ref: a630e17bd7d77df2b88ee3995e35ed1b88f00830
dir: /sys/lib/ghostscript/bdftops.ps/
% Copyright (C) 1990, 1995, 1996 Aladdin Enterprises. All rights reserved.
%
% This software is provided AS-IS with no warranty, either express or
% implied.
%
% This software is distributed under license and may not be copied,
% modified or distributed except as expressly authorized under the terms
% of the license contained in the file LICENSE in this distribution.
%
% For more information about licensing, please refer to
% http://www.ghostscript.com/licensing/. For information on
% commercial licensing, go to http://www.artifex.com/licensing/ or
% contact Artifex Software, Inc., 101 Lucas Valley Road #110,
% San Rafael, CA 94903, U.S.A., +1(415)492-9861.
% $Id: bdftops.ps,v 1.7 2003/08/06 17:05:09 alexcher Exp $
% bdftops.ps
% Convert a BDF file (possibly with (an) associated AFM file(s))
% to a PostScript Type 1 font (without eexec encryption).
% The resulting font will work with any PostScript language interpreter,
% but not with ATM or other font rasterizers lacking a complete interpreter.
/envBDF 120 dict def
envBDF begin
% "Import" the image-to-path package.
% This also brings in the Type 1 opcodes (type1ops.ps).
(impath.ps) runlibfile
% "Import" the font-writing package.
(wrfont.ps) runlibfile
wrfont_dict begin
/binary_CharStrings false def
/binary_tokens false def
/encrypt_CharStrings true def
/standard_only true def
end
/lenIV 0 def
% Invert the StandardEncoding vector.
256 dict dup begin
0 1 255 { dup StandardEncoding exch get exch def } for
end /StandardDecoding exch def
% Define the properties copied to FontInfo.
mark
(COPYRIGHT) /Notice
(FAMILY_NAME) /FamilyName
(FULL_NAME) /FullName
(WEIGHT_NAME) /Weight
.dicttomark /properties exch def
% Define the character sequences for synthesizing missing composite
% characters in the standard encoding.
mark
/AE [/A /E]
/OE [/O /E]
/ae [/a /e]
/ellipsis [/period /period /period]
/emdash [/hyphen /hyphen /hyphen]
/endash [/hyphen /hyphen]
/fi [/f /i]
/fl [/f /l]
/germandbls [/s /s]
/guillemotleft [/less /less]
/guillemotright [/greater /greater]
/oe [/o /e]
/quotedblbase [/comma /comma]
.dicttomark /composites exch def
% Define the procedure for synthesizing composites.
% This must not be bound.
/compose
{ exch pop
FontMatrix Private /composematrix get invertmatrix concat
0 0 moveto
dup gsave false charpath pathbbox currentpoint grestore
6 2 roll setcachedevice show
} def
% Define the CharString procedure that calls compose, with the string
% on the stack. This too must remain unbound.
/compose_proc
{ Private /compose get exec
} def
% Define aliases for missing characters similarly.
mark
/acute /quoteright
/bullet /asterisk
/cedilla /comma
/circumflex /asciicircum
/dieresis /quotedbl
/dotlessi /i
/exclamdown /exclam
/florin /f
/fraction /slash
/grave /quoteleft
/guilsinglleft /less
/guilsinglright /greater
/hungarumlaut /quotedbl
/periodcentered /asterisk
/questiondown /question
/quotedblleft /quotedbl
/quotedblright /quotedbl
/quotesinglbase /comma
/quotesingle /quoteright
/tilde /asciitilde
.dicttomark /aliases exch def
% Define overstruck characters that can be synthesized with seac.
mark
[ /Aacute /Acircumflex /Adieresis /Agrave /Aring /Atilde
/Ccedilla
/Eacute /Ecircumflex /Edieresis /Egrave
/Iacute /Icircumflex /Idieresis /Igrave
/Lslash
/Ntilde
/Oacute /Ocircumflex /Odieresis /Ograve /Otilde
/Scaron
/Uacute /Ucircumflex /Udieresis /Ugrave
/Yacute /Ydieresis
/Zcaron
/aacute /acircumflex /adieresis /agrave /aring /atilde
/ccedilla
/eacute /ecircumflex /edieresis /egrave
/iacute /icircumflex /idieresis /igrave
/lslash
/ntilde
/oacute /ocircumflex /odieresis /ograve /otilde
/scaron
/uacute /ucircumflex /udieresis /ugrave
/yacute /ydieresis
/zcaron
]
{ dup =string cvs
[ exch dup 0 1 getinterval cvn
exch dup length 1 sub 1 exch getinterval cvn
]
} forall
/cent [/c /slash]
/daggerdbl [/bar /equal]
/divide [/colon /hyphen]
/sterling [/L /hyphen]
/yen [/Y /equal]
.dicttomark /accentedchars exch def
% ------ Output utilities ------ %
/ws {psfile exch writestring} bind def
/wl {ws (\n) ws} bind def
/wt {=string cvs ws ( ) ws} bind def
% ------ BDF file parsing utilities ------ %
% Define a buffer for reading the BDF file.
/buffer 400 string def
% Read a line from the BDF file into the buffer.
% Ignore empty (zero-length) lines.
% Define /keyword as the first word on the line.
% Define /args as the remainder of the line.
% If the keyword is equal to commentword, skip the line.
% (If commentword is equal to a space, never skip.)
/nextline
{ { bdfile buffer readline not
{ (Premature EOF\n) print stop } if
dup length 0 ne { exit } if pop
}
loop
( ) search
{ /keyword exch def pop }
{ /keyword exch def () }
ifelse
/args exch def
keyword commentword eq { nextline } if
} bind def
% Get a word argument from args. We do *not* copy the string.
/warg % warg -> string
{ args ( ) search
{ exch pop exch }
{ () }
ifelse /args exch def
} bind def
% Get an integer argument from args.
/iarg % iarg -> int
{ warg cvi
} bind def
% Get a numeric argument from args.
/narg % narg -> int|real
{ warg cvr
dup dup cvi eq { cvi } if
} bind def
% Convert the remainder of args into a string.
/remarg % remarg -> string
{ args copystring
} bind def
% Get a string argument that occupies the remainder of args.
/sarg % sarg -> string
{ args (") anchorsearch
{ pop /args exch def } { pop } ifelse
args args length 1 sub get (") 0 get eq
{ args 0 args length 1 sub getinterval /args exch def } if
args copystring
} bind def
% Check that the keyword is the expected one.
/checkline % (EXPECTED-KEYWORD) checkline ->
{ dup keyword ne
{ (Expected ) print =
(Line=) print keyword print ( ) print args print (\n) print stop
} if
pop
} bind def
% Read a line and check its keyword.
/getline % (EXPECTED-KEYWORD) getline ->
{ nextline checkline
} bind def
% Find the first/last non-zero bit of a non-zero byte.
/fnzb
{ 0 { exch dup 128 ge { pop exit } { dup add exch 1 add } ifelse }
loop
} bind def
/lnzb
{ 7 { exch dup 1 and 0 ne { pop exit } { -1 bitshift exch 1 sub } ifelse }
loop
} bind def
% ------ Type 1 encoding utilities ------ %
% Parse the side bearing and width information that begins a CharString.
% Arguments: charstring. Result: sbx sby wx wy substring.
/parsesbw
{ mark exch lenIV
{ % stack: mark ... string dropcount
dup 2 index length exch sub getinterval
dup 0 get dup 32 lt { pop exit } if
dup 246 le
{ 139 sub exch 1 }
{ dup 250 le
{ 247 sub 8 bitshift 108 add 1 index 1 get add exch 2 }
{ dup 254 le
{ 251 sub 8 bitshift 108 add 1 index 1 get add neg exch 2 }
{ pop dup 1 get 128 xor 128 sub
8 bitshift 1 index 2 get add
8 bitshift 1 index 3 get add
8 bitshift 1 index 4 get add exch 5
} ifelse
} ifelse
} ifelse
} loop
counttomark 3 eq { 0 3 1 roll 0 exch } if
6 -1 roll pop
} bind def
% Find the side bearing and width information that begins a CharString.
% Arguments: charstring. Result: charstring sizethroughsbw.
/findsbw
{ dup parsesbw 4 { exch pop } repeat skipsbw
} bind def
/skipsbw % charstring sbwprefix -> sizethroughsbw
{ length 1 index length exch sub
2 copy get 12 eq { 2 } { 1 } ifelse add
} bind def
% Encode a number, and append it to a string.
% Arguments: str num. Result: newstr.
/concatnum
{ dup dup -107 ge exch 107 le and
{ 139 add 1 string dup 0 3 index put }
{ dup dup -1131 ge exch 1131 le and
{ dup 0 ge { 16#f694 } { neg 16#fa94 } ifelse add
2 string dup 0 3 index -8 bitshift put
dup 1 3 index 255 and put
}
{ 5 string dup 0 255 put exch
2 copy 1 exch -24 bitshift 255 and put
2 copy 2 exch -16 bitshift 255 and put
2 copy 3 exch -8 bitshift 255 and put
2 copy 4 exch 255 and put
exch
}
ifelse
}
ifelse exch pop concatstrings
} bind def
% ------ Point arithmetic utilities ------ %
/ptadd { exch 4 -1 roll add 3 1 roll add } bind def
/ptexch { 4 2 roll } bind def
/ptneg { neg exch neg exch } bind def
/ptpop { pop pop } bind def
/ptsub { ptneg ptadd } bind def
% ------ The main program ------ %
/readBDF % <infilename> <outfilename> <fontname>
% <encodingname> <uniqueID> <xuid> readBDF -> <font>
{ /xuid exch def % may be null
/uniqueID exch def % may be -1
/encodingname exch def
/encoding encodingname cvx exec def
/fontname exch def
/psname exch def
/bdfname exch def
gsave % so we can set the CTM to the font matrix
% Open the input files. We don't open the output file until
% we've done a minimal validity check on the input.
bdfname (r) file /bdfile exch def
/commentword ( ) def
% Check for the STARTFONT.
(STARTFONT) getline
args (2.1) ne { (Not version 2.1\n) print stop } if
% Initialize the font.
/Font 20 dict def
Font begin
/FontName fontname def
/PaintType 0 def
/FontType 1 def
uniqueID 0 gt { /UniqueID uniqueID def } if
xuid null ne { /XUID xuid def } if
/Encoding encoding def
/FontInfo 20 dict def
/Private 20 dict def
currentdict end currentdict end
exch begin begin % insert font above environment
% Initialize the Private dictionary in the font.
Private begin
/-! {string currentfile exch readhexstring pop} readonly def
/-| {string currentfile exch readstring pop} readonly def
/|- {readonly def} readonly def
/| {readonly put} readonly def
/BlueValues [] def
/lenIV lenIV def
/MinFeature {16 16} def
/password 5839 def
/UniqueID uniqueID def
end % Private
% Invert the Encoding, for synthesizing composite characters.
/decoding encoding length dict def
0 1 encoding length 1 sub
{ dup encoding exch get exch decoding 3 1 roll put }
for
% Now open the output file.
psname (w) file /psfile exch def
% Put out a header compatible with the Adobe "standard".
(%!FontType1-1.0: ) ws fontname wt (000.000) wl
(% This is a font description converted from ) ws
bdfname wl
(% by bdftops running on ) ws
statusdict /product get ws ( revision ) ws
revision =string cvs ws (.) wl
% Copy the initial comments, up to FONT.
true
{ nextline
keyword (COMMENT) ne {exit} if
{ (% Here are the initial comments from the BDF file:\n%) wl
} if false
(%) ws remarg wl
} loop pop
() wl
/commentword (COMMENT) def % do skip comments from now on
% Read and process the FONT, SIZE, and FONTBOUNDINGBOX.
% If we cared about FONT, we'd use it here. If the BDF files
% from MIT had PostScript names rather than X names, we would
% care; but what's there is unusable, so we discard FONT.
% The FONTBOUNDINGBOX may not be reliable, so we discard it too.
(FONT) checkline
(SIZE) getline
/pointsize iarg def /xres iarg def /yres iarg def
(FONTBOUNDINGBOX) getline
nextline
% Initialize the font bounding box bookeeping.
/fbbxo 1000 def
/fbbyo 1000 def
/fbbxe -1000 def
/fbbye -1000 def
% Read and process the properties. We only care about a few of them.
keyword (STARTPROPERTIES) eq
{ iarg
{ nextline
properties keyword known
{ FontInfo properties keyword get sarg readonly put
} if
} repeat
(ENDPROPERTIES) getline
nextline
} if
% Compute and set the FontMatrix.
Font /FontMatrix
[ 0.001 0 0 0.001 xres mul yres div 0 0 ] readonly
dup setmatrix put
% Read and process the header for the bitmaps.
(CHARS) checkline
/ccount iarg def
% Initialize the CharStrings dictionary.
/charstrings ccount
composites length add
aliases length add
accentedchars length add
1 add dict def % 1 add for .notdef
/isfixedwidth true def
/fixedwidth null def
/subrcount 0 def
/subrs [] def
% Read the bitmap data. This reads the remainder of the file.
% We do this before processing the bitmaps so that we can compute
% the correct FontBBox first.
/chardata ccount dict def
ccount -1 1
{ (STARTCHAR) getline
/charname remarg def
(ENCODING) getline
/eindex iarg def
eindex dup 0 ge exch 255 le and
{ charname /charname StandardEncoding eindex get def
charname /.notdef eq eindex 0 gt and
{ /charname (A) eindex =string cvs concatstrings cvn def
}
if
(/) print charname =string cvs print (,) print print
}
{ (/) print charname print
}
ifelse
10 mod 1 eq { (\n) print flush } if
(SWIDTH) getline
/swx iarg pointsize mul 1000 div xres mul 72 div def
/swy iarg pointsize mul 1000 div xres mul 72 div def
(DWIDTH) getline % Ignore, use SWIDTH instead
(BBX) getline
/bbw iarg def /bbh iarg def /bbox iarg def /bboy iarg def
nextline
keyword (ATTRIBUTES) eq
{ nextline
} if
(BITMAP) checkline
% Update the font bounding box.
/fbbxo fbbxo bbox .min def
/fbbyo fbbyo bboy .min def
/fbbxe fbbxe bbox bbw add .max def
/fbbye fbbye bboy bbh add .max def
% Read the bits for this character.
/raster bbw 7 add 8 idiv def
/cbits raster bbh mul string def
cbits length 0 gt
{ 0 raster cbits length raster sub
{ cbits exch raster getinterval
bdfile buffer readline not
{ (EOF in bitmap\n) print stop } if
% stack has <cbits.interval> <buffer.interval>
0 () /SubFileDecode filter
exch 2 copy readhexstring pop pop pop closefile
} for
} if
(ENDCHAR) getline
% Save the character data.
chardata charname [swx swy bbw bbh bbox bboy cbits] put
} for
(ENDFONT) getline
% Allocate the buffers for the bitmap and the outline,
% according to the font bounding box.
/fbbw fbbxe fbbxo sub def
/fbbh fbbye fbbyo sub def
/fraster fbbw 7 add 8 idiv def
/bits fraster fbbh mul 200 .max 65535 .min string def
/outline bits length 16 mul 65535 .min string def
% Process the characters.
chardata
{ exch /charname exch def aload pop
/cbits exch def
/bboy exch def /bbox exch def
/bbh exch def /bbw exch def
/swy exch def /swx exch def
% The bitmap handed to type1imagepath must have the correct height,
% because type1imagepath uses this to compute the scale factor,
% so we have to clear the unused parts of it.
/raster bbw 7 add 8 idiv def
bits dup 0 1 raster fbbh mul 1 sub
{ 0 put dup } for
pop pop
bits raster fbbh bbh sub mul cbits putinterval
% Compute the font entry, converting the bitmap to an outline.
bits 0 raster fbbh mul getinterval % the bitmap image
bbw fbbh % bitmap width & height
swx swy % width x & y
bbox neg bboy neg % origin x & y
% Account for lenIV when converting the outline.
outline lenIV outline length lenIV sub getinterval
type1imagepath
length lenIV add
outline exch 0 exch getinterval
% Check for a fixed width font.
isfixedwidth
{ fixedwidth null eq
{ /fixedwidth swx def }
{ fixedwidth swx ne { /isfixedwidth false def } if }
ifelse
} if
% Finish up the character.
copystring
charname exch charstrings 3 1 roll put
} forall
% Add CharStrings entries for aliases.
aliases
{ charstrings 2 index known not charstrings 2 index known and
{ charstrings exch get charstrings 3 1 roll put
}
{ pop pop
}
ifelse
}
forall
% If this is not a fixed-width font, synthesize missing characters
% out of available ones.
isfixedwidth not
{ false composites
{ 1 index charstrings exch known not
1 index { decoding exch known and } forall
{ ( /) print 1 index bits cvs print
/combine exch def
0 1 combine length 1 sub
{ dup combine exch get decoding exch get
bits 3 1 roll put
} for
bits 0 combine length getinterval copystring
[ exch /compose_proc load aload pop ] cvx
charstrings 3 1 roll put
pop true
}
{ pop pop }
ifelse
}
forall flush
{ Private /composematrix matrix put
Private /compose /compose load put
}
if
}
if
% Synthesize accented characters with seac if needed and possible.
accentedchars
{ aload pop /accent exch def /base exch def
buffer cvs /accented exch def
charstrings accented known not
charstrings base known and
charstrings accent known and
StandardDecoding base known and
StandardDecoding accent known and
encoding StandardDecoding base get get base eq and
encoding StandardDecoding accent get get accent eq and
{ ( /) print accented print
charstrings base get findsbw 0 exch getinterval
/acstring exch def % start with sbw of base
charstrings accent get parsesbw
4 { pop } repeat % just leave sbx
acstring exch concatnum
0 concatnum 0 concatnum % adx ady
decoding base get concatnum % bchar
decoding accent get concatnum % achar
s_seac concatstrings
charstrings exch accented copystring exch put
} if
} forall
% Make a CharStrings entry for .notdef.
outline lenIV <8b8b0d0e> putinterval % 0 0 hsbw endchar
charstrings /.notdef outline 0 lenIV 4 add getinterval copystring put
% Encrypt the CharStrings and Subrs (in place).
charstrings
{ % Be careful not to encrypt aliased characters twice,
% since they share their CharString.
aliases 2 index known
{ charstrings aliases 3 index get .knownget
{ 1 index ne }
{ true }
ifelse
}
{ true
}
ifelse
1 index type /stringtype eq and
{ 4330 exch dup .type1encrypt exch pop
readonly charstrings 3 1 roll put
}
{ pop pop
}
ifelse
}
forall
0 1 subrcount 1 sub
{ dup subrs exch get
4330 exch dup .type1encrypt exch pop
subrs 3 1 roll put
}
for
% Make most of the remaining entries in the font dictionaries.
% The Type 1 font machinery really only works with a 1000 unit
% character coordinate system. Set this up here, by computing the factor
% to make the X entry in the FontMatrix come out at exactly 0.001.
/fontscale 1000 fbbh div yres mul xres div def
Font /FontBBox
[ fbbxo fontscale mul
fbbyo fontscale mul
fbbxe fontscale mul
fbbye fontscale mul
] cvx readonly put
Font /CharStrings charstrings readonly put
FontInfo /FullName known not
{ % Some programs insist on FullName being present.
FontInfo /FullName FontName dup length string cvs put
}
if
FontInfo /isFixedPitch isfixedwidth put
subrcount 0 gt
{ Private /Subrs subrs 0 subrcount getinterval readonly put
} if
% Determine the italic angle and underline position
% by actually installing the font.
save
/_temp_ Font definefont setfont
[1000 0 0 1000 0 0] setmatrix % mitigate rounding problems
% The italic angle is the multiple of -5 degrees
% that minimizes the width of the 'I'.
0 9999 0 5 85
{ dup rotate
newpath 0 0 moveto (I) false charpath
dup neg rotate
pathbbox pop exch pop exch sub
dup 3 index lt { 4 -2 roll } if
pop pop
}
for pop
% The underline position is halfway between the bottom of the 'A'
% and the bottom of the FontBBox.
newpath 0 0 moveto (A) false charpath
FontMatrix concat
pathbbox pop pop exch pop
% Put the values in FontInfo.
3 -1 roll
restore
Font /FontBBox get 1 get add 2 div cvi
dup FontInfo /UnderlinePosition 3 -1 roll put
2 div abs FontInfo /UnderlineThickness 3 -1 roll put
FontInfo /ItalicAngle 3 -1 roll put
% Clean up and finish.
grestore
bdfile closefile
Font currentdict end end begin % remove font from dict stack
(\n) print flush
} bind def
% ------ Reader for AFM files ------ %
% Dictionary for looking up character keywords
/cmdict 6 dict dup begin
/C { /c iarg def } def
/N { /n warg copystring def } def
/WX { /w narg def } def
/W0X /WX load def
/W /WX load def
/W0 /WX load def
end def
/readAFM % fontdict afmfilename readAFM -> fontdict
{ (r) file /bdfile exch def
/Font exch def
/commentword (Comment) def
% Check for the StartFontMetrics.
(StartFontMetrics) getline
args cvr 2.0 lt { (Not version 2.0 or greater\n) print stop } if
% Look for StartCharMetrics, then parse the character metrics.
% The only information we care about is the X width.
/metrics 0 dict def
{ nextline
keyword (EndFontMetrics) eq { exit } if
keyword (StartCharMetrics) eq
{ iarg dup dict /metrics exch def
{ /c -1 def /n null def /w null def
nextline buffer
{ token not { exit } if
dup cmdict exch known
{ exch /args exch def cmdict exch get exec args }
{ pop }
ifelse
} loop
c 0 ge n null ne or w null ne and
{ n null eq { /n Font /Encoding get c get def } if
metrics n w put
}
if
}
repeat
(EndCharMetrics) getline
} if
} loop
% Insert the metrics in the font.
metrics length 0 ne
{ Font /Metrics metrics readonly put
} if
Font
} bind def
end % envBDF
% Enter the main program in the current dictionary.
/bdfafmtops % infilename afmfilename* outfilename fontname
% encodingname uniqueID xuid
{ envBDF begin
7 -2 roll exch 7 2 roll % afm* in out fontname encodingname uniqueID xuid
readBDF % afm* font
exch { readAFM } forall
save exch
dup /FontName get exch definefont
setfont
psfile writefont
restore
psfile closefile
end
} bind def
% If the program was invoked from the command line, run it now.
[ shellarguments
{ counttomark 4 ge
{ dup 0 get
dup 48 ge exch 57 le and % last arg starts with a digit?
{ /StandardEncoding } % no encodingname
{ cvn } % have encodingname
ifelse
exch (.) search % next-to-last arg has . in it?
{ mark 4 1 roll % have xuid
{ cvi exch pop exch (.) search not { exit } if }
loop cvi ]
3 -1 roll cvi exch
}
{ cvi null % no xuid
}
ifelse
counttomark 5 roll
counttomark 6 sub array astore
7 -2 roll cvn 7 -3 roll % make sure fontname is a name
bdfafmtops
}
{ cleartomark
(Usage:\n bdftops xx.bdf [yy1.afm ...] zz.gsf fontname uniqueID [xuid] [encodingname]\n) print flush
mark
}
ifelse
}
if pop