Skip to content
Navigation Menu
{{ message }}
forked from matplotlib/matplotlib
-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy path_type1font.py
More file actions
1208 lines (1063 loc) · 39.8 KB
/
Copy path_type1font.py
File metadata and controls
1208 lines (1063 loc) · 39.8 KB
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
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
"""
A class representing a Type 1 font.
This version reads pfa and pfb files and splits them for embedding in
pdf files. It also supports SlantFont and ExtendFont transformations,
similarly to pdfTeX and friends.
Usage::
font = Type1Font(filename)
clear_part, encrypted_part, finale = font.parts
slanted_font = font.transform({'slant': 0.167})
extended_font = font.transform({'extend': 1.2})
subset_font = font.subset([ord(c) for c in 'Hello World'])
Sources:
* Adobe Technical Note #5040, Supporting Downloadable PostScript
Language Fonts.
* Adobe Type 1 Font Format, Adobe Systems Incorporated, third printing,
v1.1, 1993. ISBN 0-201-57044-0.
"""
from __future__ import annotations
import binascii
import functools
import itertools
import logging
import re
import string
import struct
import typing as T
import numpy as np
from matplotlib.cbook import _format_approx
from . import _api
_log = logging.getLogger(__name__)
class _Token:
"""
A token in a PostScript stream.
Attributes
----------
pos : int
Position, i.e. offset from the beginning of the data.
raw : str
Raw text of the token.
kind : str
Description of the token (for debugging or testing).
"""
__slots__ = ('pos', 'raw')
kind = '?'
def __init__(self, pos, raw):
_log.debug('type1font._Token %s at %d: %r', self.kind, pos, raw)
self.pos = pos
self.raw = raw
def __str__(self):
return f"<{self.kind} {self.raw} @{self.pos}>"
def endpos(self):
"""Position one past the end of the token"""
return self.pos + len(self.raw)
def is_keyword(self, *names):
"""Is this a name token with one of the names?"""
return False
def is_slash_name(self):
"""Is this a name token that starts with a slash?"""
return False
def is_delim(self):
"""Is this a delimiter token?"""
return False
def is_number(self):
"""Is this a number token?"""
return False
def value(self):
return self.raw
class _NameToken(_Token):
kind = 'name'
def is_slash_name(self):
return self.raw.startswith('/')
def value(self):
return self.raw[1:]
class _BooleanToken(_Token):
kind = 'boolean'
def value(self):
return self.raw == 'true'
class _KeywordToken(_Token):
kind = 'keyword'
def is_keyword(self, *names):
return self.raw in names
class _DelimiterToken(_Token):
kind = 'delimiter'
def is_delim(self):
return True
def opposite(self):
return {'[': ']', ']': '[',
'{': '}', '}': '{',
'<<': '>>', '>>': '<<'
}[self.raw]
class _WhitespaceToken(_Token):
kind = 'whitespace'
class _StringToken(_Token):
kind = 'string'
_escapes_re = re.compile(r'\\([\\()nrtbf]|[0-7]{1,3})')
_replacements = {'\\': '\\', '(': '(', ')': ')', 'n': '\n',
'r': '\r', 't': '\t', 'b': '\b', 'f': '\f'}
_ws_re = re.compile('[\0\t\r\f\n ]')
@classmethod
def _escape(cls, match):
group = match.group(1)
try:
return cls._replacements[group]
except KeyError:
return chr(int(group, 8))
@functools.lru_cache
def value(self):
if self.raw[0] == '(':
return self._escapes_re.sub(self._escape, self.raw[1:-1])
else:
data = self._ws_re.sub('', self.raw[1:-1])
if len(data) % 2 == 1:
data += '0'
return binascii.unhexlify(data)
class _BinaryToken(_Token):
kind = 'binary'
def value(self):
return self.raw[1:]
class _NumberToken(_Token):
kind = 'number'
def is_number(self):
return True
def value(self):
if '.' not in self.raw:
return int(self.raw)
else:
return float(self.raw)
def _tokenize(data: bytes, skip_ws: bool) -> T.Generator[_Token, int, None]:
"""
A generator that produces _Token instances from Type-1 font code.
The consumer of the generator may send an integer to the tokenizer to
indicate that the next token should be _BinaryToken of the given length.
Parameters
----------
data : bytes
The data of the font to tokenize.
skip_ws : bool
If true, the generator will drop any _WhitespaceTokens from the output.
"""
text = data.decode('ascii', 'replace')
whitespace_or_comment_re = re.compile(r'[\0\t\r\f\n ]+|%[^\r\n]*')
token_re = re.compile(r'/{0,2}[^]\0\t\r\f\n ()<>{}/%[]+')
instring_re = re.compile(r'[()\\]')
hex_re = re.compile(r'^<[0-9a-fA-F\0\t\r\f\n ]*>$')
oct_re = re.compile(r'[0-7]{1,3}')
pos = 0
next_binary: int | None = None
while pos < len(text):
if next_binary is not None:
n = next_binary
next_binary = (yield _BinaryToken(pos, data[pos:pos+n]))
pos += n
continue
match = whitespace_or_comment_re.match(text, pos)
if match:
if not skip_ws:
next_binary = (yield _WhitespaceToken(pos, match.group()))
pos = match.end()
elif text[pos] == '(':
# PostScript string rules:
# - parentheses must be balanced
# - backslashes escape backslashes and parens
# - also codes \n\r\t\b\f and octal escapes are recognized
# - other backslashes do not escape anything
start = pos
pos += 1
depth = 1
while depth:
match = instring_re.search(text, pos)
if match is None:
raise ValueError(
f'Unterminated string starting at {start}')
pos = match.end()
if match.group() == '(':
depth += 1
elif match.group() == ')':
depth -= 1
else: # a backslash
char = text[pos]
if char in r'\()nrtbf':
pos += 1
else:
octal = oct_re.match(text, pos)
if octal:
pos = octal.end()
else:
pass # non-escaping backslash
next_binary = (yield _StringToken(start, text[start:pos]))
elif text[pos:pos + 2] in ('<<', '>>'):
next_binary = (yield _DelimiterToken(pos, text[pos:pos + 2]))
pos += 2
elif text[pos] == '<':
start = pos
try:
pos = text.index('>', pos) + 1
except ValueError as e:
raise ValueError(f'Unterminated hex string starting at {start}'
) from e
if not hex_re.match(text[start:pos]):
raise ValueError(f'Malformed hex string starting at {start}')
next_binary = (yield _StringToken(pos, text[start:pos]))
else:
match = token_re.match(text, pos)
if match:
raw = match.group()
if raw.startswith('/'):
next_binary = (yield _NameToken(pos, raw))
elif match.group() in ('true', 'false'):
next_binary = (yield _BooleanToken(pos, raw))
else:
try:
float(raw)
next_binary = (yield _NumberToken(pos, raw))
except ValueError:
next_binary = (yield _KeywordToken(pos, raw))
pos = match.end()
else:
next_binary = (yield _DelimiterToken(pos, text[pos]))
pos += 1
class _BalancedExpression(_Token):
pass
def _expression(initial, tokens, data):
"""
Consume some number of tokens and return a balanced PostScript expression.
Parameters
----------
initial : _Token
The token that triggered parsing a balanced expression.
tokens : iterator of _Token
Following tokens.
data : bytes
Underlying data that the token positions point to.
Returns
-------
_BalancedExpression
"""
delim_stack = []
token = initial
while True:
if token.is_delim():
if token.raw in ('[', '{'):
delim_stack.append(token)
elif token.raw in (']', '}'):
if not delim_stack:
raise RuntimeError(f"unmatched closing token {token}")
match = delim_stack.pop()
if match.raw != token.opposite():
raise RuntimeError(
f"opening token {match} closed by {token}"
)
if not delim_stack:
break
else:
raise RuntimeError(f'unknown delimiter {token}')
elif not delim_stack:
break
token = next(tokens)
return _BalancedExpression(
initial.pos,
data[initial.pos:token.endpos()].decode('ascii', 'replace')
)
class Type1Font:
"""
A class representing a Type-1 font, for use by backends.
Attributes
----------
parts : tuple
A 3-tuple of the cleartext part, the encrypted part, and the finale of
zeros.
decrypted : bytes
The decrypted form of ``parts[1]``.
prop : dict[str, Any]
A dictionary of font properties. Noteworthy keys include:
- FontName: PostScript name of the font
- Encoding: dict from numeric codes to glyph names
- FontMatrix: bytes object encoding a matrix
- UniqueID: optional font identifier, dropped when modifying the font
- CharStrings: dict from glyph names to byte code
- Subrs: array of byte code subroutines
- OtherSubrs: bytes object encoding some PostScript code
"""
__slots__ = ('parts', 'decrypted', 'prop', '_pos', '_abbr')
# the _pos dict contains (begin, end) indices to parts[0] + decrypted
# so that they can be replaced when transforming the font;
# but since sometimes a definition appears in both parts[0] and decrypted,
# _pos[name] is an array of such pairs
#
# _abbr maps three standard abbreviations to their particular names in
# this font (e.g. 'RD' is named '-|' in some fonts)
def __init__(self, input):
"""
Initialize a Type-1 font.
Parameters
----------
input : str or 3-tuple
Either a pfb file name, or a 3-tuple of already-decoded Type-1
font `~.Type1Font.parts`.
"""
if isinstance(input, tuple) and len(input) == 3:
self.parts = input
else:
with open(input, 'rb') as file:
data = self._read(file)
self.parts = self._split(data)
self.decrypted = self._decrypt(self.parts[1], 'eexec')
self._abbr = {'RD': 'RD', 'ND': 'ND', 'NP': 'NP'}
self._parse()
def _read(self, file):
"""Read the font from a file, decoding into usable parts."""
rawdata = file.read()
if not rawdata.startswith(b'\x80'):
return rawdata
data = b''
while rawdata:
if not rawdata.startswith(b'\x80'):
raise RuntimeError('Broken pfb file (expected byte 128, '
'got %d)' % rawdata[0])
type = rawdata[1]
if type in (1, 2):
length, = struct.unpack('<i', rawdata[2:6])
segment = rawdata[6:6 + length]
rawdata = rawdata[6 + length:]
if type == 1: # ASCII text: include verbatim
data += segment
elif type == 2: # binary data: encode in hexadecimal
data += binascii.hexlify(segment)
elif type == 3: # end of file
break
else:
raise RuntimeError('Unknown segment type %d in pfb file' % type)
return data
def _split(self, data):
"""
Split the Type 1 font into its three main parts.
The three parts are: (1) the cleartext part, which ends in a
eexec operator; (2) the encrypted part; (3) the fixed part,
which contains 512 ASCII zeros possibly divided on various
lines, a cleartomark operator, and possibly something else.
"""
# Cleartext part: just find the eexec and skip whitespace
idx = data.index(b'eexec')
idx += len(b'eexec')
while data[idx] in b' \t\r\n':
idx += 1
len1 = idx
# Encrypted part: find the cleartomark operator and count
# zeros backward
idx = data.rindex(b'cleartomark') - 1
zeros = 512
while zeros and data[idx] in b'0' or data[idx] in b'\r\n':
if data[idx] in b'0':
zeros -= 1
idx -= 1
if zeros:
# this may have been a problem on old implementations that
# used the zeros as necessary padding
_log.info('Insufficiently many zeros in Type 1 font')
# Convert encrypted part to binary (if we read a pfb file, we may end
# up converting binary to hexadecimal to binary again; but if we read
# a pfa file, this part is already in hex, and I am not quite sure if
# even the pfb format guarantees that it will be in binary).
idx1 = len1 + ((idx - len1 + 2) & ~1) # ensure an even number of bytes
binary = binascii.unhexlify(data[len1:idx1])
return data[:len1], binary, data[idx+1:]
@staticmethod
def _decrypt(ciphertext, key, ndiscard=4):
"""
Decrypt ciphertext using the Type-1 font algorithm.
The algorithm is described in Adobe's "Adobe Type 1 Font Format".
The key argument can be an integer, or one of the strings
'eexec' and 'charstring', which map to the key specified for the
corresponding part of Type-1 fonts.
The ndiscard argument should be an integer, usually 4.
That number of bytes is discarded from the beginning of plaintext.
"""
key = _api.check_getitem({'eexec': 55665, 'charstring': 4330}, key=key)
plaintext = []
for byte in ciphertext:
plaintext.append(byte ^ (key >> 8))
key = ((key+byte) * 52845 + 22719) & 0xffff
return bytes(plaintext[ndiscard:])
@staticmethod
def _encrypt(plaintext, key, ndiscard=4):
"""
Encrypt plaintext using the Type-1 font algorithm.
The algorithm is described in Adobe's "Adobe Type 1 Font Format".
The key argument can be an integer, or one of the strings
'eexec' and 'charstring', which map to the key specified for the
corresponding part of Type-1 fonts.
The ndiscard argument should be an integer, usually 4. That
number of bytes is prepended to the plaintext before encryption.
This function prepends NUL bytes for reproducibility, even though
the original algorithm uses random bytes, presumably to avoid
cryptanalysis.
"""
key = _api.check_getitem({'eexec': 55665, 'charstring': 4330}, key=key)
ciphertext = []
for byte in b'\0' * ndiscard + plaintext:
c = byte ^ (key >> 8)
ciphertext.append(c)
key = ((key + c) * 52845 + 22719) & 0xffff
return bytes(ciphertext)
def _parse(self):
"""
Find the values of various font properties. This limited kind
of parsing is described in Chapter 10 "Adobe Type Manager
Compatibility" of the Type-1 spec.
"""
# Start with reasonable defaults
prop = {'Weight': 'Regular', 'ItalicAngle': 0.0, 'isFixedPitch': False,
'UnderlinePosition': -100, 'UnderlineThickness': 50}
pos = {}
data = self.parts[0] + self.decrypted
source = _tokenize(data, True)
while True:
# See if there is a key to be assigned a value
# e.g. /FontName in /FontName /Helvetica def
try:
token = next(source)
except StopIteration:
break
if token.is_delim():
# skip over this - we want top-level keys only
_expression(token, source, data)
if token.is_slash_name():
key = token.value()
keypos = token.pos
else:
continue
# Some values need special parsing
if key in ('Subrs', 'CharStrings', 'Encoding', 'OtherSubrs'):
prop[key], endpos = {
'Subrs': self._parse_subrs,
'CharStrings': self._parse_charstrings,
'Encoding': self._parse_encoding,
'OtherSubrs': self._parse_othersubrs
}[key](source, data)
pos.setdefault(key, []).append((keypos, endpos))
continue
try:
token = next(source)
except StopIteration:
break
if isinstance(token, _KeywordToken):
# constructs like
# FontDirectory /Helvetica known {...} {...} ifelse
# mean the key was not really a key
continue
if token.is_delim():
value = _expression(token, source, data).raw
else:
value = token.value()
# look for a 'def' possibly preceded by access modifiers
try:
kw = next(
kw for kw in source
if not kw.is_keyword('readonly', 'noaccess', 'executeonly')
)
except StopIteration:
break
# sometimes noaccess def and readonly def are abbreviated
if kw.is_keyword('def', self._abbr['ND'], self._abbr['NP']):
prop[key] = value
pos.setdefault(key, []).append((keypos, kw.endpos()))
# detect the standard abbreviations
if value == '{noaccess def}':
self._abbr['ND'] = key
elif value == '{noaccess put}':
self._abbr['NP'] = key
elif value == '{string currentfile exch readstring pop}':
self._abbr['RD'] = key
# Fill in the various *Name properties
if 'FontName' not in prop:
prop['FontName'] = (prop.get('FullName') or
prop.get('FamilyName') or
'Unknown')
if 'FullName' not in prop:
prop['FullName'] = prop['FontName']
if 'FamilyName' not in prop:
extras = ('(?i)([ -](regular|plain|italic|oblique|(semi)?bold|'
'(ultra)?light|extra|condensed))+$')
prop['FamilyName'] = re.sub(extras, '', prop['FullName'])
# Parse FontBBox
toks = [*_tokenize(prop['FontBBox'].encode('ascii'), True)]
if ([tok.kind for tok in toks]
!= ['delimiter', 'number', 'number', 'number', 'number', 'delimiter']
or toks[-1].raw != toks[0].opposite()):
raise RuntimeError(
f"FontBBox should be a size-4 array, was {prop['FontBBox']}")
prop['FontBBox'] = [tok.value() for tok in toks[1:-1]]
# Decrypt the encrypted parts
ndiscard = prop.get('lenIV', 4)
cs = prop['CharStrings']
for key, value in cs.items():
cs[key] = self._decrypt(value, 'charstring', ndiscard)
if 'Subrs' in prop:
prop['Subrs'] = [
self._decrypt(value, 'charstring', ndiscard)
for value in prop['Subrs']
]
self.prop = prop
self._pos = pos
def _parse_subrs(self, tokens, _data):
count_token = next(tokens)
if not count_token.is_number():
raise RuntimeError(
f"Token following /Subrs must be a number, was {count_token}"
)
count = count_token.value()
array = [None] * count
next(t for t in tokens if t.is_keyword('array'))
for _ in range(count):
next(t for t in tokens if t.is_keyword('dup'))
index_token = next(tokens)
if not index_token.is_number():
raise RuntimeError(
"Token following dup in Subrs definition must be a "
f"number, was {index_token}"
)
nbytes_token = next(tokens)
if not nbytes_token.is_number():
raise RuntimeError(
"Second token following dup in Subrs definition must "
f"be a number, was {nbytes_token}"
)
token = next(tokens)
if not token.is_keyword(self._abbr['RD']):
raise RuntimeError(
f"Token preceding subr must be {self._abbr['RD']}, "
f"was {token}"
)
binary_token = tokens.send(1+nbytes_token.value())
array[index_token.value()] = binary_token.value()
return array, next(tokens).endpos()
def _parse_charstrings(self, tokens, _data):
count_token = next(tokens)
if not count_token.is_number():
raise RuntimeError(
"Token following /CharStrings must be a number, "
f"was {count_token}"
)
count = count_token.value()
charstrings = {}
next(t for t in tokens if t.is_keyword('begin'))
while True:
token = next(t for t in tokens
if t.is_keyword('end') or t.is_slash_name())
if token.raw == 'end':
return charstrings, token.endpos()
glyphname = token.value()
nbytes_token = next(tokens)
if not nbytes_token.is_number():
raise RuntimeError(
f"Token following /{glyphname} in CharStrings definition "
f"must be a number, was {nbytes_token}"
)
token = next(tokens)
if not token.is_keyword(self._abbr['RD']):
raise RuntimeError(
f"Token preceding charstring must be {self._abbr['RD']}, "
f"was {token}"
)
binary_token = tokens.send(1+nbytes_token.value())
charstrings[glyphname] = binary_token.value()
@staticmethod
def _parse_encoding(tokens, _data):
# this only works for encodings that follow the Adobe manual
# but some old fonts include non-compliant data - we log a warning
# and return a possibly incomplete encoding
encoding = {}
while True:
token = next(t for t in tokens
if t.is_keyword('StandardEncoding', 'dup', 'def'))
if token.is_keyword('StandardEncoding'):
return _StandardEncoding, token.endpos()
if token.is_keyword('def'):
return encoding, token.endpos()
index_token = next(tokens)
if not index_token.is_number():
_log.warning(
f"Parsing encoding: expected number, got {index_token}"
)
continue
name_token = next(tokens)
if not name_token.is_slash_name():
_log.warning(
f"Parsing encoding: expected slash-name, got {name_token}"
)
continue
encoding[index_token.value()] = name_token.value()
def _parse_othersubrs(self, tokens, data):
init_pos = None
while True:
token = next(tokens)
if init_pos is None:
init_pos = token.pos
if token.is_delim():
_expression(token, tokens, data)
elif token.is_keyword('def', self._abbr['ND']):
return data[init_pos:token.endpos()], token.endpos()
def transform(self, effects):
"""
Return a new font that is slanted and/or extended.
Parameters
----------
effects : dict
A dict with optional entries:
- 'slant' : float, default: 0
Tangent of the angle that the font is to be slanted to the
right. Negative values slant to the left.
- 'extend' : float, default: 1
Scaling factor for the font width. Values less than 1 condense
the glyphs.
Returns
-------
`Type1Font`
"""
fontname = self.prop['FontName']
italicangle = self.prop['ItalicAngle']
array = [
float(x) for x in (self.prop['FontMatrix']
.lstrip('[').rstrip(']').split())
]
oldmatrix = np.eye(3, 3)
oldmatrix[0:3, 0] = array[::2]
oldmatrix[0:3, 1] = array[1::2]
modifier = np.eye(3, 3)
if 'slant' in effects:
slant = effects['slant']
fontname += f'_Slant_{int(1000 * slant)}'
italicangle = round(
float(italicangle) - np.arctan(slant) / np.pi * 180,
5
)
modifier[1, 0] = slant
if 'extend' in effects:
extend = effects['extend']
fontname += f'_Extend_{int(1000 * extend)}'
modifier[0, 0] = extend
newmatrix = np.dot(modifier, oldmatrix)
array[::2] = newmatrix[0:3, 0]
array[1::2] = newmatrix[0:3, 1]
fontmatrix = (
f"[{' '.join(_format_approx(x, 6) for x in array)}]"
)
newparts = self._replace(
[(x, f'/FontName/{fontname} def')
for x in self._pos['FontName']]
+ [(x, f'/ItalicAngle {italicangle} def')
for x in self._pos['ItalicAngle']]
+ [(x, f'/FontMatrix {fontmatrix} readonly def')
for x in self._pos['FontMatrix']]
+ [(x, '') for x in self._pos.get('UniqueID', [])]
)
return Type1Font((
newparts[0],
self._encrypt(newparts[1], 'eexec'),
self.parts[2]
))
def with_encoding(self, encoding):
"""
Change the encoding of the font.
Parameters
----------
encoding : dict
A dictionary mapping character codes to glyph names.
Returns
-------
`Type1Font`
"""
newparts = self._replace(
[(x, '') for x in self._pos.get('UniqueID', [])]
+ [(self._pos['Encoding'][0], self._postscript_encoding(encoding))]
)
return Type1Font((
newparts[0],
self._encrypt(newparts[1], 'eexec'),
self.parts[2]
))
def _replace(self, replacements):
"""
Change the font according to `replacements`
Parameters
----------
replacements : list of ((int, int), str)
Each element is ((pos0, pos1), replacement) where pos0 and
pos1 are indices to the original font data (parts[0] and the
decrypted part concatenated). The data in the interval
pos0:pos1 will be replaced by the replacement text. To
accommodate binary data, the replacement is taken to be in
Latin-1 encoding.
The case where pos0 is inside parts[0] and pos1 inside
the decrypted part is not supported.
Returns
-------
(bytes, bytes)
The new parts[0] and decrypted part (which needs to be
encrypted in the transformed font).
"""
data = bytearray(self.parts[0])
data.extend(self.decrypted)
len0 = len(self.parts[0])
for (pos0, pos1), value in sorted(replacements, reverse=True):
data[pos0:pos1] = value.encode('latin-1')
if pos0 < len(self.parts[0]):
if pos1 >= len(self.parts[0]):
raise RuntimeError(
f"text to be replaced with {value} spans "
"the eexec boundary"
)
len0 += len(value) - pos1 + pos0
return bytes(data[:len0]), bytes(data[len0:])
def subset(self, characters, name_prefix):
"""
Return a new font that only defines the given characters.
Parameters
----------
characters : sequence of bytes
The subset of characters to include. These are indices into the
font's encoding array. The encoding array of a Type-1 font can
only include 256 characters, but other glyphs may be accessed
via the seac operator.
name_prefix : str
Prefix to prepend to the font name.
Returns
-------
`Type1Font`
"""
characters = frozenset(characters)
if _log.isEnabledFor(logging.DEBUG):
_log.debug(
"Subsetting font %s to characters %s = %s",
self.prop['FontName'],
sorted(characters),
[self.prop['Encoding'].get(code) for code in sorted(characters)],
)
encoding = {code: glyph
for code, glyph in self.prop['Encoding'].items()
if code in characters}
encoding[0] = '.notdef'
# todo and done include strings (glyph names)
todo = set(encoding.values())
done = set()
seen_subrs = {0, 1, 2, 3}
while todo:
glyph = todo.pop()
called_glyphs, called_subrs = _CharstringSimulator(self).run(glyph)
todo.update(called_glyphs - done)
seen_subrs.update(called_subrs)
done.add(glyph)
charstrings = self._subset_charstrings(done)
subrs = self._subset_subrs(seen_subrs)
newparts = self._replace(
[(x, f'/FontName /{name_prefix}{self.prop["FontName"]} def')
for x in self._pos['FontName']]
+ [(self._pos['CharStrings'][0], charstrings),
(self._pos['Subrs'][0], subrs),
(self._pos['Encoding'][0], self._postscript_encoding(encoding))
] + [(x, '') for x in self._pos.get('UniqueID', [])]
)
return type(self)((
newparts[0],
self._encrypt(newparts[1], 'eexec'),
self.parts[2]
))
@staticmethod
def _charstring_tokens(data):
"""Parse a Type-1 charstring
Yield opcode names and integer parameters.
"""
data = iter(data)
for byte in data:
if 32 <= byte <= 246:
yield byte - 139
elif 247 <= byte <= 250:
byte2 = next(data)
yield (byte-247) * 256 + byte2 + 108
elif 251 <= byte <= 254:
byte2 = next(data)
yield -(byte-251)*256 - byte2 - 108
elif byte == 255:
bs = bytes(itertools.islice(data, 4))
yield struct.unpack('>i', bs)[0]
elif byte == 12:
byte1 = next(data)
yield {
0: 'dotsection',
1: 'vstem3',
2: 'hstem3',
6: 'seac',
7: 'sbw',
12: 'div',
16: 'callothersubr',
17: 'pop',
33: 'setcurrentpoint'
}[byte1]
else:
yield {
1: 'hstem',
3: 'vstem',
4: 'vmoveto',
5: 'rlineto',
6: 'hlineto',
7: 'vlineto',
8: 'rrcurveto',
9: 'closepath',
10: 'callsubr',
11: 'return',
13: 'hsbw',
14: 'endchar',
21: 'rmoveto',
22: 'hmoveto',
30: 'vhcurveto',
31: 'hvcurveto'
}[byte]
def _postscript_encoding(self, encoding):
"""Return a PostScript encoding array for the encoding."""
return '\n'.join([
'/Encoding 256 array\n0 1 255 { 1 index exch /.notdef put} for',
*(
f'dup {i} /{glyph} put'
for i, glyph in sorted(encoding.items())
if glyph != '.notdef'
),
'readonly def\n',
])
def _subset_charstrings(self, glyphs):
"""Return a PostScript CharStrings array for the glyphs."""
charstrings = self.prop['CharStrings']
lenIV = self.prop.get('lenIV', 4)
ordered = sorted(glyphs)
encrypted = [
self._encrypt(charstrings[glyph], 'charstring', lenIV).decode('latin-1')
for glyph in ordered
]
RD, ND = self._abbr['RD'], self._abbr['ND']
return '\n'.join([
f'/CharStrings {len(ordered)} dict dup begin',
*(
f'/{glyph} {len(enc)} {RD} {enc} {ND}'
for glyph, enc in zip(ordered, encrypted)
),
'end\n',
])
def _subset_subrs(self, indices):
"""Return a PostScript Subrs array for the subroutines."""
# we can't remove subroutines, we just replace unused ones with a stub
subrs = self.prop['Subrs']
n_subrs = len(subrs)
lenIV = self.prop.get('lenIV', 4)
stub = self._encrypt(b'\x0b', 'charstring', lenIV).decode('latin-1')
encrypted = [
self._encrypt(subrs[i], 'charstring', lenIV).decode('latin-1')
if i in indices
else stub
for i in range(n_subrs)
]
RD, ND, NP = self._abbr['RD'], self._abbr['ND'], self._abbr['NP']
return '\n'.join([
f'/Subrs {n_subrs} array',
*(
f'dup {i} {len(enc)} {RD} {enc} {NP}'
for i, enc in enumerate(encrypted)
),
])
You can’t perform that action at this time.