1 | #!/usr/bin/python -u |
---|
2 | # |
---|
3 | # Original script modified in November 2003 to take advantage of |
---|
4 | # the character-validation range routines, and updated to the |
---|
5 | # current Unicode information (Version 4.0.1) |
---|
6 | # |
---|
7 | # NOTE: there is an 'alias' facility for blocks which are not present in |
---|
8 | # the current release, but are needed for ABI compatibility. This |
---|
9 | # must be accomplished MANUALLY! Please see the comments below under |
---|
10 | # 'blockAliases' |
---|
11 | # |
---|
12 | import sys |
---|
13 | import string |
---|
14 | import time |
---|
15 | |
---|
16 | webpage = "http://www.unicode.org/Public/4.0-Update1/UCD-4.0.1d5b.html" |
---|
17 | sources = "Blocks-4.0.1d1b.txt UnicodeData-4.0.1d1b.txt" |
---|
18 | |
---|
19 | # |
---|
20 | # blockAliases is a small hack - it is used for mapping block names which |
---|
21 | # were were used in the 3.1 release, but are missing or changed in the current |
---|
22 | # release. The format is "OldBlockName:NewBlockName1[,NewBlockName2[,...]]" |
---|
23 | blockAliases = [] |
---|
24 | blockAliases.append("CombiningMarksforSymbols:CombiningDiacriticalMarksforSymbols") |
---|
25 | blockAliases.append("Greek:GreekandCoptic") |
---|
26 | blockAliases.append("PrivateUse:PrivateUseArea,SupplementaryPrivateUseArea-A," + |
---|
27 | "SupplementaryPrivateUseArea-B") |
---|
28 | |
---|
29 | # minTableSize gives the minimum number of ranges which must be present |
---|
30 | # before a range table is produced. If there are less than this |
---|
31 | # number, inline comparisons are generated |
---|
32 | minTableSize = 8 |
---|
33 | |
---|
34 | (blockfile, catfile) = string.split(sources) |
---|
35 | |
---|
36 | |
---|
37 | # |
---|
38 | # Now process the "blocks" file, reducing it to a dictionary |
---|
39 | # indexed by blockname, containing a tuple with the applicable |
---|
40 | # block range |
---|
41 | # |
---|
42 | BlockNames = {} |
---|
43 | try: |
---|
44 | blocks = open(blockfile, "r") |
---|
45 | except: |
---|
46 | print "Missing %s, aborting ..." % blockfile |
---|
47 | sys.exit(1) |
---|
48 | |
---|
49 | for line in blocks.readlines(): |
---|
50 | if line[0] == '#': |
---|
51 | continue |
---|
52 | line = string.strip(line) |
---|
53 | if line == '': |
---|
54 | continue |
---|
55 | try: |
---|
56 | fields = string.split(line, ';') |
---|
57 | range = string.strip(fields[0]) |
---|
58 | (start, end) = string.split(range, "..") |
---|
59 | name = string.strip(fields[1]) |
---|
60 | name = string.replace(name, ' ', '') |
---|
61 | except: |
---|
62 | print "Failed to process line: %s" % (line) |
---|
63 | continue |
---|
64 | start = "0x" + start |
---|
65 | end = "0x" + end |
---|
66 | try: |
---|
67 | BlockNames[name].append((start, end)) |
---|
68 | except: |
---|
69 | BlockNames[name] = [(start, end)] |
---|
70 | blocks.close() |
---|
71 | print "Parsed %d blocks descriptions" % (len(BlockNames.keys())) |
---|
72 | |
---|
73 | for block in blockAliases: |
---|
74 | alias = string.split(block,':') |
---|
75 | alist = string.split(alias[1],',') |
---|
76 | for comp in alist: |
---|
77 | if BlockNames.has_key(comp): |
---|
78 | if alias[0] not in BlockNames: |
---|
79 | BlockNames[alias[0]] = [] |
---|
80 | for r in BlockNames[comp]: |
---|
81 | BlockNames[alias[0]].append(r) |
---|
82 | else: |
---|
83 | print "Alias %s: %s not in Blocks" % (alias[0], comp) |
---|
84 | continue |
---|
85 | |
---|
86 | # |
---|
87 | # Next process the Categories file. This is more complex, since |
---|
88 | # the file is in code sequence, and we need to invert it. We use |
---|
89 | # a dictionary with index category-name, with each entry containing |
---|
90 | # all the ranges (codepoints) of that category. Note that category |
---|
91 | # names comprise two parts - the general category, and the "subclass" |
---|
92 | # within that category. Therefore, both "general category" (which is |
---|
93 | # the first character of the 2-character category-name) and the full |
---|
94 | # (2-character) name are entered into this dictionary. |
---|
95 | # |
---|
96 | try: |
---|
97 | data = open(catfile, "r") |
---|
98 | except: |
---|
99 | print "Missing %s, aborting ..." % catfile |
---|
100 | sys.exit(1) |
---|
101 | |
---|
102 | nbchar = 0; |
---|
103 | Categories = {} |
---|
104 | for line in data.readlines(): |
---|
105 | if line[0] == '#': |
---|
106 | continue |
---|
107 | line = string.strip(line) |
---|
108 | if line == '': |
---|
109 | continue |
---|
110 | try: |
---|
111 | fields = string.split(line, ';') |
---|
112 | point = string.strip(fields[0]) |
---|
113 | value = 0 |
---|
114 | while point != '': |
---|
115 | value = value * 16 |
---|
116 | if point[0] >= '0' and point[0] <= '9': |
---|
117 | value = value + ord(point[0]) - ord('0') |
---|
118 | elif point[0] >= 'A' and point[0] <= 'F': |
---|
119 | value = value + 10 + ord(point[0]) - ord('A') |
---|
120 | elif point[0] >= 'a' and point[0] <= 'f': |
---|
121 | value = value + 10 + ord(point[0]) - ord('a') |
---|
122 | point = point[1:] |
---|
123 | name = fields[2] |
---|
124 | except: |
---|
125 | print "Failed to process line: %s" % (line) |
---|
126 | continue |
---|
127 | |
---|
128 | nbchar = nbchar + 1 |
---|
129 | # update entry for "full name" |
---|
130 | try: |
---|
131 | Categories[name].append(value) |
---|
132 | except: |
---|
133 | try: |
---|
134 | Categories[name] = [value] |
---|
135 | except: |
---|
136 | print "Failed to process line: %s" % (line) |
---|
137 | # update "general category" name |
---|
138 | try: |
---|
139 | Categories[name[0]].append(value) |
---|
140 | except: |
---|
141 | try: |
---|
142 | Categories[name[0]] = [value] |
---|
143 | except: |
---|
144 | print "Failed to process line: %s" % (line) |
---|
145 | |
---|
146 | blocks.close() |
---|
147 | print "Parsed %d char generating %d categories" % (nbchar, len(Categories.keys())) |
---|
148 | |
---|
149 | # |
---|
150 | # The data is now all read. Time to process it into a more useful form. |
---|
151 | # |
---|
152 | # reduce the number list into ranges |
---|
153 | for cat in Categories.keys(): |
---|
154 | list = Categories[cat] |
---|
155 | start = -1 |
---|
156 | prev = -1 |
---|
157 | end = -1 |
---|
158 | ranges = [] |
---|
159 | for val in list: |
---|
160 | if start == -1: |
---|
161 | start = val |
---|
162 | prev = val |
---|
163 | continue |
---|
164 | elif val == prev + 1: |
---|
165 | prev = val |
---|
166 | continue |
---|
167 | elif prev == start: |
---|
168 | ranges.append((prev, prev)) |
---|
169 | start = val |
---|
170 | prev = val |
---|
171 | continue |
---|
172 | else: |
---|
173 | ranges.append((start, prev)) |
---|
174 | start = val |
---|
175 | prev = val |
---|
176 | continue |
---|
177 | if prev == start: |
---|
178 | ranges.append((prev, prev)) |
---|
179 | else: |
---|
180 | ranges.append((start, prev)) |
---|
181 | Categories[cat] = ranges |
---|
182 | |
---|
183 | # |
---|
184 | # Assure all data is in alphabetic order, since we will be doing binary |
---|
185 | # searches on the tables. |
---|
186 | # |
---|
187 | bkeys = BlockNames.keys() |
---|
188 | bkeys.sort() |
---|
189 | |
---|
190 | ckeys = Categories.keys() |
---|
191 | ckeys.sort() |
---|
192 | |
---|
193 | # |
---|
194 | # Generate the resulting files |
---|
195 | # |
---|
196 | try: |
---|
197 | header = open("include/libxml/xmlunicode.h", "w") |
---|
198 | except: |
---|
199 | print "Failed to open include/libxml/xmlunicode.h" |
---|
200 | sys.exit(1) |
---|
201 | |
---|
202 | try: |
---|
203 | output = open("xmlunicode.c", "w") |
---|
204 | except: |
---|
205 | print "Failed to open xmlunicode.c" |
---|
206 | sys.exit(1) |
---|
207 | |
---|
208 | date = time.asctime(time.localtime(time.time())) |
---|
209 | |
---|
210 | header.write( |
---|
211 | """/* |
---|
212 | * Summary: Unicode character APIs |
---|
213 | * Description: API for the Unicode character APIs |
---|
214 | * |
---|
215 | * This file is automatically generated from the |
---|
216 | * UCS description files of the Unicode Character Database |
---|
217 | * %s |
---|
218 | * using the genUnicode.py Python script. |
---|
219 | * |
---|
220 | * Generation date: %s |
---|
221 | * Sources: %s |
---|
222 | * Author: Daniel Veillard |
---|
223 | */ |
---|
224 | |
---|
225 | #ifndef __XML_UNICODE_H__ |
---|
226 | #define __XML_UNICODE_H__ |
---|
227 | |
---|
228 | #include <libxml/xmlversion.h> |
---|
229 | |
---|
230 | #ifdef __cplusplus |
---|
231 | extern "C" { |
---|
232 | #endif |
---|
233 | |
---|
234 | """ % (webpage, date, sources)); |
---|
235 | |
---|
236 | output.write( |
---|
237 | """/* |
---|
238 | * xmlunicode.c: this module implements the Unicode character APIs |
---|
239 | * |
---|
240 | * This file is automatically generated from the |
---|
241 | * UCS description files of the Unicode Character Database |
---|
242 | * %s |
---|
243 | * using the genUnicode.py Python script. |
---|
244 | * |
---|
245 | * Generation date: %s |
---|
246 | * Sources: %s |
---|
247 | * Daniel Veillard <veillard@redhat.com> |
---|
248 | */ |
---|
249 | |
---|
250 | #define IN_LIBXML |
---|
251 | #include "libxml.h" |
---|
252 | |
---|
253 | #ifdef LIBXML_UNICODE_ENABLED |
---|
254 | |
---|
255 | #include <string.h> |
---|
256 | #include <libxml/xmlversion.h> |
---|
257 | #include <libxml/xmlunicode.h> |
---|
258 | #include <libxml/chvalid.h> |
---|
259 | |
---|
260 | typedef int (xmlIntFunc)(int); /* just to keep one's mind untwisted */ |
---|
261 | |
---|
262 | typedef struct { |
---|
263 | const char *rangename; |
---|
264 | xmlIntFunc *func; |
---|
265 | } xmlUnicodeRange; |
---|
266 | |
---|
267 | typedef struct { |
---|
268 | xmlUnicodeRange *table; |
---|
269 | int numentries; |
---|
270 | } xmlUnicodeNameTable; |
---|
271 | |
---|
272 | |
---|
273 | static xmlIntFunc *xmlUnicodeLookup(xmlUnicodeNameTable *tptr, const char *tname); |
---|
274 | |
---|
275 | static xmlUnicodeRange xmlUnicodeBlocks[] = { |
---|
276 | """ % (webpage, date, sources)); |
---|
277 | |
---|
278 | flag = 0 |
---|
279 | for block in bkeys: |
---|
280 | name = string.replace(block, '-', '') |
---|
281 | if flag: |
---|
282 | output.write(',\n') |
---|
283 | else: |
---|
284 | flag = 1 |
---|
285 | output.write(' {"%s", xmlUCSIs%s}' % (block, name)) |
---|
286 | output.write('};\n\n') |
---|
287 | |
---|
288 | output.write('static xmlUnicodeRange xmlUnicodeCats[] = {\n') |
---|
289 | flag = 0; |
---|
290 | for name in ckeys: |
---|
291 | if flag: |
---|
292 | output.write(',\n') |
---|
293 | else: |
---|
294 | flag = 1 |
---|
295 | output.write(' {"%s", xmlUCSIsCat%s}' % (name, name)) |
---|
296 | output.write('};\n\n') |
---|
297 | |
---|
298 | # |
---|
299 | # For any categories with more than minTableSize ranges we generate |
---|
300 | # a range table suitable for xmlCharInRange |
---|
301 | # |
---|
302 | for name in ckeys: |
---|
303 | if len(Categories[name]) > minTableSize: |
---|
304 | numshort = 0 |
---|
305 | numlong = 0 |
---|
306 | ranges = Categories[name] |
---|
307 | sptr = "NULL" |
---|
308 | lptr = "NULL" |
---|
309 | for range in ranges: |
---|
310 | (low, high) = range |
---|
311 | if high < 0x10000: |
---|
312 | if numshort == 0: |
---|
313 | pline = "static xmlChSRange xml%sS[] = {" % name |
---|
314 | sptr = "xml%sS" % name |
---|
315 | else: |
---|
316 | pline += ", " |
---|
317 | numshort += 1 |
---|
318 | else: |
---|
319 | if numlong == 0: |
---|
320 | if numshort > 0: |
---|
321 | output.write(pline + " };\n") |
---|
322 | pline = "static xmlChLRange xml%sL[] = {" % name |
---|
323 | lptr = "xml%sL" % name |
---|
324 | else: |
---|
325 | pline += ", " |
---|
326 | numlong += 1 |
---|
327 | if len(pline) > 60: |
---|
328 | output.write(pline + "\n") |
---|
329 | pline = " " |
---|
330 | pline += "{%s, %s}" % (hex(low), hex(high)) |
---|
331 | output.write(pline + " };\nstatic xmlChRangeGroup xml%sG = {%s,%s,%s,%s};\n\n" |
---|
332 | % (name, numshort, numlong, sptr, lptr)) |
---|
333 | |
---|
334 | |
---|
335 | output.write( |
---|
336 | """xmlUnicodeNameTable xmlUnicodeBlockTbl = {xmlUnicodeBlocks, %s}; |
---|
337 | xmlUnicodeNameTable xmlUnicodeCatTbl = {xmlUnicodeCats, %s}; |
---|
338 | |
---|
339 | /** |
---|
340 | * xmlUnicodeLookup: |
---|
341 | * @tptr: pointer to the name table |
---|
342 | * @name: name to be found |
---|
343 | * |
---|
344 | * binary table lookup for user-supplied name |
---|
345 | * |
---|
346 | * Returns pointer to range function if found, otherwise NULL |
---|
347 | */ |
---|
348 | static xmlIntFunc |
---|
349 | *xmlUnicodeLookup(xmlUnicodeNameTable *tptr, const char *tname) { |
---|
350 | int low, high, mid, cmp; |
---|
351 | xmlUnicodeRange *sptr; |
---|
352 | |
---|
353 | low = 0; |
---|
354 | high = tptr->numentries - 1; |
---|
355 | sptr = tptr->table; |
---|
356 | while (low <= high) { |
---|
357 | mid = (low + high) / 2; |
---|
358 | if ((cmp=strcmp(tname, sptr[mid].rangename)) == 0) |
---|
359 | return (sptr[mid].func); |
---|
360 | if (cmp < 0) |
---|
361 | high = mid - 1; |
---|
362 | else |
---|
363 | low = mid + 1; |
---|
364 | } |
---|
365 | return (NULL); |
---|
366 | } |
---|
367 | |
---|
368 | """ % (len(BlockNames), len(Categories)) ) |
---|
369 | |
---|
370 | for block in bkeys: |
---|
371 | name = string.replace(block, '-', '') |
---|
372 | header.write("XMLPUBFUN int XMLCALL xmlUCSIs%s\t(int code);\n" % name) |
---|
373 | output.write("/**\n * xmlUCSIs%s:\n * @code: UCS code point\n" % (name)) |
---|
374 | output.write(" *\n * Check whether the character is part of %s UCS Block\n"% |
---|
375 | (block)) |
---|
376 | output.write(" *\n * Returns 1 if true 0 otherwise\n */\n"); |
---|
377 | output.write("int\nxmlUCSIs%s(int code) {\n return(" % name) |
---|
378 | flag = 0 |
---|
379 | for (start, end) in BlockNames[block]: |
---|
380 | if flag: |
---|
381 | output.write(" ||\n ") |
---|
382 | else: |
---|
383 | flag = 1 |
---|
384 | output.write("((code >= %s) && (code <= %s))" % (start, end)) |
---|
385 | output.write(");\n}\n\n") |
---|
386 | |
---|
387 | header.write("\nXMLPUBFUN int XMLCALL xmlUCSIsBlock\t(int code, const char *block);\n\n") |
---|
388 | output.write( |
---|
389 | """/** |
---|
390 | * xmlUCSIsBlock: |
---|
391 | * @code: UCS code point |
---|
392 | * @block: UCS block name |
---|
393 | * |
---|
394 | * Check whether the character is part of the UCS Block |
---|
395 | * |
---|
396 | * Returns 1 if true, 0 if false and -1 on unknown block |
---|
397 | */ |
---|
398 | int |
---|
399 | xmlUCSIsBlock(int code, const char *block) { |
---|
400 | xmlIntFunc *func; |
---|
401 | |
---|
402 | func = xmlUnicodeLookup(&xmlUnicodeBlockTbl, block); |
---|
403 | if (func == NULL) |
---|
404 | return (-1); |
---|
405 | return (func(code)); |
---|
406 | } |
---|
407 | |
---|
408 | """) |
---|
409 | |
---|
410 | for name in ckeys: |
---|
411 | ranges = Categories[name] |
---|
412 | header.write("XMLPUBFUN int XMLCALL xmlUCSIsCat%s\t(int code);\n" % name) |
---|
413 | output.write("/**\n * xmlUCSIsCat%s:\n * @code: UCS code point\n" % (name)) |
---|
414 | output.write(" *\n * Check whether the character is part of %s UCS Category\n"% |
---|
415 | (name)) |
---|
416 | output.write(" *\n * Returns 1 if true 0 otherwise\n */\n"); |
---|
417 | output.write("int\nxmlUCSIsCat%s(int code) {\n" % name) |
---|
418 | if len(Categories[name]) > minTableSize: |
---|
419 | output.write(" return(xmlCharInRange((unsigned int)code, &xml%sG)" |
---|
420 | % name) |
---|
421 | else: |
---|
422 | start = 1 |
---|
423 | for range in ranges: |
---|
424 | (begin, end) = range; |
---|
425 | if start: |
---|
426 | output.write(" return("); |
---|
427 | start = 0 |
---|
428 | else: |
---|
429 | output.write(" ||\n "); |
---|
430 | if (begin == end): |
---|
431 | output.write("(code == %s)" % (hex(begin))) |
---|
432 | else: |
---|
433 | output.write("((code >= %s) && (code <= %s))" % ( |
---|
434 | hex(begin), hex(end))) |
---|
435 | output.write(");\n}\n\n") |
---|
436 | |
---|
437 | header.write("\nXMLPUBFUN int XMLCALL xmlUCSIsCat\t(int code, const char *cat);\n") |
---|
438 | output.write( |
---|
439 | """/** |
---|
440 | * xmlUCSIsCat: |
---|
441 | * @code: UCS code point |
---|
442 | * @cat: UCS Category name |
---|
443 | * |
---|
444 | * Check whether the character is part of the UCS Category |
---|
445 | * |
---|
446 | * Returns 1 if true, 0 if false and -1 on unknown category |
---|
447 | */ |
---|
448 | int |
---|
449 | xmlUCSIsCat(int code, const char *cat) { |
---|
450 | xmlIntFunc *func; |
---|
451 | |
---|
452 | func = xmlUnicodeLookup(&xmlUnicodeCatTbl, cat); |
---|
453 | if (func == NULL) |
---|
454 | return (-1); |
---|
455 | return (func(code)); |
---|
456 | } |
---|
457 | |
---|
458 | |
---|
459 | #endif /* LIBXML_UNICODE_ENABLED */ |
---|
460 | """) |
---|
461 | |
---|
462 | header.write(""" |
---|
463 | #ifdef __cplusplus |
---|
464 | } |
---|
465 | #endif |
---|
466 | #endif /* __XML_UNICODE_H__ */ |
---|
467 | """); |
---|
468 | |
---|
469 | header.close() |
---|
470 | output.close() |
---|