1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
|
#
# Copyright (C) 2017-2020 Dimitar Toshkov Zhekov <dimitar.zhekov@gmail.com>
#
# 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.,
# 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
#
import re
import codecs
from collections import OrderedDict
from enum import IntEnum, unique
import fnutil
# -- Width --
DPARSE_LIMIT = 512
SPARSE_LIMIT = 32000
class Width:
def __init__(self, x, y):
self.x = x
self.y = y
@staticmethod
def parse(name, value, limit):
words = fnutil.split_words(name, value, 2)
return Width(fnutil.parse_dec(name + '.x', words[0], -limit, limit),
fnutil.parse_dec(name + '.y', words[1], -limit, limit))
@staticmethod
def parse_s(name, value):
return Width.parse(name, value, SPARSE_LIMIT)
@staticmethod
def parse_d(name, value):
return Width.parse(name, value, DPARSE_LIMIT)
def __str__(self):
return '%d %d' % (self.x, self.y)
# -- BXX --
class BBX:
def __init__(self, width, height, xoff, yoff):
self.width = width
self.height = height
self.xoff = xoff
self.yoff = yoff
@staticmethod
def parse(name, value):
words = fnutil.split_words(name, value, 4)
return BBX(fnutil.parse_dec('width', words[0], 1, DPARSE_LIMIT),
fnutil.parse_dec('height', words[1], 1, DPARSE_LIMIT),
fnutil.parse_dec('bbxoff', words[2], -DPARSE_LIMIT, DPARSE_LIMIT),
fnutil.parse_dec('bbyoff', words[3], -DPARSE_LIMIT, DPARSE_LIMIT))
def row_size(self):
return (self.width + 7) >> 3
def __str__(self):
return '%d %d %d %d' % (self.width, self.height, self.xoff, self.yoff)
# -- Props --
def skip_comments(line):
return None if line[:7] == b'COMMENT' else line
class Props(OrderedDict):
def __iter__(self):
return self.items().__iter__()
def read(self, input, name, callback=None):
return self.parse(input.read_lines(skip_comments), name, callback)
def parse(self, line, name, callback=None):
if not line or not line.startswith(bytes(name, 'ascii')):
raise Exception(name + ' expected')
value = line[len(name):].lstrip()
self[name] = value
return value if callback is None else callback(name, value)
def set(self, name, value):
self[name] = value if isinstance(value, (bytes, bytearray)) else bytes(str(value), 'ascii')
# -- Base --
class Base:
def __init__(self):
self.props = Props()
self.bbx = None
# -- Char
HEX_BYTES = (48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 65, 66, 67, 68, 69, 70)
class Char(Base):
def __init__(self):
Base.__init__(self)
self.code = -1
self.swidth = None
self.dwidth = None
self.data = None
def bitmap(self):
bitmap = ''
row_size = self.bbx.row_size()
for index in range(0, len(self.data), row_size):
bitmap += self.data[index : index + row_size].hex() + '\n'
return bytes(bitmap, 'ascii').upper()
def _read(self, input):
# HEADER
self.props.read(input, 'STARTCHAR')
self.code = self.props.read(input, 'ENCODING', fnutil.parse_dec)
self.swidth = self.props.read(input, 'SWIDTH', Width.parse_s)
self.dwidth = self.props.read(input, 'DWIDTH', Width.parse_d)
self.bbx = self.props.read(input, 'BBX', BBX.parse)
line = input.read_lines(skip_comments)
if line and line.startswith(b'ATTRIBUTES'):
self.props.parse(line, 'ATTRIBUTES')
line = input.read_lines(skip_comments)
# BITMAP
if self.props.parse(line, 'BITMAP'):
raise Exception('BITMAP expected')
row_len = self.bbx.row_size() * 2
self.data = bytearray()
for _ in range(0, self.bbx.height):
line = input.read_lines(skip_comments)
if not line:
raise Exception('bitmap data expected')
if len(line) == row_len:
self.data += codecs.decode(line, 'hex')
else:
raise Exception('invalid bitmap length')
# FINAL
if input.read_lines(skip_comments) != b'ENDCHAR':
raise Exception('ENDCHAR expected')
return self
@staticmethod
def read(input):
return Char()._read(input) # pylint: disable=protected-access
def write(self, output):
for [name, value] in self.props:
output.write_prop(name, value)
output.write_line(self.bitmap() + b'ENDCHAR')
# -- Font --
@unique
class XLFD(IntEnum):
FOUNDRY = 1
FAMILY_NAME = 2
WEIGHT_NAME = 3
SLANT = 4
SETWIDTH_NAME = 5
ADD_STYLE_NAME = 6
PIXEL_SIZE = 7
POINT_SIZE = 8
RESOLUTION_X = 9
RESOLUTION_Y = 10
SPACING = 11
AVERAGE_WIDTH = 12
CHARSET_REGISTRY = 13
CHARSET_ENCODING = 14
CHARS_MAX = 65535
class Font(Base):
def __init__(self):
Base.__init__(self)
self.xlfd = []
self.chars = []
self.default_code = -1
@property
def bold(self):
return b'bold' in self.xlfd[XLFD.WEIGHT_NAME].lower()
@property
def italic(self):
return self.xlfd[XLFD.SLANT] in [b'I', b'O']
@property
def proportional(self):
return self.xlfd[XLFD.SPACING] == b'P'
def _read(self, input):
# HEADER
line = input.read_line()
if self.props.parse(line, 'STARTFONT') != b'2.1':
raise Exception('STARTFONT 2.1 expected')
self.xlfd = self.props.read(input, 'FONT', lambda name, value: value.split(b'-', 15))
if len(self.xlfd) != 15 or self.xlfd[0] != b'':
raise Exception('non-XLFD font names are not supported')
self.props.read(input, 'SIZE')
self.bbx = self.props.read(input, 'FONTBOUNDINGBOX', BBX.parse)
line = input.read_lines(skip_comments)
if line and line.startswith(b'STARTPROPERTIES'):
num_props = self.props.parse(line, 'STARTPROPERTIES', fnutil.parse_dec)
for _ in range(0, num_props):
line = input.read_lines(skip_comments)
if line is None:
raise Exception('property expected')
match = re.fullmatch(br'(\w+)\s+([-\d"].*)', line)
if not match:
raise Exception('invalid property format')
name = str(match.group(1), 'ascii')
value = match.group(2)
if self.props.get(name) is not None:
raise Exception('duplicate property')
if name == 'DEFAULT_CHAR':
self.default_code = fnutil.parse_dec(name, value)
self.props[name] = value
if self.props.read(input, 'ENDPROPERTIES') != b'':
raise Exception('ENDPROPERTIES expected')
line = input.read_lines(skip_comments)
# GLYPHS
num_chars = fnutil.parse_dec('CHARS', self.props.parse(line, 'CHARS'), 1, CHARS_MAX)
for _ in range(0, num_chars):
self.chars.append(Char.read(input))
if next((char.code for char in self.chars if char.code == self.default_code), -1) != self.default_code:
raise Exception('invalid DEFAULT_CHAR')
# FINAL
if input.read_lines(skip_comments) != b'ENDFONT':
raise Exception('ENDFONT expected')
if input.read_line() is not None:
raise Exception('garbage after ENDFONT')
return self
@staticmethod
def read(input):
return Font()._read(input) # pylint: disable=protected-access
def write(self, output):
for [name, value] in self.props:
output.write_prop(name, value)
for char in self.chars:
char.write(output)
output.write_line(b'ENDFONT')
|