Copyright (C) 1990, 1995, 1996, 1997 Aladdin Enterprises. All rights reserved. This file is part of Aladdin Ghostscript. Aladdin Ghostscript is distributed with NO WARRANTY OF ANY KIND. No author or distributor accepts any responsibility for the consequences of using it, or for whether it serves any particular purpose or works at all, unless he or she says so in writing. Refer to the Aladdin Ghostscript Free Public License (the "License") for full details. Every copy of Aladdin Ghostscript must include a copy of the License, normally in a plain ASCII text file named PUBLIC. The License grants you the right to copy, modify and redistribute Aladdin Ghostscript, but only under certain conditions described in the License. Among other things, the License requires that the copyright notice and this notice be preserved on all copies. - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - This file, fonts.txt, describes the fonts and font facilities supplied with Ghostscript. For an overview of Ghostscript and a list of the documentation files, see README. About Ghostscript fonts ======================= Most of the font files supplied with Ghostscript have a .pfb extension; a few have a .pfa or .gsf extension. Each file defines one (transformable) font specified in outline form. They are ordinary Type 1 PostScript outline fonts, and can be given to any PostScript language interpreter. Files with .pfa or .pfb extension are also compatible with Adobe Type Manager and with tools that don't include a full PostScript language interpreter; files with .gsf extension are not compatible with ATM or other tools. Starting with release 4.0, Ghostscript configurations compiled with the `ttfont' option can also use TrueType fonts (.ttf), but since Ghostscript's TrueType rasterizer currently ignores the `hints', the output is not of very good quality. The only other font-related file that Ghostscript needs for proper operation is a file called Fontmap. This file maps font names (such as /Times-Roman) to font file names (such as n021003l.pfb) or aliases (such as NimbusNo9L-Regu). The free fonts supplied with Ghostscript fall into three groups: - 35 basic PostScript fonts (Times, Helvetica, Courier, Symbol, etc.) These are commercial-quality Type 1 fonts. See the file `Fontmap' for the complete list. - Fonts derived from the public domain Hershey fonts, with improvements (such as adding accented characters) by Thomas Wolff. These are quite different from traditional printer or display fonts; the file `hershey.txt' describes them in more detail. - A few miscellaneous fonts including Cyrillic and kana fonts. The 35 basic fonts are normally distributed in a file called `ghostscript-fonts-std-N.NN.tar.gz', the rest in a file called `ghostscript-fonts-other-N.NN.fonts.tar.gz'. The file gs_fonts.ps, which is loaded as part of Ghostscript initialization, arranges to load fonts on demand using the information from Fontmap. If you want to preload all of the known fonts, invoke the procedure loadallfonts This is not done by default, since the fonts occupy about 50K each and there are a lot of them. If you want to try out the fonts, prfont.ps contains code for printing a sampler. Load this program, by including it in the gs command line or by invoking (prfont.ps) run and then to produce a sampler of a particular font, invoke /fontName DoFont e.g. /Times-Roman DoFont About the Kanji fonts --------------------- Mr. Tetsurou Tanaka of the Department of Engineering, University of Tokyo, has created a set of free Kanji fonts that is freely available on the Internet for anonymous FTP from moe.ipl.t.u-tokyo.ac.jp:/Font and is distributed with Ghostscript. Anyone can use these fonts as they are or with some format translation, and redistribute them without reporting. If you redistribute them, you must inform the recipient that he can get the original from the abovementioned FTP site, so that he could get the newest version later. The fonts include a README file in Japanese. Here is some English documentation supplied by Mr. Kiyotaka Sakai, also of the University of Tokyo. The following four fonts are different styles of the same characters (JISX208). These fonts also needs wadalab-sym.tar.Z in common. wadalab-mincho-0-8.tar.gz wadalab-mincho-0-12.6.tar.gz wadalab-gothic-0-13.5.tar.gz wadalab-maru-0-8.4.tar.gz And the following two fonts are the other different styles of the same characters(JISX0212). wadalab-mincho-1-8.tar.gz wadalab-maru-1-8.tar.gz These fonts are postscript(type1) fonts. You can convert them to Metafont, Type1(PFA) font, TeX font(for japanese TeX), BDF font, Shotaikurabu font by using wftomf.c, wftopfa.c, wftodm.c, bdfmerge.c wftovf.c in tools directory. In order to conform to MS-DOS naming restrictions, we have renamed the original font files as follows: Name Original name ---- ------------- got013-5.taz wadalab-gothic-0-13.5.tar.gz mar08-4.taz wadalab-maru-0-8.4.tar.gz mar18.taz wadalab-maru-1-8.tar.gz min012-6.taz wadalab-mincho-0-12.6.tar.gz min08-4.taz wadalab-mincho-0-8.4.tar.gz min18.taz wadalab-mincho-1-8.tar.gz sym-4.taz wadalab-sym.4.tar.gz Platform fonts ============== Starting with release 2.6.1, Ghostscript uses whatever font technology is provided by the system on which it runs, by using the system's API for displaying text. On MS Windows this may be TrueType, or it may be ATM; Ghostscript neither knows nor cares. The PostScript language specifies that fonts are data structures with particular contents (e.g., they include a bounding box for the font, an Encoding vector for specifying the character set, etc.), and it is fairly common for PostScript files to make use of this fact; also, characters can be used as clipping regions, and can be arbitrarily rotated, skewed, expanded/condensed, etc. algorithmically. Most of this information is available in one form or another from the underlying graphics system, but one crucial piece is not: the actual scalable outlines of the characters, which Ghostscript needs in order to implement clipping with character shapes and to implement arbitrarily transformed characters. Consequently, Ghostscript needs the scalable outlines of any font mentioned in a document, and will load them from the disk (.PFA, .PFB, or .GSF file) in the usual way, even if it uses the platform's font machinery for displaying the characters. To make matters worse, platforms use different names for their standard fonts. For example, the Times Roman font, for which PostScript files use the name "Times-Roman", may be known as "Times-Roman", "Times Roman", "Tms Rmn", "Times New Roman", or "TimesNewRoman". The name may even be completely different: the usual Helvetica-equivalent TrueType font is called "Arial". Now, it is possible to cope with this situation by introducing aliases in Fontmap, but there are two reasons why the current Ghostscript release does not do this: 1) Naming in different systems is so unstandardized that there does not seem to be a small set of alternative names that is likely to cover most of the situations. All 5 of the above names for Times Roman have been seen in Windows and OS/2 environments, depending on system version, TrueType vs. ATM, and other unknown factors. 2) Each alias takes up a substantial amount of space (several hundred bytes) at run time. If each of the standard 35 fonts has 3 additional aliases, this might amount to 50K of wasted space. This is a lot on a PC, although running under Windows in enhanced mode, it might not be a problem. If you don't seem to be getting nice characters on the screen under MS Windows, you can try adding aliases to the Fontmap, according to the documentation found there. Adding your own fonts ===================== Ghostscript can use any Type 0, Type 1, Type 3, Type 4, or Type 42 font that is acceptable to other PostScript language interpreters or to ATM, including MultiMaster fonts. Starting with release 4.0, Ghostscript can also use TrueType fonts (if Ghostscript was compiled with the `ttfont' option). Ghostscript also provides a way to construct a (low-quality) Type 1 font from a bitmap font in BDF format, which is a popular format in the Unix world. If you want to add fonts of your own, you must edit Fontmap to include an entry for your new font at the end. The format for entries is documented in the Fontmap file. Since later entries in Fontmap override earlier entries, any fonts you add will supersede the corresponding fonts supplied with Ghostscript. In the PC world, Type 1 fonts are customarily given names ending in .PFA or .PFB. Ghostscript can use these directly; you just need to make the entry in Fontmap. If you are going to use a commercial Type 1 font (such as fonts obtained in conjunction with Adobe Type Manager) with Ghostscript, please read carefully the license that accompanies the font; Aladdin Enterprises takes no responsibility for any possible violations of such licenses. The same applies to TrueType fonts, whose names customarily end in .TTF. Converting BDF fonts -------------------- If you want to convert a BDF file to a scalable outline, use the program bdftops.ps (and invoking shell script bdftops.bat or bdftops). Run the shell command bdftops [ ...] [] [] e.g., bdftops pzdr.bdf ZapfDingbats.afm pzdr.gsf ZapfDingbats 4100000 1000000.1.41 (Obviously, you would enter this all on one line; the example is split so it will fit on the page.) Then make an entry for the .gsf file in Fontmap as described above. See the end of this document for more detail. For developers only =================== The rest of this document is very unlikely to be of value to ordinary users. Contents of fonts ----------------- As noted above, Ghostscript accepts fonts in the same formats as PostScript interpreters. Type 0, 1, and 3 fonts are documented in the PostScript Language Reference Manual (Second Edition); detailed documentation for Type 1 fonts appears in a separate Adobe book. Type 42 (encapsulated TrueType) fonts are documented in an Adobe supplement; the TrueType format is documented in publications available from Apple and Microsoft. Type 4 fonts are not documented anywhere; they are essentially Type 1 fonts with a BuildChar or BuildGlyph procedure. Precompiling fonts ------------------ You can compile any Type 1 font into C and link it into the Ghostscript executable. (Type 1 fonts include any font whose name ends with .pfa or .pfb, and it also includes all the Ghostscript .gsf fonts except for the Hershey fonts.) This doesn't have any effect on rendering speed, but it eliminates the time for loading the font dynamically, which may make a big difference in total rendering time, especially for multi-page documents. Fonts that have been precompiled and linked in this way do not need to appear in the Fontmap, although if they do appear there, no harm is done. The utility for precompiling fonts is called font2c. Note that font2c is a PostScript language program, so you must have Ghostscript already running to be able to run font2c; you must also have entries in the Fontmap for the fonts you want to compile. For example, to precompile the Times-Italic font, font2c Times-Italic ptmri.c where the first argument is the font name and the second is the name of the .c file. You can use any file name you want, as long as it ends in .c. It doesn't have to be limited to 8 characters, unless your operating system requires this. We suggest that you use names xxxx.c, where xxxx.gsf or xxxx.pfa is the name of the font file in the Fontmap file, just so you don't have to keep track of another set of names. (If you are running on a VMS platform, or another platform where the C compiler has a limit on the length of identifiers, you must do something slightly more complicated; see the section 'Platforms with identifier length limits' below. Also, on VMS, you must put quotes "" around the font name so that the VMS command processor doesn't convert the name to lower case.) For VMS environments, see the directions in the file make.txt, and ignore the rest of this section. Note that ncrr.c, ptmr.c, etc. are not supplied with the Ghostscript fileset, since they are quite large and can easily be recreated using the font2c program as described above. There is a makefile called cfonts.mak that will run font2c on all the fonts supplied with Ghostscript. Invoke it with make -f cfonts.mak On some systems, you may have to omit the space following the -f, i.e., make -fcfonts.mak Besides running font2c, you must arrange things so that the fonts will be compiled, and linked into the executable. To do this, add the compiled fonts "feature" to your platform-specific makefile. On MS-DOS systems, you edit tc.mak, bc.mak, bcwin.mak, msc.mak, or watc.mak; on Unix systems, you edit ansihead.mak, cc-head.mak, or gcc-head.mak, and then execute tar_cat. Find the definition of the FEATURE_DEVS macro in the makefile, e.g., FEATURE_DEVS=level2.dev pdf.dev and add ccfonts.dev, e.g., FEATURE_DEVS=level2.dev pdf.dev ccfonts.dev Next, you must add the specific fonts to the generic makefile. On MS-DOS systems, you edit gs.mak; on Unix systems, you edit makefile. The makefile already has rules for the standard 35 fonts supplied with Ghostscript. Find the line in the relevant makefile that says #ccfonts1_= Remove the # mark from this line, and from all the following lines down to and including the one that says #ccfonts9= This is all you need to do for the standard fonts. The next couple of paragraphs describe how to compile in other fonts, such as the Utopia or Kana fonts or your own fonts. Suppose you want to compile the Kana fonts into the executable. First, pick one of ccfonts10 through 15 as the place you will do this, say ccfonts10. Add your compiled font file names, e.g., ccfonts10_=fhirw.$(OBJ) If the line gets too long, use another line of the same form, e.g., ccfonts11_=fkarw.$(OBJ) Just below this, you will find a line that says ccfonts10_= Add your own fonts to the end of this line, e.g., ccfonts10=Calligraphic_Hiragana Notice that you must replace `-' by `_' in the font name. Again, if the line gets too long, add another line of the same form, e.g., ccfonts10=Calligraphic_Hiragana ccfonts11=Calligraphic_Katakana After all the lines of this form, add a pair of lines to compile each font, separating these entries from the ccfonts* lines and from each other by a blank line. In our example: fhirw.$(OBJ): fhirw.c $(CCFONT) $(CCCF) fhirw.c fkarw.$(OBJ): fkarw.c $(CCFONT) $(CCCF) fkarw.c Finally, run `make'. The executable will now include the fonts you added. They will be present in FontDirectory when Ghostscript starts up. Precompiling fonts on platforms with identifier length limits ------------------------------------------------------------- On some platforms, the C compiler and/or linker have a limit on the number of significant characters in an identifier. On such platforms, you must do a little extra work. Let N be the maximum number of significant characters in an identifier (typically 31). For each font whose name is longer than N-5 characters, pick an arbitrary identifier that we will call the "short name". This can be any string you want, as long as it contains only letters, digits, and underscores; is no longer than N-5 characters; and is not the same as any other font name or short name. A good choice for this would be to use the name of the C file. (There is no harm in doing this for fonts with names shorter than N-5 characters, it's just not necessary.) You must do two different things for fonts that require a short name. First, you must supply the short name as a third argument to the font2c program. For example, to compile NewCenturySchlbk-BoldItalic using the short name "pncbi", font2c NewCenturySchlbk-BoldItalic pncbi.c pncbi Then when you add the font to the gsaddmod line in the makefile, use the short name, not the actual font name, e.g., ccfonts2=pncbi instead of ccfonts2=NewCenturySchlbk_BoldItalic Everything else is as described above. This procedure doesn't change the name of the font in the Fontmap, or as seen from within Ghostscript; it's just a workaround for a limitation of some older compilers. Font names and unique IDs ------------------------- If you create your own fonts, and are only going to use them within your own organization, you should use UniqueID values between 4000000 and 4999999 as described just below; if you are going to distribute fonts, call Adobe and get them to assign you some UniqueIDs and also an XUID for your organization. The current (September 1993) UniqueID Coordinator is Terry O'Donnell; he is very helpful and will probably be able to assign you the numbers over the phone. The suggested temporary UniqueID for a font looks like: 4TTWVE0 where TT is a two-digit number representing the typeface, W represents the weight (normal, bold, ...), V represents the variant (normal, italic, oblique, ...), and E represents the expansion (normal, condensed, ...). This scheme will not work forever. As soon there are more 99 typefaces, or more than 9 weights or variants, we will have to do something else. But it suffices for the near future. XUIDs are a Level 2 PostScript feature that serves the same function as UniqueIDs, but is not limited to a single 24-bit integer. The bdftops program creates XUIDs of the form [-X- 0 -U-] where -X- is the organization XUID and -U- is the UniqueID. (Aladdin Enterprises' organization XUID is 107; do not use this for your own fonts that you distribute.) Ghostscript fonts are stored on files, and the file names must comply with the 8-character limit imposed by MS-DOS and other operating systems. We therefore construct the filename for a font in a way somewhat similar to the construction for temporary UniqueIDs: FTTWVVVE.gsf where F is the foundry, TT a two-letter abbreviation for the typeface, and W, V, and E the weight, variant, and expansion. Since a font can have multiple variants, we allocate three letters to that (for example, Lucida Regular Sans Typewriter Italic). If a font has four variants, you're on your own. If a font does have multiple variants, it's best to add the expansion letter `r', so that it is clear which letters are variants and which the expansion. This scheme is very close to the one proposed in `Filenames for fonts', published in the first 1990 issue of TUGboat (the journal of the TeX Users Group). In the following tables, we made no attempt to be exhaustive. Instead, we have simply allocated entries for those things that we needed for the fonts that we are actually distributing. foundries: ---------- b = Bitstream f = freely distributable/public domain fonts n = IBM p = Adobe (`p' for PostScript) u = URW[++] typefaces: id name filename prefix ---------------------------------------------------------------- 08 = Avant Garde = pag (Adobe) 11 = Bookman = pbk (Adobe) 01 = CharterBT = bch (Bitstream) 02 = Courier = ncr (IBM) 03 = Helvetica = phv (Adobe) 04 = New Century Schoolbook = pnc (Adobe) 09 = Palatino = ppl (Adobe) 05 = Symbol = psy (Adobe) 06 = Times = ptm (Adobe) -- Utopia = put (Adobe) 07 = Zapf Chancery = zc (public domain) 10 = Zapf Dingbats = pzd (Adobe) 12 = public domain Cyrillic = fcy (public domain) 13 = Kevin Hartig Hiragana = fhi (shareware) 14 = Kevin Hartig Katakana = fka (shareware) 90 = Hershey Gothic English = hrge 91 = Hershey Gothic Italian = hrit 92 = Hershey Gothic German = hrgr 93 = Hershey Greek = hrgk 94 = Hershey Plain = hrpl 95 = Hershey Script = hrsc 96 = Hershey Symbol = hrsy weights: -------- 0 = normal = r 1 = bold = b 2 = book = k 3 = demi = d 4 = light = l variants: --------- 0 = normal = r (omitted when the weight is normal) 1 = italic = i 2 = oblique = o 3 = script/handwritten/swash = w expansions: ----------- 0 = normal = r (omitted when the weight and variant are normal) 1 = narrow = n Converting BDF fonts to .gsf (Type 1) fonts ------------------------------------------- The bdftops conversion program takes the following arguments: bdftops xx.bdf [yy1.afm ...] zz.gsf fontname UniqueID [XUID] [encodingname] These arguments have the following meanings: xx.bdf - the input bitmap file, a BDF file yy*.afm - the AFM files giving the metrics (optional) zz.gsf - the output file fontname - the font name UniqueID - the UniqueID, as described above XUID - the XUID, in the form n1.n2.n3... (optional) encodingname - the encoding for the font (optional) Currently, the defined encodings are StandardEncoding, ISOLatin1Encoding, SymbolEncoding, and DingbatsEncoding. If the encoding is omitted, StandardEncoding is assumed.