1998 Free Software Foundation, Inc.
+   Copyright (C) 1989-1998, 2000 Free Software Foundation, Inc.
    written by Douglas C. Schmidt (schmidt@ics.uci.edu)
 
 This file is part of GNU GPERF.
@@ -67,6 +67,10 @@ main (int argc, char *argv[])
   /* Generates and prints the Gen_Perf hash table. */
   int status = generate_table ();
 
+  /* Check for write error on stdout. */
+  if (fflush (stdout) || ferror (stdout))
+    status = 1;
+
   /* Don't use exit() here, it skips the destructors. */
   return status;
 }
diff --git a/contrib/gperf/src/options.cc b/contrib/gperf/src/options.cc
index 7be9746c8f91..d15e21cee8ce 100644
--- a/contrib/gperf/src/options.cc
+++ b/contrib/gperf/src/options.cc
@@ -1,5 +1,5 @@
 /* Handles parsing the Options provided to the user.
-   Copyright (C) 1989-1998 Free Software Foundation, Inc.
+   Copyright (C) 1989-1998, 2000 Free Software Foundation, Inc.
    written by Douglas C. Schmidt (schmidt@ics.uci.edu)
 
 This file is part of GNU GPERF.
@@ -43,6 +43,9 @@ static const char *const DEFAULT_NAME = "in_word_set";
 /* Default name for the key component. */
 static const char *const DEFAULT_KEY = "name";
 
+/* Default struct initializer suffix. */
+static const char *const DEFAULT_INITIALIZER_SUFFIX = "";
+
 /* Default name for the generated class. */
 static const char *const DEFAULT_CLASS_NAME = "Perfect_Hash";
 
@@ -67,6 +70,7 @@ int Options::iterations;
 char **Options::argument_vector;
 const char *Options::function_name;
 const char *Options::key_name;
+const char *Options::initializer_suffix;
 const char *Options::class_name;
 const char *Options::hash_name;
 const char *Options::wordlist_name;
@@ -79,7 +83,7 @@ void
 Options::short_usage (FILE * strm)
 {
   T (Trace t ("Options::short_usage");)
-  fprintf (strm, "Usage: %s [-cCdDef[num]GhH<hashname>i<init>Ijk<keys>K<keyname>lL<language>nN<function name>ors<size>S<switches>tTvW<wordlistname>Z<class name>7] [input-file]\n"
+  fprintf (strm, "Usage: %s [-cCdDef[num]F<initializers>GhH<hashname>i<init>Ijk<keys>K<keyname>lL<language>nN<function name>ors<size>S<switches>tTvW<wordlistname>Z<class name>7] [input-file]\n"
                  "Try `%s --help' for more information.\n",
                  program_name, program_name);
 }
@@ -116,6 +120,9 @@ Options::long_usage (FILE * strm)
            "Details in the output code:\n"
            "  -K, --slot-name=NAME   Select name of the keyword component in the keyword\n"
            "                         structure.\n"
+           "  -F, --initializer-suffix=INITIALIZERS\n"
+           "                         Initializers for additional components in the keyword\n"
+           "                         structure.\n"
            "  -H, --hash-fn-name=NAME\n"
            "                         Specify name of generated hash function. Default is\n"
            "                         `hash'.\n"
@@ -222,7 +229,50 @@ Options::print_options (void)
   printf ("/* Command-line: ");
 
   for (i = 0; i < argument_count; i++)
-    printf ("%s ", argument_vector[i]);
+    {
+      const char *arg = argument_vector[i];
+
+      /* Escape arg if it contains shell metacharacters. */
+      if (*arg == '-')
+        {
+          putchar (*arg);
+          arg++;
+          if (*arg >= 'A' && *arg <= 'Z' || *arg >= 'a' && *arg <= 'z')
+            {
+              putchar (*arg);
+              arg++;
+            }
+        }
+      if (strpbrk (arg, "\t\n !\"#$&'()*;<>?[\\]`{|}~") != NULL)
+        {
+          if (strchr (arg, '\'') != NULL)
+            {
+              putchar ('"');
+              for (; *arg; arg++)
+                {
+                  if (*arg == '\"' || *arg == '\\' || *arg == '$')
+                    putchar ('\\');
+                  putchar (*arg);
+                }
+              putchar ('"');
+            }
+          else
+            {
+              putchar ('\'');
+              for (; *arg; arg++)
+                {
+                  if (*arg == '\\')
+                    putchar ('\\');
+                  putchar (*arg);
+                }
+              putchar ('\'');
+            }
+        }
+      else
+        printf ("%s", arg);
+
+      printf (" ");
+    }
 
   printf (" */");
 }
@@ -266,6 +316,7 @@ Options::Options (void)
   option_word         = DEFAULTCHARS | C;
   function_name       = DEFAULT_NAME;
   key_name            = DEFAULT_KEY;
+  initializer_suffix  = DEFAULT_INITIALIZER_SUFFIX;
   hash_name           = DEFAULT_HASH_NAME;
   wordlist_name       = DEFAULT_WORDLIST_NAME;
   class_name          = DEFAULT_CLASS_NAME;
@@ -309,6 +360,7 @@ Options::~Options (void)
                "\nhash function name = %s"
                "\nword list name = %s"
                "\nkey name = %s"
+               "\ninitializer suffix = %s"
                "\njump value = %d"
                "\nmax associated value = %d"
                "\ninitial associated value = %d"
@@ -337,7 +389,8 @@ Options::~Options (void)
                option_word & SEVENBIT ? "enabled" : "disabled",
                iterations,
                function_name, hash_name, wordlist_name, key_name,
-               jump, size - 1, initial_asso_value, delimiters, total_switches);
+               initializer_suffix, jump, size - 1, initial_asso_value,
+               delimiters, total_switches);
       if (option_word & ALLCHARS)
         fprintf (stderr, "all characters are used in the hash function\n");
 
@@ -363,6 +416,7 @@ static const struct option long_options[] =
   { "struct-type", no_argument, 0, 't' },
   { "language", required_argument, 0, 'L' },
   { "slot-name", required_argument, 0, 'K' },
+  { "initializer-suffix", required_argument, 0, 'F' },
   { "hash-fn-name", required_argument, 0, 'H' },
   { "lookup-fn-name", required_argument, 0, 'N' },
   { "class-name", required_argument, 0, 'Z' },
@@ -403,7 +457,7 @@ Options::operator() (int argc, char *argv[])
 
   while ((option_char =
             getopt_long (argument_count, argument_vector,
-                         "adcCDe:Ef:gGhH:i:Ij:k:K:lL:nN:oprs:S:tTvW:Z:7",
+                         "adcCDe:Ef:F:gGhH:i:Ij:k:K:lL:nN:oprs:S:tTvW:Z:7",
                          long_options, (int *)0))
          != -1)
     {
@@ -453,11 +507,16 @@ Options::operator() (int argc, char *argv[])
               }
             break;
           }
+        case 'F':
+          {
+            initializer_suffix = /*getopt*/optarg;
+            break;
+          }
         case 'g':               /* Use the ``inline'' keyword for generated sub-routines, ifdef __GNUC__. */
           break;                /* This is now the default. */
         case 'G':               /* Make the keyword table a global variable. */
           {
-                                                option_word |= GLOBAL;
+            option_word |= GLOBAL;
             break;
           }
         case 'h':               /* Displays a list of helpful Options to the user. */
diff --git a/contrib/gperf/src/options.h b/contrib/gperf/src/options.h
index a81afe67e939..9025c86f1732 100644
--- a/contrib/gperf/src/options.h
+++ b/contrib/gperf/src/options.h
@@ -2,7 +2,7 @@
 
 /* Handles parsing the Options provided to the user.
 
-   Copyright (C) 1989-1998 Free Software Foundation, Inc.
+   Copyright (C) 1989-1998, 2000 Free Software Foundation, Inc.
    written by Douglas C. Schmidt (schmidt@ics.uci.edu)
 
 This file is part of GNU GPERF.
@@ -98,6 +98,7 @@ public:
   static int          get_total_switches (void);
   static const char  *get_function_name (void);
   static const char  *get_key_name (void);
+  static const char  *get_initializer_suffix (void);
   static const char  *get_class_name (void);
   static const char  *get_hash_name (void);
   static const char  *get_wordlist_name (void);
@@ -116,6 +117,7 @@ private:
   static char       **argument_vector;                    /* Stores a pointer to command-line vector. */
   static const char  *function_name;                      /* Names used for generated lookup function. */
   static const char  *key_name;                           /* Name used for keyword key. */
+  static const char  *initializer_suffix;                 /* Suffix for empty struct initializers. */
   static const char  *class_name;                         /* Name used for generated C++ class. */
   static const char  *hash_name;                          /* Name used for generated hash function. */
   static const char  *wordlist_name;                      /* Name used for hash table array. */
diff --git a/contrib/gperf/src/options.icc b/contrib/gperf/src/options.icc
index 0d9668f3a6a4..82fe5375c23f 100644
--- a/contrib/gperf/src/options.icc
+++ b/contrib/gperf/src/options.icc
@@ -1,6 +1,6 @@
 /* Inline Functions for options.{h,cc}.
 
-   Copyright (C) 1989-1998 Free Software Foundation, Inc.
+   Copyright (C) 1989-1998, 2000 Free Software Foundation, Inc.
    written by Douglas C. Schmidt (schmidt@ics.uci.edu)
 
 This file is part of GNU GPERF.
@@ -118,6 +118,14 @@ Options::get_key_name (void)
   return key_name;
 }
 
+/* Returns the struct initializer suffix. */
+INLINE const char *
+Options::get_initializer_suffix (void)
+{
+  T (Trace t ("Options::get_initializer_suffix");)
+  return initializer_suffix;
+}
+
 /* Returns the hash function name. */
 INLINE const char *
 Options::get_hash_name (void)
diff --git a/contrib/gperf/src/version.cc b/contrib/gperf/src/version.cc
index 2e5cf68203f9..8f07c695d537 100644
--- a/contrib/gperf/src/version.cc
+++ b/contrib/gperf/src/version.cc
@@ -1,6 +1,6 @@
 /* Current program version number.
 
-   Copyright (C) 1989-1998 Free Software Foundation, Inc.
+   Copyright (C) 1989-1998, 2000 Free Software Foundation, Inc.
    written by Douglas C. Schmidt (schmidt@ics.uci.edu)
 
 This file is part of GNU GPERF.
@@ -19,4 +19,4 @@ You should have received a copy of the GNU General Public License
 along with GNU GPERF; see the file COPYING.  If not, write to the Free
 Software Foundation, 59 Temple Place - Suite 330, Boston, MA 02111, USA.  */
 
-const char *version_string = "2.7";
+const char *version_string = "2.7.2";