1 // Copyright (C) 2016 and later: Unicode, Inc. and others.
2 // License & terms of use: http://www.unicode.org/copyright.html
3 /*
4 **********************************************************************
5 * Copyright (C) 2005-2016, International Business Machines
6 * Corporation and others. All Rights Reserved.
7 **********************************************************************
8 */
9
10 #include "unicode/utypes.h"
11
12 #if !UCONFIG_NO_CONVERSION
13
14 #include "unicode/ucsdet.h"
15
16 #include "csdetect.h"
17 #include "csmatch.h"
18 #include "uenumimp.h"
19
20 #include "cmemory.h"
21 #include "cstring.h"
22 #include "umutex.h"
23 #include "ucln_in.h"
24 #include "uarrsort.h"
25 #include "inputext.h"
26 #include "csrsbcs.h"
27 #include "csrmbcs.h"
28 #include "csrutf8.h"
29 #include "csrucode.h"
30 #include "csr2022.h"
31
32 #define NEW_ARRAY(type,count) (type *) uprv_malloc((count) * sizeof(type))
33 #define DELETE_ARRAY(array) uprv_free((void *) (array))
34
35 U_NAMESPACE_BEGIN
36
37 struct CSRecognizerInfo : public UMemory {
CSRecognizerInfoCSRecognizerInfo38 CSRecognizerInfo(CharsetRecognizer *recognizer, UBool isDefaultEnabled)
39 : recognizer(recognizer), isDefaultEnabled(isDefaultEnabled) {};
40
~CSRecognizerInfoCSRecognizerInfo41 ~CSRecognizerInfo() {delete recognizer;};
42
43 CharsetRecognizer *recognizer;
44 UBool isDefaultEnabled;
45 };
46
47 U_NAMESPACE_END
48
49 static icu::CSRecognizerInfo **fCSRecognizers = NULL;
50 static icu::UInitOnce gCSRecognizersInitOnce;
51 static int32_t fCSRecognizers_size = 0;
52
53 U_CDECL_BEGIN
csdet_cleanup(void)54 static UBool U_CALLCONV csdet_cleanup(void)
55 {
56 U_NAMESPACE_USE
57 if (fCSRecognizers != NULL) {
58 for(int32_t r = 0; r < fCSRecognizers_size; r += 1) {
59 delete fCSRecognizers[r];
60 fCSRecognizers[r] = NULL;
61 }
62
63 DELETE_ARRAY(fCSRecognizers);
64 fCSRecognizers = NULL;
65 fCSRecognizers_size = 0;
66 }
67 gCSRecognizersInitOnce.reset();
68
69 return TRUE;
70 }
71
72 static int32_t U_CALLCONV
charsetMatchComparator(const void *,const void * left,const void * right)73 charsetMatchComparator(const void * /*context*/, const void *left, const void *right)
74 {
75 U_NAMESPACE_USE
76
77 const CharsetMatch **csm_l = (const CharsetMatch **) left;
78 const CharsetMatch **csm_r = (const CharsetMatch **) right;
79
80 // NOTE: compare is backwards to sort from highest to lowest.
81 return (*csm_r)->getConfidence() - (*csm_l)->getConfidence();
82 }
83
initRecognizers(UErrorCode & status)84 static void U_CALLCONV initRecognizers(UErrorCode &status) {
85 U_NAMESPACE_USE
86 ucln_i18n_registerCleanup(UCLN_I18N_CSDET, csdet_cleanup);
87 CSRecognizerInfo *tempArray[] = {
88 new CSRecognizerInfo(new CharsetRecog_UTF8(), TRUE),
89
90 new CSRecognizerInfo(new CharsetRecog_UTF_16_BE(), TRUE),
91 new CSRecognizerInfo(new CharsetRecog_UTF_16_LE(), TRUE),
92 new CSRecognizerInfo(new CharsetRecog_UTF_32_BE(), TRUE),
93 new CSRecognizerInfo(new CharsetRecog_UTF_32_LE(), TRUE),
94
95 new CSRecognizerInfo(new CharsetRecog_8859_1(), TRUE),
96 new CSRecognizerInfo(new CharsetRecog_8859_2(), TRUE),
97 new CSRecognizerInfo(new CharsetRecog_8859_5_ru(), TRUE),
98 new CSRecognizerInfo(new CharsetRecog_8859_6_ar(), TRUE),
99 new CSRecognizerInfo(new CharsetRecog_8859_7_el(), TRUE),
100 new CSRecognizerInfo(new CharsetRecog_8859_8_I_he(), TRUE),
101 new CSRecognizerInfo(new CharsetRecog_8859_8_he(), TRUE),
102 new CSRecognizerInfo(new CharsetRecog_windows_1251(), TRUE),
103 new CSRecognizerInfo(new CharsetRecog_windows_1256(), TRUE),
104 new CSRecognizerInfo(new CharsetRecog_KOI8_R(), TRUE),
105 new CSRecognizerInfo(new CharsetRecog_8859_9_tr(), TRUE),
106 new CSRecognizerInfo(new CharsetRecog_sjis(), TRUE),
107 new CSRecognizerInfo(new CharsetRecog_gb_18030(), TRUE),
108 new CSRecognizerInfo(new CharsetRecog_euc_jp(), TRUE),
109 new CSRecognizerInfo(new CharsetRecog_euc_kr(), TRUE),
110 new CSRecognizerInfo(new CharsetRecog_big5(), TRUE),
111
112 new CSRecognizerInfo(new CharsetRecog_2022JP(), TRUE),
113 #if !UCONFIG_ONLY_HTML_CONVERSION
114 new CSRecognizerInfo(new CharsetRecog_2022KR(), TRUE),
115 new CSRecognizerInfo(new CharsetRecog_2022CN(), TRUE),
116
117 new CSRecognizerInfo(new CharsetRecog_IBM424_he_rtl(), FALSE),
118 new CSRecognizerInfo(new CharsetRecog_IBM424_he_ltr(), FALSE),
119 new CSRecognizerInfo(new CharsetRecog_IBM420_ar_rtl(), FALSE),
120 new CSRecognizerInfo(new CharsetRecog_IBM420_ar_ltr(), FALSE)
121 #endif
122 };
123 int32_t rCount = UPRV_LENGTHOF(tempArray);
124
125 fCSRecognizers = NEW_ARRAY(CSRecognizerInfo *, rCount);
126
127 if (fCSRecognizers == NULL) {
128 status = U_MEMORY_ALLOCATION_ERROR;
129 }
130 else {
131 fCSRecognizers_size = rCount;
132 for (int32_t r = 0; r < rCount; r += 1) {
133 fCSRecognizers[r] = tempArray[r];
134 if (fCSRecognizers[r] == NULL) {
135 status = U_MEMORY_ALLOCATION_ERROR;
136 }
137 }
138 }
139 }
140
141 U_CDECL_END
142
143 U_NAMESPACE_BEGIN
144
setRecognizers(UErrorCode & status)145 void CharsetDetector::setRecognizers(UErrorCode &status)
146 {
147 umtx_initOnce(gCSRecognizersInitOnce, &initRecognizers, status);
148 }
149
CharsetDetector(UErrorCode & status)150 CharsetDetector::CharsetDetector(UErrorCode &status)
151 : textIn(new InputText(status)), resultArray(NULL),
152 resultCount(0), fStripTags(FALSE), fFreshTextSet(FALSE),
153 fEnabledRecognizers(NULL)
154 {
155 if (U_FAILURE(status)) {
156 return;
157 }
158
159 setRecognizers(status);
160
161 if (U_FAILURE(status)) {
162 return;
163 }
164
165 resultArray = (CharsetMatch **)uprv_malloc(sizeof(CharsetMatch *)*fCSRecognizers_size);
166
167 if (resultArray == NULL) {
168 status = U_MEMORY_ALLOCATION_ERROR;
169 return;
170 }
171
172 for(int32_t i = 0; i < fCSRecognizers_size; i += 1) {
173 resultArray[i] = new CharsetMatch();
174
175 if (resultArray[i] == NULL) {
176 status = U_MEMORY_ALLOCATION_ERROR;
177 break;
178 }
179 }
180 }
181
~CharsetDetector()182 CharsetDetector::~CharsetDetector()
183 {
184 delete textIn;
185
186 for(int32_t i = 0; i < fCSRecognizers_size; i += 1) {
187 delete resultArray[i];
188 }
189
190 uprv_free(resultArray);
191
192 if (fEnabledRecognizers) {
193 uprv_free(fEnabledRecognizers);
194 }
195 }
196
setText(const char * in,int32_t len)197 void CharsetDetector::setText(const char *in, int32_t len)
198 {
199 textIn->setText(in, len);
200 fFreshTextSet = TRUE;
201 }
202
setStripTagsFlag(UBool flag)203 UBool CharsetDetector::setStripTagsFlag(UBool flag)
204 {
205 UBool temp = fStripTags;
206 fStripTags = flag;
207 fFreshTextSet = TRUE;
208 return temp;
209 }
210
getStripTagsFlag() const211 UBool CharsetDetector::getStripTagsFlag() const
212 {
213 return fStripTags;
214 }
215
setDeclaredEncoding(const char * encoding,int32_t len) const216 void CharsetDetector::setDeclaredEncoding(const char *encoding, int32_t len) const
217 {
218 textIn->setDeclaredEncoding(encoding,len);
219 }
220
getDetectableCount()221 int32_t CharsetDetector::getDetectableCount()
222 {
223 UErrorCode status = U_ZERO_ERROR;
224
225 setRecognizers(status);
226
227 return fCSRecognizers_size;
228 }
229
detect(UErrorCode & status)230 const CharsetMatch *CharsetDetector::detect(UErrorCode &status)
231 {
232 int32_t maxMatchesFound = 0;
233
234 detectAll(maxMatchesFound, status);
235
236 if(maxMatchesFound > 0) {
237 return resultArray[0];
238 } else {
239 return NULL;
240 }
241 }
242
detectAll(int32_t & maxMatchesFound,UErrorCode & status)243 const CharsetMatch * const *CharsetDetector::detectAll(int32_t &maxMatchesFound, UErrorCode &status)
244 {
245 if(!textIn->isSet()) {
246 status = U_MISSING_RESOURCE_ERROR;// TODO: Need to set proper status code for input text not set
247
248 return NULL;
249 } else if (fFreshTextSet) {
250 CharsetRecognizer *csr;
251 int32_t i;
252
253 textIn->MungeInput(fStripTags);
254
255 // Iterate over all possible charsets, remember all that
256 // give a match quality > 0.
257 resultCount = 0;
258 for (i = 0; i < fCSRecognizers_size; i += 1) {
259 csr = fCSRecognizers[i]->recognizer;
260 if (csr->match(textIn, resultArray[resultCount])) {
261 resultCount++;
262 }
263 }
264
265 if (resultCount > 1) {
266 uprv_sortArray(resultArray, resultCount, sizeof resultArray[0], charsetMatchComparator, NULL, TRUE, &status);
267 }
268 fFreshTextSet = FALSE;
269 }
270
271 maxMatchesFound = resultCount;
272
273 return resultArray;
274 }
275
setDetectableCharset(const char * encoding,UBool enabled,UErrorCode & status)276 void CharsetDetector::setDetectableCharset(const char *encoding, UBool enabled, UErrorCode &status)
277 {
278 if (U_FAILURE(status)) {
279 return;
280 }
281
282 int32_t modIdx = -1;
283 UBool isDefaultVal = FALSE;
284 for (int32_t i = 0; i < fCSRecognizers_size; i++) {
285 CSRecognizerInfo *csrinfo = fCSRecognizers[i];
286 if (uprv_strcmp(csrinfo->recognizer->getName(), encoding) == 0) {
287 modIdx = i;
288 isDefaultVal = (csrinfo->isDefaultEnabled == enabled);
289 break;
290 }
291 }
292 if (modIdx < 0) {
293 // No matching encoding found
294 status = U_ILLEGAL_ARGUMENT_ERROR;
295 return;
296 }
297
298 if (fEnabledRecognizers == NULL && !isDefaultVal) {
299 // Create an array storing the non default setting
300 fEnabledRecognizers = NEW_ARRAY(UBool, fCSRecognizers_size);
301 if (fEnabledRecognizers == NULL) {
302 status = U_MEMORY_ALLOCATION_ERROR;
303 return;
304 }
305 // Initialize the array with default info
306 for (int32_t i = 0; i < fCSRecognizers_size; i++) {
307 fEnabledRecognizers[i] = fCSRecognizers[i]->isDefaultEnabled;
308 }
309 }
310
311 if (fEnabledRecognizers != NULL) {
312 fEnabledRecognizers[modIdx] = enabled;
313 }
314 }
315
316 /*const char *CharsetDetector::getCharsetName(int32_t index, UErrorCode &status) const
317 {
318 if( index > fCSRecognizers_size-1 || index < 0) {
319 status = U_INDEX_OUTOFBOUNDS_ERROR;
320
321 return 0;
322 } else {
323 return fCSRecognizers[index]->getName();
324 }
325 }*/
326
327 U_NAMESPACE_END
328
329 U_CDECL_BEGIN
330 typedef struct {
331 int32_t currIndex;
332 UBool all;
333 UBool *enabledRecognizers;
334 } Context;
335
336
337
338 static void U_CALLCONV
enumClose(UEnumeration * en)339 enumClose(UEnumeration *en) {
340 if(en->context != NULL) {
341 DELETE_ARRAY(en->context);
342 }
343
344 DELETE_ARRAY(en);
345 }
346
347 static int32_t U_CALLCONV
enumCount(UEnumeration * en,UErrorCode *)348 enumCount(UEnumeration *en, UErrorCode *) {
349 if (((Context *)en->context)->all) {
350 // ucsdet_getAllDetectableCharsets, all charset detector names
351 return fCSRecognizers_size;
352 }
353
354 // Otherwise, ucsdet_getDetectableCharsets - only enabled ones
355 int32_t count = 0;
356 UBool *enabledArray = ((Context *)en->context)->enabledRecognizers;
357 if (enabledArray != NULL) {
358 // custom set
359 for (int32_t i = 0; i < fCSRecognizers_size; i++) {
360 if (enabledArray[i]) {
361 count++;
362 }
363 }
364 } else {
365 // default set
366 for (int32_t i = 0; i < fCSRecognizers_size; i++) {
367 if (fCSRecognizers[i]->isDefaultEnabled) {
368 count++;
369 }
370 }
371 }
372 return count;
373 }
374
375 static const char* U_CALLCONV
enumNext(UEnumeration * en,int32_t * resultLength,UErrorCode *)376 enumNext(UEnumeration *en, int32_t *resultLength, UErrorCode * /*status*/) {
377 const char *currName = NULL;
378
379 if (((Context *)en->context)->currIndex < fCSRecognizers_size) {
380 if (((Context *)en->context)->all) {
381 // ucsdet_getAllDetectableCharsets, all charset detector names
382 currName = fCSRecognizers[((Context *)en->context)->currIndex]->recognizer->getName();
383 ((Context *)en->context)->currIndex++;
384 } else {
385 // ucsdet_getDetectableCharsets
386 UBool *enabledArray = ((Context *)en->context)->enabledRecognizers;
387 if (enabledArray != NULL) {
388 // custome set
389 while (currName == NULL && ((Context *)en->context)->currIndex < fCSRecognizers_size) {
390 if (enabledArray[((Context *)en->context)->currIndex]) {
391 currName = fCSRecognizers[((Context *)en->context)->currIndex]->recognizer->getName();
392 }
393 ((Context *)en->context)->currIndex++;
394 }
395 } else {
396 // default set
397 while (currName == NULL && ((Context *)en->context)->currIndex < fCSRecognizers_size) {
398 if (fCSRecognizers[((Context *)en->context)->currIndex]->isDefaultEnabled) {
399 currName = fCSRecognizers[((Context *)en->context)->currIndex]->recognizer->getName();
400 }
401 ((Context *)en->context)->currIndex++;
402 }
403 }
404 }
405 }
406
407 if(resultLength != NULL) {
408 *resultLength = currName == NULL ? 0 : (int32_t)uprv_strlen(currName);
409 }
410
411 return currName;
412 }
413
414
415 static void U_CALLCONV
enumReset(UEnumeration * en,UErrorCode *)416 enumReset(UEnumeration *en, UErrorCode *) {
417 ((Context *)en->context)->currIndex = 0;
418 }
419
420 static const UEnumeration gCSDetEnumeration = {
421 NULL,
422 NULL,
423 enumClose,
424 enumCount,
425 uenum_unextDefault,
426 enumNext,
427 enumReset
428 };
429
430 U_CDECL_END
431
432 U_NAMESPACE_BEGIN
433
getAllDetectableCharsets(UErrorCode & status)434 UEnumeration * CharsetDetector::getAllDetectableCharsets(UErrorCode &status)
435 {
436
437 /* Initialize recognized charsets. */
438 setRecognizers(status);
439
440 if(U_FAILURE(status)) {
441 return 0;
442 }
443
444 UEnumeration *en = NEW_ARRAY(UEnumeration, 1);
445 if (en == NULL) {
446 status = U_MEMORY_ALLOCATION_ERROR;
447 return 0;
448 }
449 memcpy(en, &gCSDetEnumeration, sizeof(UEnumeration));
450 en->context = (void*)NEW_ARRAY(Context, 1);
451 if (en->context == NULL) {
452 status = U_MEMORY_ALLOCATION_ERROR;
453 DELETE_ARRAY(en);
454 return 0;
455 }
456 uprv_memset(en->context, 0, sizeof(Context));
457 ((Context*)en->context)->all = TRUE;
458 return en;
459 }
460
getDetectableCharsets(UErrorCode & status) const461 UEnumeration * CharsetDetector::getDetectableCharsets(UErrorCode &status) const
462 {
463 if(U_FAILURE(status)) {
464 return 0;
465 }
466
467 UEnumeration *en = NEW_ARRAY(UEnumeration, 1);
468 if (en == NULL) {
469 status = U_MEMORY_ALLOCATION_ERROR;
470 return 0;
471 }
472 memcpy(en, &gCSDetEnumeration, sizeof(UEnumeration));
473 en->context = (void*)NEW_ARRAY(Context, 1);
474 if (en->context == NULL) {
475 status = U_MEMORY_ALLOCATION_ERROR;
476 DELETE_ARRAY(en);
477 return 0;
478 }
479 uprv_memset(en->context, 0, sizeof(Context));
480 ((Context*)en->context)->all = FALSE;
481 ((Context*)en->context)->enabledRecognizers = fEnabledRecognizers;
482 return en;
483 }
484
485 U_NAMESPACE_END
486
487 #endif
488