2 * "$Id: transcode.c 6649 2007-07-11 21:46:42Z mike $"
4 * Transcoding support for the Common UNIX Printing System (CUPS).
6 * Copyright 2007-2008 by Apple Inc.
7 * Copyright 1997-2007 by Easy Software Products.
9 * These coded instructions, statements, and computer programs are the
10 * property of Apple Inc. and are protected by Federal copyright
11 * law. Distribution and use rights are outlined in the file "LICENSE.txt"
12 * which should have been included with this file. If this file is
13 * file is missing or damaged, see the license at "http://www.cups.org/".
15 * This file is subject to the Apple OS-Developed Software exception.
19 * _cupsCharmapFlush() - Flush all character set maps out of cache.
20 * _cupsCharmapFree() - Free a character set map.
21 * _cupsCharmapGet() - Get a character set map.
22 * cupsCharsetToUTF8() - Convert legacy character set to UTF-8.
23 * cupsUTF8ToCharset() - Convert UTF-8 to legacy character set.
24 * cupsUTF8ToUTF32() - Convert UTF-8 to UTF-32.
25 * cupsUTF32ToUTF8() - Convert UTF-32 to UTF-8.
26 * compare_wide() - Compare key for wide (VBCS) match.
27 * conv_sbcs_to_utf8() - Convert legacy SBCS to UTF-8.
28 * conv_utf8_to_sbcs() - Convert UTF-8 to legacy SBCS.
29 * conv_utf8_to_vbcs() - Convert UTF-8 to legacy DBCS/VBCS.
30 * conv_vbcs_to_utf8() - Convert legacy DBCS/VBCS to UTF-8.
31 * free_sbcs_charmap() - Free memory used by a single byte character set.
32 * free_vbcs_charmap() - Free memory used by a variable byte character set.
33 * get_charmap() - Lookup or get a character set map (private).
34 * get_charmap_count() - Count lines in a charmap file.
35 * get_sbcs_charmap() - Get SBCS Charmap.
36 * get_vbcs_charmap() - Get DBCS/VBCS Charmap.
40 * Include necessary headers...
56 static pthread_mutex_t map_mutex
= PTHREAD_MUTEX_INITIALIZER
;
57 /* Mutex to control access to maps */
58 #endif /* HAVE_PTHREAD_H */
59 static _cups_cmap_t
*cmap_cache
= NULL
;
60 /* SBCS Charmap Cache */
61 static _cups_vmap_t
*vmap_cache
= NULL
;
62 /* VBCS Charmap Cache */
69 static int compare_wide(const void *k1
, const void *k2
);
70 static int conv_sbcs_to_utf8(cups_utf8_t
*dest
,
71 const cups_sbcs_t
*src
,
73 const cups_encoding_t encoding
);
74 static int conv_utf8_to_sbcs(cups_sbcs_t
*dest
,
75 const cups_utf8_t
*src
,
77 const cups_encoding_t encoding
);
78 static int conv_utf8_to_vbcs(cups_sbcs_t
*dest
,
79 const cups_utf8_t
*src
,
81 const cups_encoding_t encoding
);
82 static int conv_vbcs_to_utf8(cups_utf8_t
*dest
,
83 const cups_sbcs_t
*src
,
85 const cups_encoding_t encoding
);
86 static void free_sbcs_charmap(_cups_cmap_t
*sbcs
);
87 static void free_vbcs_charmap(_cups_vmap_t
*vbcs
);
88 static void *get_charmap(const cups_encoding_t encoding
);
89 static int get_charmap_count(cups_file_t
*fp
);
90 static _cups_cmap_t
*get_sbcs_charmap(const cups_encoding_t encoding
,
91 const char *filename
);
92 static _cups_vmap_t
*get_vbcs_charmap(const cups_encoding_t encoding
,
93 const char *filename
);
97 * '_cupsCharmapFlush()' - Flush all character set maps out of cache.
101 _cupsCharmapFlush(void)
103 _cups_cmap_t
*cmap
, /* Legacy SBCS / Unicode Charset Map */
104 *cnext
; /* Next Legacy SBCS Charset Map */
105 _cups_vmap_t
*vmap
, /* Legacy VBCS / Unicode Charset Map */
106 *vnext
; /* Next Legacy VBCS Charset Map */
109 #ifdef HAVE_PTHREAD_H
110 pthread_mutex_lock(&map_mutex
);
111 #endif /* HAVE_PTHREAD_H */
114 * Loop through SBCS charset map cache, free all memory...
117 for (cmap
= cmap_cache
; cmap
; cmap
= cnext
)
121 free_sbcs_charmap(cmap
);
127 * Loop through DBCS/VBCS charset map cache, free all memory...
130 for (vmap
= vmap_cache
; vmap
; vmap
= vnext
)
134 free_vbcs_charmap(vmap
);
139 #ifdef HAVE_PTHREAD_H
140 pthread_mutex_unlock(&map_mutex
);
141 #endif /* HAVE_PTHREAD_H */
146 * '_cupsCharmapFree()' - Free a character set map.
148 * This does not actually free; use '_cupsCharmapFlush()' for that.
153 const cups_encoding_t encoding
) /* I - Encoding */
155 _cups_cmap_t
*cmap
; /* Legacy SBCS / Unicode Charset Map */
156 _cups_vmap_t
*vmap
; /* Legacy VBCS / Unicode Charset Map */
160 * See if we already have this SBCS charset map loaded...
163 #ifdef HAVE_PTHREAD_H
164 pthread_mutex_lock(&map_mutex
);
165 #endif /* HAVE_PTHREAD_H */
167 for (cmap
= cmap_cache
; cmap
; cmap
= cmap
->next
)
169 if (cmap
->encoding
== encoding
)
178 * See if we already have this DBCS/VBCS charset map loaded...
181 for (vmap
= vmap_cache
; vmap
; vmap
= vmap
->next
)
183 if (vmap
->encoding
== encoding
)
191 #ifdef HAVE_PTHREAD_H
192 pthread_mutex_unlock(&map_mutex
);
193 #endif /* HAVE_PTHREAD_H */
198 * '_cupsCharmapGet()' - Get a character set map.
200 * This code handles single-byte (SBCS), double-byte (DBCS), and
201 * variable-byte (VBCS) character sets _without_ charset escapes...
202 * This code does not handle multiple-byte character sets (MBCS)
203 * (such as ISO-2022-JP) with charset switching via escapes...
206 void * /* O - Charset map pointer */
208 const cups_encoding_t encoding
) /* I - Encoding */
210 void *charmap
; /* Charset map pointer */
213 DEBUG_printf(("_cupsCharmapGet(encoding=%d)\n", encoding
));
216 * Check for valid arguments...
219 if (encoding
< 0 || encoding
>= CUPS_ENCODING_VBCS_END
)
221 DEBUG_puts(" Bad encoding, returning NULL!");
226 * Lookup or get the charset map pointer and return...
229 #ifdef HAVE_PTHREAD_H
230 pthread_mutex_lock(&map_mutex
);
231 #endif /* HAVE_PTHREAD_H */
233 charmap
= get_charmap(encoding
);
235 #ifdef HAVE_PTHREAD_H
236 pthread_mutex_unlock(&map_mutex
);
237 #endif /* HAVE_PTHREAD_H */
244 * 'cupsCharsetToUTF8()' - Convert legacy character set to UTF-8.
246 * This code handles single-byte (SBCS), double-byte (DBCS), and
247 * variable-byte (VBCS) character sets _without_ charset escapes...
248 * This code does not handle multiple-byte character sets (MBCS)
249 * (such as ISO-2022-JP) with charset switching via escapes...
252 int /* O - Count or -1 on error */
254 cups_utf8_t
*dest
, /* O - Target string */
255 const char *src
, /* I - Source string */
256 const int maxout
, /* I - Max output */
257 const cups_encoding_t encoding
) /* I - Encoding */
259 int bytes
; /* Number of bytes converted */
263 * Check for valid arguments...
266 DEBUG_printf(("cupsCharsetToUTF8(dest=%p, src=\"%s\", maxout=%d, encoding=%d)\n",
267 dest
, src
, maxout
, encoding
));
272 if (!dest
|| !src
|| maxout
< 1 || maxout
> CUPS_MAX_USTRING
)
274 DEBUG_puts(" Bad arguments, returning -1");
279 * Handle identity conversions...
282 if (encoding
== CUPS_UTF8
||
283 encoding
< 0 || encoding
>= CUPS_ENCODING_VBCS_END
)
285 strlcpy((char *)dest
, src
, maxout
);
286 return ((int)strlen((char *)dest
));
290 * Handle ISO-8859-1 to UTF-8 directly...
293 if (encoding
== CUPS_ISO8859_1
)
295 int ch
; /* Character from string */
296 cups_utf8_t
*destptr
, /* Pointer into UTF-8 buffer */
297 *destend
; /* End of UTF-8 buffer */
301 destend
= dest
+ maxout
- 2;
303 while (*src
&& destptr
< destend
)
309 *destptr
++ = 0xc0 | (ch
>> 6);
310 *destptr
++ = 0x80 | (ch
& 0x3f);
318 return ((int)(destptr
- dest
));
322 * Convert input legacy charset to UTF-8...
325 #ifdef HAVE_PTHREAD_H
326 pthread_mutex_lock(&map_mutex
);
327 #endif /* HAVE_PTHREAD_H */
329 if (encoding
< CUPS_ENCODING_SBCS_END
)
330 bytes
= conv_sbcs_to_utf8(dest
, (cups_sbcs_t
*)src
, maxout
, encoding
);
332 bytes
= conv_vbcs_to_utf8(dest
, (cups_sbcs_t
*)src
, maxout
, encoding
);
334 #ifdef HAVE_PTHREAD_H
335 pthread_mutex_unlock(&map_mutex
);
336 #endif /* HAVE_PTHREAD_H */
343 * 'cupsUTF8ToCharset()' - Convert UTF-8 to legacy character set.
345 * This code handles single-byte (SBCS), double-byte (DBCS), and
346 * variable-byte (VBCS) character sets _without_ charset escapes...
347 * This code does not handle multiple-byte character sets (MBCS)
348 * (such as ISO-2022-JP) with charset switching via escapes...
351 int /* O - Count or -1 on error */
353 char *dest
, /* O - Target string */
354 const cups_utf8_t
*src
, /* I - Source string */
355 const int maxout
, /* I - Max output */
356 const cups_encoding_t encoding
) /* I - Encoding */
358 int bytes
; /* Number of bytes converted */
362 * Check for valid arguments...
365 if (!dest
|| !src
|| maxout
< 1 || maxout
> CUPS_MAX_USTRING
)
374 * Handle identity conversions...
377 if (encoding
== CUPS_UTF8
||
378 encoding
< 0 || encoding
>= CUPS_ENCODING_VBCS_END
)
380 strlcpy(dest
, (char *)src
, maxout
);
381 return ((int)strlen(dest
));
385 * Handle UTF-8 to ISO-8859-1 directly...
388 if (encoding
== CUPS_ISO8859_1
)
390 int ch
; /* Character from string */
391 char *destptr
, /* Pointer into ISO-8859-1 buffer */
392 *destend
; /* End of ISO-8859-1 buffer */
396 destend
= dest
+ maxout
- 1;
398 while (*src
&& destptr
< destend
)
402 if ((ch
& 0xe0) == 0xc0)
404 ch
= ((ch
& 0x1f) << 6) | (*src
++ & 0x3f);
411 else if ((ch
& 0xf0) == 0xe0 ||
414 else if (!(ch
& 0x80))
420 return ((int)(destptr
- dest
));
424 * Convert input UTF-8 to legacy charset...
427 #ifdef HAVE_PTHREAD_H
428 pthread_mutex_lock(&map_mutex
);
429 #endif /* HAVE_PTHREAD_H */
431 if (encoding
< CUPS_ENCODING_SBCS_END
)
432 bytes
= conv_utf8_to_sbcs((cups_sbcs_t
*)dest
, src
, maxout
, encoding
);
434 bytes
= conv_utf8_to_vbcs((cups_sbcs_t
*)dest
, src
, maxout
, encoding
);
436 #ifdef HAVE_PTHREAD_H
437 pthread_mutex_unlock(&map_mutex
);
438 #endif /* HAVE_PTHREAD_H */
445 * 'cupsUTF8ToUTF32()' - Convert UTF-8 to UTF-32.
447 * 32-bit UTF-32 (actually 21-bit) maps to UTF-8 as follows...
449 * UTF-32 char UTF-8 char(s)
450 * --------------------------------------------------
451 * 0 to 127 = 0xxxxxxx (US-ASCII)
452 * 128 to 2047 = 110xxxxx 10yyyyyy
453 * 2048 to 65535 = 1110xxxx 10yyyyyy 10zzzzzz
454 * > 65535 = 11110xxx 10yyyyyy 10zzzzzz 10xxxxxx
456 * UTF-32 prohibits chars beyond Plane 16 (> 0x10ffff) in UCS-4,
457 * which would convert to five- or six-octet UTF-8 sequences...
460 int /* O - Count or -1 on error */
462 cups_utf32_t
*dest
, /* O - Target string */
463 const cups_utf8_t
*src
, /* I - Source string */
464 const int maxout
) /* I - Max output */
466 int i
; /* Looping variable */
467 cups_utf8_t ch
; /* Character value */
468 cups_utf8_t next
; /* Next character value */
469 cups_utf32_t ch32
; /* UTF-32 character value */
473 * Check for valid arguments and clear output...
479 if (!dest
|| !src
|| maxout
< 1 || maxout
> CUPS_MAX_USTRING
)
483 * Convert input UTF-8 to output UTF-32 (and insert BOM)...
488 for (i
= maxout
- 1; *src
&& i
> 0; i
--)
493 * Convert UTF-8 character(s) to UTF-32 character...
499 * One-octet UTF-8 <= 127 (US-ASCII)...
505 else if ((ch
& 0xe0) == 0xc0)
508 * Two-octet UTF-8 <= 2047 (Latin-x)...
515 ch32
= ((ch
& 0x1f) << 6) | (next
& 0x3f);
518 * Check for non-shortest form (invalid UTF-8)...
526 else if ((ch
& 0xf0) == 0xe0)
529 * Three-octet UTF-8 <= 65535 (Plane 0 - BMP)...
536 ch32
= ((ch
& 0x0f) << 6) | (next
& 0x3f);
542 ch32
= (ch32
<< 6) | (next
& 0x3f);
545 * Check for non-shortest form (invalid UTF-8)...
553 else if ((ch
& 0xf8) == 0xf0)
556 * Four-octet UTF-8...
563 ch32
= ((ch
& 0x07) << 6) | (next
& 0x3f);
569 ch32
= (ch32
<< 6) | (next
& 0x3f);
575 ch32
= (ch32
<< 6) | (next
& 0x3f);
578 * Check for non-shortest form (invalid UTF-8)...
589 * More than 4-octet (invalid UTF-8 sequence)...
596 * Check for UTF-16 surrogate (illegal UTF-8)...
599 if (ch32
>= 0xd800 && ch32
<= 0xdfff)
610 * 'cupsUTF32ToUTF8()' - Convert UTF-32 to UTF-8.
612 * 32-bit UTF-32 (actually 21-bit) maps to UTF-8 as follows...
614 * UTF-32 char UTF-8 char(s)
615 * --------------------------------------------------
616 * 0 to 127 = 0xxxxxxx (US-ASCII)
617 * 128 to 2047 = 110xxxxx 10yyyyyy
618 * 2048 to 65535 = 1110xxxx 10yyyyyy 10zzzzzz
619 * > 65535 = 11110xxx 10yyyyyy 10zzzzzz 10xxxxxx
621 * UTF-32 prohibits chars beyond Plane 16 (> 0x10ffff) in UCS-4,
622 * which would convert to five- or six-octet UTF-8 sequences...
625 int /* O - Count or -1 on error */
627 cups_utf8_t
*dest
, /* O - Target string */
628 const cups_utf32_t
*src
, /* I - Source string */
629 const int maxout
) /* I - Max output */
631 cups_utf8_t
*start
; /* Start of destination string */
632 int i
; /* Looping variable */
633 int swap
; /* Byte-swap input to output */
634 cups_utf32_t ch
; /* Character value */
638 * Check for valid arguments and clear output...
644 if (!dest
|| !src
|| maxout
< 1)
648 * Check for leading BOM in UTF-32 and inverted BOM...
652 swap
= *src
== 0xfffe0000;
654 if (*src
== 0xfffe0000 || *src
== 0xfeff)
658 * Convert input UTF-32 to output UTF-8...
661 for (i
= maxout
- 1; *src
&& i
> 0;)
666 * Byte swap input UTF-32, if necessary...
667 * (only byte-swapping 24 of 32 bits)
671 ch
= ((ch
>> 24) | ((ch
>> 8) & 0xff00) | ((ch
<< 8) & 0xff0000));
674 * Check for beyond Plane 16 (invalid UTF-32)...
681 * Convert UTF-32 character to UTF-8 character(s)...
687 * One-octet UTF-8 <= 127 (US-ASCII)...
690 *dest
++ = (cups_utf8_t
)ch
;
696 * Two-octet UTF-8 <= 2047 (Latin-x)...
702 *dest
++ = (cups_utf8_t
)(0xc0 | ((ch
>> 6) & 0x1f));
703 *dest
++ = (cups_utf8_t
)(0x80 | (ch
& 0x3f));
706 else if (ch
< 0x10000)
709 * Three-octet UTF-8 <= 65535 (Plane 0 - BMP)...
715 *dest
++ = (cups_utf8_t
)(0xe0 | ((ch
>> 12) & 0x0f));
716 *dest
++ = (cups_utf8_t
)(0x80 | ((ch
>> 6) & 0x3f));
717 *dest
++ = (cups_utf8_t
)(0x80 | (ch
& 0x3f));
723 * Four-octet UTF-8...
729 *dest
++ = (cups_utf8_t
)(0xf0 | ((ch
>> 18) & 0x07));
730 *dest
++ = (cups_utf8_t
)(0x80 | ((ch
>> 12) & 0x3f));
731 *dest
++ = (cups_utf8_t
)(0x80 | ((ch
>> 6) & 0x3f));
732 *dest
++ = (cups_utf8_t
)(0x80 | (ch
& 0x3f));
739 return ((int)(dest
- start
));
744 * 'compare_wide()' - Compare key for wide (VBCS) match.
748 compare_wide(const void *k1
, /* I - Key char */
749 const void *k2
) /* I - Map char */
751 cups_vbcs_t key
; /* Legacy key character */
752 cups_vbcs_t map
; /* Legacy map character */
755 key
= *((cups_vbcs_t
*)k1
);
756 map
= ((_cups_wide2uni_t
*)k2
)->widechar
;
758 return ((int)(key
- map
));
763 * 'conv_sbcs_to_utf8()' - Convert legacy SBCS to UTF-8.
766 static int /* O - Count or -1 on error */
768 cups_utf8_t
*dest
, /* O - Target string */
769 const cups_sbcs_t
*src
, /* I - Source string */
770 int maxout
, /* I - Max output */
771 const cups_encoding_t encoding
) /* I - Encoding */
773 _cups_cmap_t
*cmap
; /* Legacy SBCS / Unicode Charset Map */
774 cups_ucs2_t
*crow
; /* Pointer to UCS-2 row in 'char2uni' */
775 cups_sbcs_t legchar
; /* Legacy character value */
776 cups_utf32_t work
[CUPS_MAX_USTRING
], /* Internal UCS-4 string */
777 *workptr
; /* Pointer into string */
781 * Find legacy charset map in cache...
784 if ((cmap
= (_cups_cmap_t
*)get_charmap(encoding
)) == NULL
)
788 * Convert input legacy charset to internal UCS-4 (and insert BOM)...
792 for (workptr
= work
+ 1; *src
&& workptr
< (work
+ CUPS_MAX_USTRING
- 1);)
797 * Convert ASCII verbatim (optimization)...
801 *workptr
++ = (cups_utf32_t
)legchar
;
805 * Convert unknown character to Replacement Character...
808 crow
= cmap
->char2uni
+ legchar
;
813 *workptr
++ = (cups_utf32_t
)*crow
;
820 * Convert internal UCS-4 to output UTF-8 (and delete BOM)...
825 return (cupsUTF32ToUTF8(dest
, work
, maxout
));
830 * 'conv_utf8_to_sbcs()' - Convert UTF-8 to legacy SBCS.
833 static int /* O - Count or -1 on error */
835 cups_sbcs_t
*dest
, /* O - Target string */
836 const cups_utf8_t
*src
, /* I - Source string */
837 int maxout
, /* I - Max output */
838 const cups_encoding_t encoding
) /* I - Encoding */
840 cups_sbcs_t
*start
; /* Start of destination string */
841 _cups_cmap_t
*cmap
; /* Legacy SBCS / Unicode Charset Map */
842 cups_sbcs_t
*srow
; /* Pointer to SBCS row in 'uni2char' */
843 cups_utf32_t unichar
; /* Character value */
844 cups_utf32_t work
[CUPS_MAX_USTRING
], /* Internal UCS-4 string */
845 *workptr
; /* Pointer into string */
849 * Find legacy charset map in cache...
852 if ((cmap
= (_cups_cmap_t
*)get_charmap(encoding
)) == NULL
)
856 * Convert input UTF-8 to internal UCS-4 (and insert BOM)...
859 if (cupsUTF8ToUTF32(work
, src
, CUPS_MAX_USTRING
) < 0)
863 * Convert internal UCS-4 to SBCS legacy charset (and delete BOM)...
866 for (workptr
= work
+ 1, start
= dest
; *workptr
&& maxout
> 1; maxout
--)
868 unichar
= *workptr
++;
873 * Convert ASCII verbatim (optimization)...
878 *dest
++ = (cups_sbcs_t
)unichar
;
883 * Convert unknown character to visible replacement...
886 srow
= cmap
->uni2char
[(int)((unichar
>> 8) & 0xff)];
889 srow
+= (int)(unichar
& 0xff);
901 return ((int)(dest
- start
));
906 * 'conv_utf8_to_vbcs()' - Convert UTF-8 to legacy DBCS/VBCS.
909 static int /* O - Count or -1 on error */
911 cups_sbcs_t
*dest
, /* O - Target string */
912 const cups_utf8_t
*src
, /* I - Source string */
913 int maxout
, /* I - Max output */
914 const cups_encoding_t encoding
) /* I - Encoding */
916 cups_sbcs_t
*start
; /* Start of destination string */
917 _cups_vmap_t
*vmap
; /* Legacy DBCS / Unicode Charset Map */
918 cups_vbcs_t
*vrow
; /* Pointer to VBCS row in 'uni2char' */
919 cups_utf32_t unichar
; /* Character value */
920 cups_vbcs_t legchar
; /* Legacy character value */
921 cups_utf32_t work
[CUPS_MAX_USTRING
], /* Internal UCS-4 string */
922 *workptr
; /* Pointer into string */
926 * Find legacy charset map in cache...
929 if ((vmap
= (_cups_vmap_t
*)get_charmap(encoding
)) == NULL
)
933 * Convert input UTF-8 to internal UCS-4 (and insert BOM)...
936 if (cupsUTF8ToUTF32(work
, src
, CUPS_MAX_USTRING
) < 0)
940 * Convert internal UCS-4 to VBCS legacy charset (and delete BOM)...
943 for (start
= dest
, workptr
= work
+ 1; *workptr
&& maxout
> 1; maxout
--)
945 unichar
= *workptr
++;
950 * Convert ASCII verbatim (optimization)...
955 *dest
++ = (cups_sbcs_t
)unichar
;
960 * Convert unknown character to visible replacement...
963 vrow
= vmap
->uni2char
[(int)((unichar
>> 8) & 0xff)];
966 vrow
+= (int)(unichar
& 0xff);
969 legchar
= (cups_vbcs_t
)'?';
971 legchar
= (cups_vbcs_t
)*vrow
;
974 * Save n-byte legacy character...
977 if (legchar
> 0xffffff)
982 *dest
++ = (cups_sbcs_t
)(legchar
>> 24);
983 *dest
++ = (cups_sbcs_t
)(legchar
>> 16);
984 *dest
++ = (cups_sbcs_t
)(legchar
>> 8);
985 *dest
++ = (cups_sbcs_t
)legchar
;
989 else if (legchar
> 0xffff)
994 *dest
++ = (cups_sbcs_t
)(legchar
>> 16);
995 *dest
++ = (cups_sbcs_t
)(legchar
>> 8);
996 *dest
++ = (cups_sbcs_t
)legchar
;
1000 else if (legchar
> 0xff)
1002 *dest
++ = (cups_sbcs_t
)(legchar
>> 8);
1003 *dest
++ = (cups_sbcs_t
)legchar
;
1013 return ((int)(dest
- start
));
1018 * 'conv_vbcs_to_utf8()' - Convert legacy DBCS/VBCS to UTF-8.
1021 static int /* O - Count or -1 on error */
1023 cups_utf8_t
*dest
, /* O - Target string */
1024 const cups_sbcs_t
*src
, /* I - Source string */
1025 int maxout
, /* I - Max output */
1026 const cups_encoding_t encoding
) /* I - Encoding */
1028 _cups_vmap_t
*vmap
; /* Legacy VBCS / Unicode Charset Map */
1029 cups_ucs2_t
*crow
; /* Pointer to UCS-2 row in 'char2uni' */
1030 _cups_wide2uni_t
*wide2uni
; /* Pointer to row in 'wide2uni' */
1031 cups_sbcs_t leadchar
; /* Lead char of n-byte legacy char */
1032 cups_vbcs_t legchar
; /* Legacy character value */
1033 cups_utf32_t work
[CUPS_MAX_USTRING
], /* Internal UCS-4 string */
1034 *workptr
; /* Pointer into string */
1038 * Find legacy charset map in cache...
1041 if ((vmap
= (_cups_vmap_t
*)get_charmap(encoding
)) == NULL
)
1045 * Convert input legacy charset to internal UCS-4 (and insert BOM)...
1049 for (workptr
= work
+ 1; *src
&& workptr
< (work
+ CUPS_MAX_USTRING
- 1);)
1052 leadchar
= (cups_sbcs_t
)legchar
;
1055 * Convert ASCII verbatim (optimization)...
1060 *workptr
++ = (cups_utf32_t
)legchar
;
1065 * Convert 2-byte legacy character...
1068 if (vmap
->lead2char
[(int)leadchar
] == leadchar
)
1073 legchar
= (legchar
<< 8) | *src
++;
1076 * Convert unknown character to Replacement Character...
1079 crow
= vmap
->char2uni
[(int)((legchar
>> 8) & 0xff)];
1081 crow
+= (int) (legchar
& 0xff);
1083 if (!crow
|| !*crow
)
1084 *workptr
++ = 0xfffd;
1086 *workptr
++ = (cups_utf32_t
)*crow
;
1091 * Fetch 3-byte or 4-byte legacy character...
1094 if (vmap
->lead3char
[(int)leadchar
] == leadchar
)
1096 if (!*src
|| !src
[1])
1099 legchar
= (legchar
<< 8) | *src
++;
1100 legchar
= (legchar
<< 8) | *src
++;
1102 else if (vmap
->lead4char
[(int)leadchar
] == leadchar
)
1104 if (!*src
|| !src
[1] || !src
[2])
1107 legchar
= (legchar
<< 8) | *src
++;
1108 legchar
= (legchar
<< 8) | *src
++;
1109 legchar
= (legchar
<< 8) | *src
++;
1115 * Find 3-byte or 4-byte legacy character...
1118 wide2uni
= (_cups_wide2uni_t
*)bsearch(&legchar
,
1121 sizeof(_cups_wide2uni_t
),
1125 * Convert unknown character to Replacement Character...
1128 if (!wide2uni
|| !wide2uni
->unichar
)
1129 *workptr
++ = 0xfffd;
1131 *workptr
++ = wide2uni
->unichar
;
1139 * Convert internal UCS-4 to output UTF-8 (and delete BOM)...
1142 return (cupsUTF32ToUTF8(dest
, work
, maxout
));
1147 * 'free_sbcs_charmap()' - Free memory used by a single byte character set.
1151 free_sbcs_charmap(_cups_cmap_t
*cmap
) /* I - Character set */
1153 int i
; /* Looping variable */
1156 for (i
= 0; i
< 256; i
++)
1157 if (cmap
->uni2char
[i
])
1158 free(cmap
->uni2char
[i
]);
1165 * 'free_vbcs_charmap()' - Free memory used by a variable byte character set.
1169 free_vbcs_charmap(_cups_vmap_t
*vmap
) /* I - Character set */
1171 int i
; /* Looping variable */
1174 for (i
= 0; i
< 256; i
++)
1175 if (vmap
->char2uni
[i
])
1176 free(vmap
->char2uni
[i
]);
1178 for (i
= 0; i
< 256; i
++)
1179 if (vmap
->uni2char
[i
])
1180 free(vmap
->uni2char
[i
]);
1183 free(vmap
->wide2uni
);
1190 * 'get_charmap()' - Lookup or get a character set map (private).
1192 * This code handles single-byte (SBCS), double-byte (DBCS), and
1193 * variable-byte (VBCS) character sets _without_ charset escapes...
1194 * This code does not handle multiple-byte character sets (MBCS)
1195 * (such as ISO-2022-JP) with charset switching via escapes...
1199 static void * /* O - Charset map pointer */
1201 const cups_encoding_t encoding
) /* I - Encoding */
1203 char filename
[1024]; /* Filename for charset map file */
1204 _cups_globals_t
*cg
= _cupsGlobals(); /* Global data */
1208 * Get the data directory and charset map name...
1211 snprintf(filename
, sizeof(filename
), "%s/charmaps/%s.txt",
1212 cg
->cups_datadir
, _cupsEncodingName(encoding
));
1214 DEBUG_printf((" filename=\"%s\"\n", filename
));
1217 * Read charset map input file into cache...
1220 if (encoding
< CUPS_ENCODING_SBCS_END
)
1221 return (get_sbcs_charmap(encoding
, filename
));
1222 else if (encoding
< CUPS_ENCODING_VBCS_END
)
1223 return (get_vbcs_charmap(encoding
, filename
));
1230 * 'get_charmap_count()' - Count lines in a charmap file.
1233 static int /* O - Count or -1 on error */
1234 get_charmap_count(cups_file_t
*fp
) /* I - File to read from */
1236 int count
; /* Number of lines */
1237 char line
[256]; /* Line from input map file */
1241 * Count lines in map input file...
1246 while (cupsFileGets(fp
, line
, sizeof(line
)))
1251 * Return the number of lines...
1262 * 'get_sbcs_charmap()' - Get SBCS Charmap.
1265 static _cups_cmap_t
* /* O - Charmap or 0 on error */
1267 const cups_encoding_t encoding
, /* I - Charmap Encoding */
1268 const char *filename
) /* I - Charmap Filename */
1270 unsigned long legchar
; /* Legacy character value */
1271 cups_utf32_t unichar
; /* Unicode character value */
1272 _cups_cmap_t
*cmap
; /* Legacy SBCS / Unicode Charset Map */
1273 cups_file_t
*fp
; /* Charset map file pointer */
1274 char *s
; /* Line parsing pointer */
1275 cups_ucs2_t
*crow
; /* Pointer to UCS-2 row in 'char2uni' */
1276 cups_sbcs_t
*srow
; /* Pointer to SBCS row in 'uni2char' */
1277 char line
[256]; /* Line from charset map file */
1281 * See if we already have this SBCS charset map loaded...
1284 for (cmap
= cmap_cache
; cmap
; cmap
= cmap
->next
)
1286 if (cmap
->encoding
== encoding
)
1289 DEBUG_printf((" returning existing cmap=%p\n", cmap
));
1291 return ((void *)cmap
);
1296 * Open SBCS charset map input file...
1299 if ((fp
= cupsFileOpen(filename
, "r")) == NULL
)
1303 * Allocate memory for SBCS charset map...
1306 if ((cmap
= (_cups_cmap_t
*)calloc(1, sizeof(_cups_cmap_t
))) == NULL
)
1309 DEBUG_puts(" Unable to allocate memory!");
1315 cmap
->encoding
= encoding
;
1318 * Save SBCS charset map into memory for transcoding...
1321 while (cupsFileGets(fp
, line
, sizeof(line
)))
1326 legchar
= strtol(line
, &s
, 16);
1327 if (legchar
< 0 || legchar
> 0xff)
1330 unichar
= strtol(s
, NULL
, 16);
1331 if (unichar
< 0 || unichar
> 0xffff)
1335 * Save legacy to Unicode mapping in direct lookup table...
1338 crow
= cmap
->char2uni
+ legchar
;
1339 *crow
= (cups_ucs2_t
)(unichar
& 0xffff);
1342 * Save Unicode to legacy mapping in indirect lookup table...
1345 srow
= cmap
->uni2char
[(unichar
>> 8) & 0xff];
1348 srow
= (cups_sbcs_t
*)calloc(256, sizeof(cups_sbcs_t
));
1352 cmap
->uni2char
[(unichar
>> 8) & 0xff] = srow
;
1355 srow
+= unichar
& 0xff;
1358 * Convert Replacement Character to visible replacement...
1361 if (unichar
== 0xfffd)
1362 legchar
= (unsigned long)'?';
1365 * First (oldest) legacy character uses Unicode mapping cell...
1369 *srow
= (cups_sbcs_t
)legchar
;
1375 * Add it to the cache and return...
1378 cmap
->next
= cmap_cache
;
1381 DEBUG_printf((" returning new cmap=%p\n", cmap
));
1386 * If we get here, there was an error in the cmap file...
1391 free_sbcs_charmap(cmap
);
1395 DEBUG_puts(" Error, returning NULL!");
1402 * 'get_vbcs_charmap()' - Get DBCS/VBCS Charmap.
1405 static _cups_vmap_t
* /* O - Charmap or 0 on error */
1407 const cups_encoding_t encoding
, /* I - Charmap Encoding */
1408 const char *filename
) /* I - Charmap Filename */
1410 _cups_vmap_t
*vmap
; /* Legacy VBCS / Unicode Charset Map */
1411 cups_ucs2_t
*crow
; /* Pointer to UCS-2 row in 'char2uni' */
1412 cups_vbcs_t
*vrow
; /* Pointer to VBCS row in 'uni2char' */
1413 _cups_wide2uni_t
*wide2uni
; /* Pointer to row in 'wide2uni' */
1414 cups_sbcs_t leadchar
; /* Lead char of 2-byte legacy char */
1415 unsigned long legchar
; /* Legacy character value */
1416 cups_utf32_t unichar
; /* Unicode character value */
1417 int mapcount
; /* Count of lines in charmap file */
1418 cups_file_t
*fp
; /* Charset map file pointer */
1419 char *s
; /* Line parsing pointer */
1420 char line
[256]; /* Line from charset map file */
1421 int i
; /* Loop variable */
1422 int legacy
; /* 32-bit legacy char */
1425 DEBUG_printf(("get_vbcs_charmap(encoding=%d, filename=\"%s\")\n",
1426 encoding
, filename
));
1429 * See if we already have this DBCS/VBCS charset map loaded...
1432 for (vmap
= vmap_cache
; vmap
; vmap
= vmap
->next
)
1434 if (vmap
->encoding
== encoding
)
1437 DEBUG_printf((" returning existing vmap=%p\n", vmap
));
1439 return ((void *)vmap
);
1444 * Open VBCS charset map input file...
1447 if ((fp
= cupsFileOpen(filename
, "r")) == NULL
)
1449 DEBUG_printf((" Unable to open file: %s\n", strerror(errno
)));
1455 * Count lines in charmap file...
1458 if ((mapcount
= get_charmap_count(fp
)) <= 0)
1460 DEBUG_puts(" Unable to get charmap count!");
1467 DEBUG_printf((" mapcount=%d\n", mapcount
));
1470 * Allocate memory for DBCS/VBCS charset map...
1473 if ((vmap
= (_cups_vmap_t
*)calloc(1, sizeof(_cups_vmap_t
))) == NULL
)
1475 DEBUG_puts(" Unable to allocate memory!");
1483 vmap
->encoding
= encoding
;
1486 * Save DBCS/VBCS charset map into memory for transcoding...
1497 while (cupsFileGets(fp
, line
, sizeof(line
)))
1502 legchar
= strtoul(line
, &s
, 16);
1503 if (legchar
== ULONG_MAX
)
1506 unichar
= strtol(s
, NULL
, 16);
1507 if (unichar
< 0 || unichar
> 0xffff)
1512 /* DEBUG_printf((" i=%d, legchar=0x%08lx, unichar=0x%04x\n", i,
1513 legchar, (unsigned)unichar)); */
1516 * Save lead char of 2/3/4-byte legacy char...
1519 if (legchar
> 0xff && legchar
<= 0xffff)
1521 leadchar
= (cups_sbcs_t
)(legchar
>> 8);
1522 vmap
->lead2char
[leadchar
] = leadchar
;
1525 if (legchar
> 0xffff && legchar
<= 0xffffff)
1527 leadchar
= (cups_sbcs_t
)(legchar
>> 16);
1528 vmap
->lead3char
[leadchar
] = leadchar
;
1531 if (legchar
> 0xffffff)
1533 leadchar
= (cups_sbcs_t
)(legchar
>> 24);
1534 vmap
->lead4char
[leadchar
] = leadchar
;
1538 * Save Legacy to Unicode mapping...
1541 if (legchar
<= 0xffff)
1544 * Save DBCS 16-bit to Unicode mapping in indirect lookup table...
1547 crow
= vmap
->char2uni
[(int)leadchar
];
1550 crow
= (cups_ucs2_t
*)calloc(256, sizeof(cups_ucs2_t
));
1554 vmap
->char2uni
[(int)leadchar
] = crow
;
1557 crow
[(int)(legchar
& 0xff)] = (cups_ucs2_t
)unichar
;
1562 * Save VBCS 32-bit to Unicode mapping in sorted list table...
1568 vmap
->widecount
= (mapcount
- i
+ 1);
1569 wide2uni
= (_cups_wide2uni_t
*)calloc(vmap
->widecount
,
1570 sizeof(_cups_wide2uni_t
));
1574 vmap
->wide2uni
= wide2uni
;
1577 wide2uni
->widechar
= (cups_vbcs_t
)legchar
;
1578 wide2uni
->unichar
= (cups_ucs2_t
)unichar
;
1583 * Save Unicode to legacy mapping in indirect lookup table...
1586 vrow
= vmap
->uni2char
[(int)((unichar
>> 8) & 0xff)];
1589 vrow
= (cups_vbcs_t
*)calloc(256, sizeof(cups_vbcs_t
));
1593 vmap
->uni2char
[(int) ((unichar
>> 8) & 0xff)] = vrow
;
1596 vrow
+= (int)(unichar
& 0xff);
1599 * Convert Replacement Character to visible replacement...
1602 if (unichar
== 0xfffd)
1603 legchar
= (unsigned long)'?';
1606 * First (oldest) legacy character uses Unicode mapping cell...
1610 *vrow
= (cups_vbcs_t
)legchar
;
1613 vmap
->charcount
= (i
- vmap
->widecount
);
1618 * Add it to the cache and return...
1621 vmap
->next
= vmap_cache
;
1624 DEBUG_printf((" returning new vmap=%p\n", vmap
));
1629 * If we get here, the file contains errors...
1634 free_vbcs_charmap(vmap
);
1638 DEBUG_puts(" Error, returning NULL!");
1645 * End of "$Id: transcode.c 6649 2007-07-11 21:46:42Z mike $"