| blob | 7e6bad948722da99f96200954a955a5377ad2935 |
1 .\" man page for ttf2tfm
2 .
3 .TH TTF2TFM 1 15-Aug-1999 "FreeType version 1.3"
4 .SH NAME
5 ttf2tfm \- build TeX metric files from a TrueType font
6 .SH SYNOPSIS
7 .na
8 .nh
9 .B ttf2tfm
10 .in +\n(.ku
11 .sp -1
12 .IR ttffile [ .ttf | .ttc ]
13 [\c
14 .BI -c \ \%caps-height-factor\c
15 ]
16 [\c
17 .BI -e \ \%extension-factor\c
18 ]
19 [\c
20 .BI -E \ \%encoding-id\^\c
21 ]
22 [\c
23 .BI -f \ \%font-index\c
24 ]
25 [\c
26 .B -l\c
27 ]
28 [\c
29 .B -n\c
30 ]
31 [\c
32 .B -N\c
33 ]
34 [\c
35 .B -O\c
36 ]
37 [\c
38 .B -p\ \c
39 .IR \%inencfile [ .enc ]\c
40 ]
41 [\c
42 .BI -P \ \%platform-id\^\c
43 ]
44 [\c
45 .B -q\c
46 ]
47 [\c
48 .BI -r \ \%old-glyphname\ \%new-glyphname\c
49 ]
50 [\c
51 .B -R\ \c
52 .IR \%replacement-file [ .rpl ]\c
53 ]
54 [\c
55 .BI -s \ \%slant-factor\c
56 ]
57 [\c
58 .B -t\ \c
59 .IR \%outencfile [ .enc ]\c
60 ]
61 [\c
62 .B -T\ \c
63 .IR \%inoutencfile [ .enc ]\c
64 ]
65 [\c
66 .B -u\c
67 ]
68 [\c
69 .B -v\ \c
70 .IR \%vplfile [ .vpl ]\c
71 ]
72 [\c
73 .B -V\ \c
74 .IR \%scvplfile [ .vpl ]\c
75 ]
76 [\c
77 .B -x\c
78 ]
79 [\c
80 .BI -y \ \%vertical-shift-factor\c
81 ]
82 [\c
83 .IR \%tfmfile [ .tfm ]\c
84 ]
85 .br
86 .in
87 .B "ttf2tfm --version"
88 |
89 .B --help
90 .ad
91 .hy
92 .
93 .
94 .
95 .\" ====
96 .\" ==== macro definitions
97 .\" ====
98 .
99 .\" here we define \TeX for troff and nroff
100 .if t .ds TX \fRT\\h'-0.1667m'\\v'0.20v'E\\v'-0.20v'\\h'-0.125m'X\fP
101 .if n .ds TX TeX
102 .
103 .\" and here the same for \LaTeX
104 .if t \{\
105 .ie '\*(.T'dvi' \
106 .ds LX \fRL\h'-0.36m'\v'-0.15v'\s-3A\s0\h'-0.15m'\v'0.15v'\fP\*(TX
107 .el .ds LX \fRL\h'-0.36m'\v'-0.22v'\s-2A\s0\h'-0.15m'\v'0.22v'\fP\*(TX
108 .\}
109 .if n .ds LX LaTeX
110 .
111 .\" \LaTeXe
112 .\" note that we need \vareps for TeX instead of \eps which can only be
113 .\" accessed with the \N escape sequence (in the Math Italic font)
114 .if t \{\
115 .ie '\*(.T'dvi' .ds LE \*(LX\h'0.15m'2\v'0.20v'\f(MI\N'34'\fP\v'-0.20v'
116 .el .ds LE \*(LX\h'0.15m'2\v'0.20v'\(*e\v'-0.20v'
117 .\}
118 .if n .ds LE LaTeX\ 2e
119 .
120 .\" a typewriter font
121 .if t \{\
122 .de C
123 \fC\\$1\fP\\$2\fC\\$3\fP\\$4
124 ..
125 .\}
126 .if n \{\
127 .de C
128 \\$1\\$2\\$3\\$4
129 ..
130 .\}
131 .
132 .\" ====
133 .\" ==== end of macro definitions
134 .\" ====
135 .
136 .
137 .
138 .SH DESCRIPTION
139 This program extracts the metric and kerning information of a TrueType
140 font and converts it into metric files usable by \*(TX
141 (quite similar to
142 .B afm2tfm
143 which is part of the
144 .B dvips
145 package; please consult its info files for more details on the various
146 parameters (especially encoding files).
147 .PP
148 Since a TrueType font often contains more than 256\ glyphs, some means
149 are necessary to map a subset of the TrueType glyphs into a \*(TX
150 font.
151 To do this, two mapping tables are needed: the first (called `input' or
152 `raw' encoding) maps the TrueType font to a raw \*(TX font (this mapping
153 table is used by both
154 .B ttf2tfm
155 and
156 .BR ttf2pk ),
157 and the second (called `output' or `virtual' encoding) maps the raw \*(TX
158 font to another (virtual) \*(TX
159 font, providing all kerning and ligature information needed by \*(TX.
160 .PP
161 This two stage mapping has the advantage that one raw font can be
162 accessed with various \*(TX
163 encodings (e.g.\ T1 and OT1) via the virtual font mechanism, and just
164 one
165 .C PK
166 file is necessary.
167 .PP
168 For CJKV (Chinese/Japanese/Korean/old Vietnamese) fonts, a different
169 mechanism is provided (see
170 .B "SUBFONT DEFINITION FILES"
171 below).
172 .
173 .
174 .SH PARAMETERS
175 Most of the command line switch names are the same as in
176 .B afm2tfm
177 for convenience.
178 One or more space characters between an option and its value is mandatory;
179 options can't be concatenated.
180 For historical reasons, the first parameter can
181 .I not
182 be a switch but must be the font name.
183 .TP
184 .BI -c \ caps-height-factor
185 The height of small caps made with the
186 .B -V
187 switch.
188 Default value of this real number is\ 0.8 times the height of uppercase
189 glyphs.
190 Will be ignored in subfont mode.
191 .TP
192 .BI -e \ extension-factor
193 The extension factor to stretch the characters horizontally.
194 Default value of this real number is\ 1.0; if less than\ 1.0, you get a
195 condensed font.
196 .TP
197 .BI -E \ encoding-id
198 The TrueType encoding ID.
199 Default value of this non-negative integer is\ 1.
200 Will be ignored if
201 .B -N
202 is used.
203 .TP
204 .BI -f \ font-index
205 The font index in a TrueType Collection.
206 Default is the first font (index\ 0).
207 Will be ignored for ordinary TrueType fonts.
208 [TrueType collections are usually found in some CJK fonts; e.g.\ the first
209 font index specifies glyphs and metrics for horizontal writing, and the
210 second font index does the same for vertical writing.
211 TrueType collections usually have the extension `\c
212 .C \&.ttc '.]
213 .TP
214 .B -l
215 Create ligatures in subfonts between first and second bytes of all the
216 original character codes.
217 Example: Character code\ 0xABCD maps to character position\ 123 in
218 subfont\ 45.
219 Then a ligature in subfont\ 45 between position 0xAB and\ 0xCD pointing
220 to character\ 123 will be produced.
221 The fonts of the Korean H\*(LX
222 package use this feature.
223 Will be ignored if not in subfont mode.
224 .TP
225 .B -n
226 Use PS names (of glyphs) of the TrueType font.
227 Will be ignored in subfont mode.
228 Only glyphs with a valid entry in the selected cmap are used.
229 .TP
230 .B -N
231 Use only PS names of the TrueType font.
232 No cmap is used, thus the switches
233 .B -E
234 and
235 .B -P
236 have no effect, causing a warning message.
237 Will be ignored in subfont mode.
238 .TP
239 .B -O
240 Use octal values for all character codes in the
241 .C VPL
242 file rather than names; this is useful for symbol or CJK fonts where
243 character names such as `A' are meaningless.
244 .TP
245 .BI -p \ inencfile
246 The input encoding file name for the TTF\(->raw\ \*(TX
247 mapping.
248 This parameter has to be specified in
249 .C \%ttfonts.map
250 for successive
251 .B ttf2pk
252 calls.
253 Will be ignored in subfont mode.
254 .TP
255 .BI -P \ platform-id
256 The TrueType platform ID.
257 Default value of this non-negative integer is\ 3.
258 Will be ignored if
259 .B -N
260 is used.
261 .TP
262 .B -q
263 Make
264 .B ttf2tfm
265 quiet.
266 It suppresses any informational output except warning and error
267 messages.
268 For CJK fonts, the output can get quite large if you don't specify
269 this switch.
270 .TP
271 .BI -r \ old-glyphname\ new-glyphname
272 Replaces
273 .I \%old-glyphname
274 with
275 .IR \%new-glyphname .
276 This switch is useful if you want to give an unnamed glyph (i.e., a glyph
277 which can be represented with `.gXXX' or `.cXXX' only) a name or if you want
278 to rename an already existing glyph name.
279 You can't use the `.gXXX' or `.cXXX' glyph name constructs for
280 .IR \%new-glyphname ;
281 multiple occurrences of
282 .B -r
283 are possible.
284 If in subfont mode or if no encoding file is specified, this switch is
285 ignored.
286 .TP
287 .BI -R \ replacement-file
288 Use this switch if you have many replacement pairs; they can be collected
289 in a file which should have `\c
290 .C \&.rpl '
291 as extension.
292 The syntax used in such replacement files is simple: Each non-empty
293 line must contain a pair `\c
294 .IR "\%old-glyphname \%new-glyphname" '
295 separated by whitespace (without the quotation marks).
296 A percent sign starts a line comment; you can continue a line on the next
297 line with a backslash as the last character.
298 If in subfont mode or if no encoding file is specified, this switch is
299 ignored.
300 .TP
301 .BI -s \ slant-factor
302 The obliqueness factor to slant the font, usually much smaller than\ 1.
303 Default of this real number is\ 0.0; if the value is larger than zero,
304 the characters slope to the right, otherwise to the left.
305 .TP
306 .BI -t \ outencfile
307 The output encoding file name for the virtual font(s).
308 Only characters in the raw \*(TX
309 font are used.
310 Will be ignored in subfont mode.
311 .TP
312 .BI -T \ inoutencfile
313 This is equivalent to
314 .RB ` -p
315 .I inoutencfile
316 .B -t
317 .IR inoutencfile '.
318 Will be ignored in subfont mode.
319 .TP
320 .B -u
321 Use only those characters specified in the output encoding, and no
322 others.
323 By default,
324 .B ttf2tfm
325 tries to include all characters in the virtual font, even those not
326 present in the encoding for the virtual font (it puts them into
327 otherwise-unused positions, rather arbitrarily).
328 Will be ignored in subfont mode.
329 .TP
330 .BI -v \ vplfile
331 Output a
332 .C VPL
333 file in addition to the
334 .C TFM
335 file.
336 If no output encoding file is specified,
337 .B ttf2tfm
338 uses a default font encoding (cmtt10).
339 Will be ignored in subfont mode.
340 .TP
341 .BI -V \ scvplfile
342 Same as
343 .BR -v ,
344 but the virtual font generated is a pseudo small caps font obtained by
345 scaling uppercase letters by\ 0.8 (resp. the value specified with
346 .BR -c )
347 to typeset lowercase.
348 This font handles accented letters and retains proper kerning.
349 Will be ignored in subfont mode.
350 .TP
351 .B -x
352 Rotate all glyphs by 90 degrees counter-clockwise.
353 If no
354 .B -y
355 parameter is given, the rotated glyphs are shifted down vertically
356 by\ 0.25em.
357 Will be ignored if not in subfont mode.
358 .TP
359 .BI -y \ vertical-shift-factor
360 Shift down rotated glyphs by the given amount (the unit is
361 .IR em ).
362 Ignored if not in subfont mode or glyphs are not rotated.
363 .TP
364 .B --version
365 Shows the current version of
366 .B ttf2tfm
367 and the used file search library (e.g.
368 .BR kpathsea ).
369 .TP
370 .B --help
371 Shows usage information.
372 .PP
373 If no
374 .C TFM
375 file name is given, the name of the
376 .C TTF
377 file is used, including the full path and replacing the extension with `\c
378 .C \&.tfm '.
379 .
380 .
381 .SH CMAPS
382 Contrary to Type\ 1 PostScript fonts (but similar to the new CID
383 PostScript font format), most TrueType fonts have more than one native
384 mapping table, also called `cmap', which maps the (internal) TTF glyph
385 indices to the (external) TTF character codes.
386 Common examples are a mapping table to Unicode encoded character
387 positions, and the standard Macintosh mapping.
389 To specify a TrueType mapping table, use the options
390 .B -P
391 and
392 .BR -E .
393 With
394 .B -P
395 you specify the platform ID; defined values are:
396 .PP
397 .in +4m
398 .ta 3iC
399 .I "platform platform ID (pid)"
400 .sp
401 .ta 3iR
402 Apple Unicode 0
403 .br
404 Macintosh 1
405 .br
406 ISO 2
407 .br
408 Microsoft 3
409 .PP
410 The encoding ID depends on the platform.
411 For pid=0, we ignore the
412 .B -E
413 parameter (setting it to zero) since the mapping table is always
414 Unicode version\ 2.0.
415 For pid=1, the following table lists the defined values:
416 .PP
417 .in +4m
418 .ta 3iC
419 .ti -2m
420 platform ID = 1
421 .sp
422 .I "script encoding ID (eid)"
423 .sp
424 .ta 3iR
425 Roman 0
426 .br
427 Japanese 1
428 .br
429 Chinese 2
430 .br
431 Korean 3
432 .br
433 Arabic 4
434 .br
435 Hebrew 5
436 .br
437 Greek 6
438 .br
439 Russian 7
440 .br
441 Roman Symbol 8
442 .br
443 Devanagari 9
444 .br
445 Gurmukhi 10
446 .br
447 Gujarati 11
448 .br
449 Oriya 12
450 .br
451 Bengali 13
452 .br
453 Tamil 14
454 .br
455 Telugu 15
456 .br
457 Kannada 16
458 .br
459 Malayalam 17
460 .br
461 Sinhalese 18
462 .br
463 Burmese 19
464 .br
465 Khmer 20
466 .br
467 Thai 21
468 .br
469 Laotian 22
470 .br
471 Georgian 23
472 .br
473 Armenian 24
474 .br
475 Maldivian 25
476 .br
477 Tibetan 26
478 .br
479 Mongolian 27
480 .br
481 Geez 28
482 .br
483 Slavic 29
484 .br
485 Vietnamese 30
486 .br
487 Sindhi 31
488 .br
489 Uninterpreted 32
490 .PP
491 Here are the ISO encoding IDs:
492 .PP
493 .in +4m
494 .ta 3iC
495 .ti -2m
496 platform ID = 2
497 .sp
498 .I "encoding encoding ID (eid)"
499 .sp
500 .ta 3iR
501 ASCII 0
502 .br
503 ISO 10646 1
504 .br
505 ISO 8859-1 2
506 .PP
507 And finally, the Microsoft encoding IDs:
508 .PP
509 .in +4m
510 .ta 3iC
511 .ti -2m
512 platform ID = 3
513 .sp
514 .I "encoding encoding ID (eid)"
515 .sp
516 .ta 3iR
517 Symbol 0
518 .br
519 Unicode 2.0 1
520 .br
521 Shift JIS 2
522 .br
523 GB 2312 (1980) 3
524 .br
525 Big 5 4
526 .br
527 KSC 5601 (Wansung) 5
528 .br
529 KSC 5601 (Johab) 6
530 .PP
531 The program will abort if you specify an invalid platform/encoding ID
532 pair.
533 It will then show the possible pid/eid pairs.
534 Please note that most fonts have at most two or three cmaps, usually
535 corresponding to the pid/eid pairs (1,0), (3,0), or (3,1) in case of
536 Latin based fonts.
537 Valid Microsoft fonts should have a (3,1) mapping table, but some
538 fonts exist (mostly Asian fonts) which have a (3,1) cmap not encoded
539 in Unicode.
540 The reason for this strange behavior is the fact that some MS\ Windows
541 versions will reject fonts having a non-(3,1) cmap (since all non-Unicode
542 Microsoft encoding IDs are for Asian MS\ Windows versions).
543 .PP
544 The
545 .B -P
546 and
547 .B -E
548 options of
549 .B ttf2tfm
550 must be equally specified for
551 .BR ttf2pk ;
552 the corresponding parameters in
553 .C \%ttfonts.map
554 are `Pid' and `Eid', respectively.
555 .PP
556 The default pid/eid pair is (3,1).
557 .PP
558 Similarly, an
559 .B -f
560 option must be specified as `Fontindex' parameter in
561 .C \%ttfonts.map .
562 .PP
563 If you use the
564 .B -N
565 switch, all cmaps are ignored, using only the PostScript names in the
566 TrueType font.
567 The corresponding option in
568 .C \%ttfonts.map
569 is \%`PS=Only'.
570 If you use the
571 .B -n
572 switch, the default glyph names built into
573 .B ttf2tfm
574 are replaced with the PS glyph names found in the font.
575 In many cases this is not what you want because the glyph names in the
576 font are often incorrect or non-standard.
577 The corresponding option in
578 .C \%ttfonts.map
579 is \%`PS=Yes'.
580 .PP
581 Single replacement glyph names specified with
582 .B -r
583 must be given directly as `\c
584 .IR old-glyphname = new-glyphname ';
585 .B -R
586 is equivalent to the `Replacement' option.
587 .
588 .
589 .SH INPUT AND OUTPUT ENCODINGS
590 You must specify the encoding vectors from the TrueType font to the
591 raw \*(TX
592 font and from the raw \*(TX
593 font to the virtual \*(TX
594 font exactly as with
595 .BR afm2tfm ,
596 but you have more possibilities to address the character codes.
597 [With `encoding vector' a mapping table with 256\ entries in form of a
598 PostScript vector is meant; see the file
599 .C \%T1-WGL4.enc
600 of this package for an example.]
601 With
602 .BR afm2tfm ,
603 you must access each glyph with its Adobe glyph name, e.g.\ \c
604 \%`/quotedsingle' or \%`/Acircumflex'.
605 This has been extended with
606 .BR ttf2tfm ;
607 now you can (and sometimes must) access the code points and/or glyphs
608 directly, using the following syntax for specifying the character position
609 in decimal, octal, or hexadecimal notation:
610 `/.c\c
611 .IR <decimal-number> ',
612 `/.c0\c
613 .IR <octal-number> ',
614 or `/.c0x\c
615 .IR <hexadecimal-number> '.
616 Examples: \%`/.c72', \%`/.c0646', \%`/.c0x48'.
617 To access a glyph index directly, using the character `g' instead of
618 `c' in the just introduced notation.
619 Example: \%`/.g0x32'.
620 [Note: The `.cXXX' notation makes no sense if
621 .B -N
622 is used.]
623 .PP
624 For pid/eid pairs (1,0) and (3,1), both
625 .B ttf2tfm
626 and
627 .B ttf2pk
628 recognize built-in default Adobe glyph names; the former follows the names
629 given in Appendix\ E of the book `Inside Macintosh', volume\ 6, the latter
630 uses the names given in the TrueType Specification (WGL4, a Unicode subset).
631 Note that Adobe names for a given glyph are often not unique and do
632 sometimes differ, e.g., many PS fonts have the glyph `mu', whereas this
633 glyph is called `mu1' in the WGL4 character set to distinguish it from the
634 real Greek letter mu.
635 Be also aware that OpenType (i.e. TrueType\ 2.0) fonts use an updated WGL4
636 table; we use the data from the latest published TrueType specification
637 (1.66).
638 You can find those mapping tables in the source code file
639 .C \%ttfenc.c .
640 .PP
641 On the other hand, the switches
642 .B -n
643 and
644 .B -N
645 makes
646 .B ttf2tfm
647 read in and use the PostScript names in the TrueType font itself (stored
648 in the `post' table) instead of the default Adobe glyph names.
649 .PP
650 Use the
651 .B -r
652 switch to remap single glyph names and
653 .B -R
654 to specify a file containing replacement glyph name pairs.
655 .PP
656 If you don't select an input encoding, the first 256\ glyphs of the
657 TrueType font with a valid entry in the selected cmap will be mapped
658 to the \*(TX
659 raw font (without the
660 .B -q
661 option,
662 .B ttf2tfm
663 prints this mapping table to standard output), followed by all glyphs
664 not yet addressed in the selected cmap.
665 However, some code points for the (1,0) pid/eid pair are omitted since
666 they do not represent glyphs useful for \*(TX:
667 0x00 (null), 0x08 (backspace), 0x09 (horizontal tabulation), 0x0d
668 (carriage return), and 0x1d (group separator).
669 The `invalid character' with glyph index\ 0 will be omitted too.
670 .PP
671 If you select the
672 .B -N
673 switch, the first 256\ glyphs of the TrueType font with a valid PostScript
674 name will be used in case no input encoding is specified.
675 Again, some glyphs are omitted: `.notdef', `.null', and
676 `nonmarkingreturn'.
677 .PP
678 If you don't select an output encoding,
679 .B ttf2tfm
680 uses the same mapping table as
681 .B afm2tfm
682 would use (you can find it in the source code file
683 .C \%texenc.c );
684 it corresponds to \*(TX
685 typewriter text.
686 Unused positions (either caused by empty code points in the mapping
687 table or missing glyphs in the TrueType font) will be filled (rather
688 arbitrarily) with characters present in the input encoding but not
689 specified in the output encoding (without the
690 .B -q
691 option
692 .B ttf2tfm
693 prints the final output encoding to standard output).
694 Use the
695 .B -u
696 option if you want only glyphs in the virtual font which are defined
697 in the output encoding file, and nothing more.
698 .PP
699 One feature missing in
700 .B afm2tfm
701 has been added which is needed by the T1 encoding:
702 .B ttf2tfm
703 will construct the glyph `Germandbls' (by simply concatenating to `S'
704 glyphs) even for normal fonts if possible.
705 It appears in the glyph list as the last item, marked with an asterisk.
706 Since this isn't a real glyph it will be available only in the virtual
707 font.
708 .PP
709 For both input and output encoding, an empty code position is
710 represented by the glyph name \%`/.notdef'.
711 .PP
712 In encoding files, you can use `\\' as the final character of a line to
713 indicate that the input is continued on the next line.
714 The backslash and the following newline character will be removed.
715 .PP
716 .
717 .
718 .SH SUBFONT DEFINITION FILES
719 CJKV (Chinese/Japanese/Korean/old Vietnamese) fonts usually contain several
720 thousand glyphs; to use them with \*(TX
721 it is necessary to split such large fonts into subfonts.
722 Subfont definition files (usually having the extension `\c
723 .C \&.sfd ')
724 are a simple means to do this smoothly.
725 .PP
726 A subfont file name usually consists of a prefix, a subfont infix, and
727 a postfix (which is empty in most cases), e.g.
728 .PP
729 .in +2m
730 ntukai23 \(-> prefix: ntukai, infix: 23, postfix: (empty)
731 .PP
732 Here the syntax of a line in an
733 .C SFD
734 file, describing one subfont:
735 .in +2m
736 .TP
737 .I <whitespace> <infix> <whitespace> <ranges> <whitespace>
738 .sp
739 .TP
740 .IR <infix> \ :=
741 anything except whitespace.
742 It's best to use only alphanumerical characters.
743 .TP
744 .IR <whitespace> \ :=
745 space, formfeed, carriage return, horizontal and vertical tabs -- no
746 newline characters.
747 .TP
748 .IR <ranges> \ :=
749 .IR "<ranges> <whitespace> <codepoint>" \ |
750 .br
751 .IR "<ranges> <whitespace> <range>" \ |
752 .br
753 .I <ranges> <whitespace> <offset> <whitespace> <range>
754 .TP
755 .IR <codepoint> \ :=
756 .I <number>
757 .br
758 .TP
759 .IR <range> \ :=
760 .IR <number> \ `_' \ <number>
761 .br
762 .TP
763 .IR <offset> \ :=
764 .IR <number> \ `:'
765 .TP
766 .IR <number> \ :=
767 hexadecimal (prefix `0x'), decimal, or octal (prefix `0')
768 .PP
769 A line can be continued on the next line with a backslash ending the line.
770 The ranges must not overlap; offsets have to be in the range 0-255.
771 .PP
772 Example:
773 .PP
774 .in +2m
775 The line
776 .PP
777 .in +4m
778 .C "03 10: 0x2349 0x2345_0x2347"
779 .PP
780 .in +2m
781 assigns to the code positions 10, 11, 12, and\ 13 of the subfont
782 having the infix `03' the character codes 0x2349, 0x2345, 0x2346, and
783 0x2347 respectively.
784 .PP
785 The
786 .C SFD
787 files in the distribution are customized for the CJK package for
788 \*(LX.
789 .PP
790 You have to embed the
791 .C SFD
792 file name into the
793 .C TFM
794 font name (at the place where the infix will appear) surrounded by two
795 `@' signs, on the command line resp.\ the
796 .C \%ttfonts.map
797 file; both
798 .B ttf2tfm
799 and
800 .B ttf2pk
801 switch then to subfont mode.
802 .PP
803 Subfont mode disables the options
804 .BR -n , \ -N , \ -p ,
805 .BR -r , \ -R , \ -t ,
806 .BR -T , \ -u , \ -v ,
807 and
808 .B -V
809 for
810 .BR ttf2tfm ;
811 similarly, no `Encoding' or `Replacement' parameter is allowed in
812 .C \%ttfonts.map .
813 Single replacement glyph names are ignored too.
814 .PP
815 .B ttf2tfm
816 will create all subfont
817 .C TFM
818 files specified in the
819 .C SFD
820 files (provided the subfont contains glyphs) in one run.
821 .PP
822 Example:
823 .PP
824 .in +2m
825 The call
826 .PP
827 .in +4m
828 .C "ttf2tfm ntukai.ttf ntukai@/usr/local/lib/ttf2tfm/Big5@"
829 .PP
830 .in +2m
831 will use
832 .C /usr/local/lib/ttf2tfm/Big5.sfd ,
833 producing
834 .I all
835 subfont files
836 .C ntukai01.tfm ,
837 .C ntukai02.tfm ,
838 etc.
839 .
840 .
841 .SH "RETURN VALUE"
842 ttf2tfm returns 0 on success and 1 on error; warning and error
843 messages are written to standard error.
844 .
845 .
846 .SH "SOME NOTES ON FILE SEARCHING"
847 Both
848 .B ttf2pk
849 and
850 .B ttf2tfm
851 use either the
852 .BR kpathsea ,
853 .BR emtexdir ,
854 or
855 .B MiKTeX
856 library for searching files
857 .RB ( emtexdir
858 will work only on operating systems which have an MS-DOSish background, i.e.
859 MS-DOS, OS/2, Windows;
860 .B MiKTeX
861 is specific to MS Windows).
862 .PP
863 As a last resort, both programs can be compiled without a search library;
864 the searched files must be then in the current directory or specified with a
865 path.
866 Default extensions will be appended also (with the exception that only `\c
867 .C \&.ttf '
868 is appended and not `\c
869 .C \&.ttc ').
870 .PP
871 KPATHSEA
872 .PP
873 Please note that older versions of
874 .B kpathsea
875 (<3.2) have no special means to seach for TrueType fonts and related
876 files, thus we use the paths for PostScript related stuff.
877 The actual version of kpathsea is displayed on screen if you call
878 either
879 .B ttf2pk
880 or
881 .B ttf2tfm
882 with the
883 .B --version
884 command line switch.
885 .PP
886 Here is a table of the file type and the corresponding
887 .B kpathsea
888 variables.
889 .C TTF2PKINPUTS
890 and
891 .C TTF2TFMINPUTS
892 are program specific environment variables introduced in
893 .B kpathsea
894 version\ 3.2:
895 .PP
896 .in +4m
897 .ta 2i
898 .C \&.ttf \ and "\ .ttc TTFONTS"
899 .br
900 .C "ttfonts.map TTF2PKINPUTS"
901 .br
902 .C "\&.enc TTF2PKINPUTS, TTF2TFMINPUTS"
903 .br
904 .C "\&.rpl TTF2PKINPUTS, TTF2TFMINPUTS"
905 .br
906 .C "\&.tfm TFMFONTS"
907 .br
908 .C "\&.sfd TTF2PKINPUTS, TTF2TFMINPUTS"
909 .PP
910 And here the same for pre-3.2-versions of
911 .B kpathsea:
912 .PP
913 .in +4m
914 .ta 2i
915 .C \&.ttf \ and "\ .ttc T1FONTS"
916 .br
917 .C "ttfonts.map TEXCONFIG"
918 .br
919 .C "\&.enc TEXPSHEADERS"
920 .br
921 .C "\&.rpl TEXPSHEADERS"
922 .br
923 .C "\&.tfm TFMFONTS"
924 .br
925 .C "\&.sfd TEXPSHEADERS"
926 .PP
927 Finally, the same for pre-3.0-versions (as used e.g. in te\*(TX\ 0.4):
928 .PP
929 .in +4m
930 .ta 2i
931 .C \&.ttf \ and "\ .ttc DVIPSHEADERS"
932 .br
933 .C "ttfonts.map TEXCONFIG"
934 .br
935 .C "\&.enc DVIPSHEADERS"
936 .br
937 .C "\&.rpl DVIPSHEADERS"
938 .br
939 .C "\&.tfm TFMFONTS"
940 .br
941 .C "\&.sfd DVIPSHEADERS"
942 .PP
943 Please consult the info files of
944 .B kpathsea
945 for details on these variables.
946 The decision whether to use the old or the new scheme will be done
947 during compilation.
948 .PP
949 You should set the
950 .C TEXMFCNF
951 variable to the directory where your
952 .C texmf.cnf
953 configuration file resides.
954 .PP
955 Here is the proper command to find out to which value a
956 .B kpathsea
957 variable is set (we use
958 .C TTFONTS
959 as an example).
960 This is especially useful if a variable isn't set in
961 .C texmf.cnf
962 or in the environment, thus pointing to the default value which is
963 hard-coded into the
964 .B kpathsea
965 library.
966 .PP
967 .in +2m
968 .C "kpsewhich -progname=ttf2tfm -expand-var='$TTFONTS'"
969 .PP
970 We select the program name also since it is possible to specify
971 variables which are searched only for a certain program -- in our
972 example it would be
973 .C TTFONTS.ttf2tfm .
974 .PP
975 A similar but not identical method is to say
976 .PP
977 .in +2m
978 .C "kpsewhich -progname=ttf2tfm -show-path='truetype fonts'"
979 .PP
980 [A full list of format types can be obtained by saying `\c
981 .C "kpsewhich --help" '
982 on the command line prompt.]
983 This is exactly the how
984 .B ttf2tfm
985 (and
986 .BR ttf2pk )
987 search for files; the disadvantage is that all variables are expanded
988 which can cause very long strings.
989 .PP
990 EMTEXDIR
991 .PP
992 Here the list of suffixes and its related environment variables to be
993 set in
994 .C autoexec.bat
995 (resp. in
996 .C config.sys
997 for OS/2):
998 .PP
999 .in +4m
1000 .ta 2i
1001 .C \&.ttf \ and "\ .ttc TTFONTS"
1002 .br
1003 .C "ttfonts.map TTFCFG"
1004 .br
1005 .C "\&.enc TTFCFG"
1006 .br
1007 .C "\&.rpl TTFCFG"
1008 .br
1009 .C "\&.tfm TEXTFM"
1010 .br
1011 .C "\&.sfd TTFCFG"
1012 .PP
1013 If one of the variables isn't set, a warning message is emitted.
1014 The current directory will always be searched.
1015 As usual, one exclamation mark appended to a directory path causes
1016 subdirectories one level deep to be searched, two exclamation marks causes
1017 all subdirectories to be searched.
1018 Example:
1019 .PP
1020 .in +2m
1021 .C TTFONTS=c:\\\\fonts\\\\truetype!!;d:\\\\myfonts\\\\truetype!
1022 .PP
1023 Constructions like `\c
1024 .C c:\\\\fonts!!\\\\truetype '
1025 aren't possible.
1026 .PP
1027 MIKTEX
1028 .PP
1029 Both
1030 .B ttf2tfm
1031 and
1032 .B ttf2pk
1033 have been fully integrated into
1034 .BR MiKTeX .
1035 Please refer to the documentation of
1036 .B MiKTeX
1037 for more details on file searching.
1038 .
1039 .
1040 .SH PROBLEMS
1041 Most
1042 .B vptovf
1043 implementations allow only 100\ bytes for the
1044 .C TFM
1045 header (the limit is 1024 in the
1046 .C TFM
1047 file itself): 8\ bytes for checksum and design size, 40\ bytes for the
1048 family name, 20\ bytes for the encoding, and 4\ bytes for a face byte.
1049 There remain only 28\ bytes for some additional information which is used by
1050 .B ttf2tfm
1051 for an identification string (which is essentially a copy of the command
1052 line), and this limit is always exceeded.
1054 The optimal solution is to increase the value of
1055 .I \%max_header_bytes
1056 in the file
1057 .C vptovf.web
1058 (and probably
1059 .C pltotf.web
1060 too) to, say,\ 400
1061 and recompile
1062 .B vptovf
1063 (and
1064 .BR pltotf ).
1065 Otherwise you'll get some (harmless) error messages like
1066 .PP
1067 .in +2m
1068 .C "This HEADER index is too big for my present table size"
1069 .PP
1070 which can be safely ignored.
1071 .
1072 .
1073 .SH "SEE ALSO"
1074 .BR ttf2pk (1),
1075 .BR afm2tfm (1),
1076 .BR vptovf (1),
1077 .br
1078 the info pages for
1079 .B dvips
1080 and
1081 .B kpathsea
1082 .
1083 .
1084 .SH AVAILABILITY
1085 .B ttf2tfm
1086 is part of the FreeType package, a high quality TrueType rendering
1087 library.
1088 .
1089 .
1090 .SH AUTHORS
1091 Werner LEMBERG
1092 .C <wl@gnu.org>
1093 .br
1094 Fr\('ed\('eric LOYER
1095 .C <loyer@ensta.fr>