LLVM  6.0.0svn
ConvertUTF.cpp
Go to the documentation of this file.
1 /*===--- ConvertUTF.c - Universal Character Names conversions ---------------===
2  *
3  * The LLVM Compiler Infrastructure
4  *
5  * This file is distributed under the University of Illinois Open Source
6  * License. See LICENSE.TXT for details.
7  *
8  *===------------------------------------------------------------------------=*/
9 /*
10  * Copyright 2001-2004 Unicode, Inc.
11  *
12  * Disclaimer
13  *
14  * This source code is provided as is by Unicode, Inc. No claims are
15  * made as to fitness for any particular purpose. No warranties of any
16  * kind are expressed or implied. The recipient agrees to determine
17  * applicability of information provided. If this file has been
18  * purchased on magnetic or optical media from Unicode, Inc., the
19  * sole remedy for any claim will be exchange of defective media
20  * within 90 days of receipt.
21  *
22  * Limitations on Rights to Redistribute This Code
23  *
24  * Unicode, Inc. hereby grants the right to freely use the information
25  * supplied in this file in the creation of products supporting the
26  * Unicode Standard, and to make copies of this file in any form
27  * for internal or external distribution as long as this notice
28  * remains attached.
29  */
30 
31 /* ---------------------------------------------------------------------
32 
33  Conversions between UTF32, UTF-16, and UTF-8. Source code file.
34  Author: Mark E. Davis, 1994.
35  Rev History: Rick McGowan, fixes & updates May 2001.
36  Sept 2001: fixed const & error conditions per
37  mods suggested by S. Parent & A. Lillich.
38  June 2002: Tim Dodd added detection and handling of incomplete
39  source sequences, enhanced error detection, added casts
40  to eliminate compiler warnings.
41  July 2003: slight mods to back out aggressive FFFE detection.
42  Jan 2004: updated switches in from-UTF8 conversions.
43  Oct 2004: updated to use UNI_MAX_LEGAL_UTF32 in UTF-32 conversions.
44 
45  See the header file "ConvertUTF.h" for complete documentation.
46 
47 ------------------------------------------------------------------------ */
48 
50 #ifdef CVTUTF_DEBUG
51 #include <stdio.h>
52 #endif
53 #include <assert.h>
54 
55 /*
56  * This code extensively uses fall-through switches.
57  * Keep the compiler from warning about that.
58  */
59 #if defined(__clang__) && defined(__has_warning)
60 # if __has_warning("-Wimplicit-fallthrough")
61 # define ConvertUTF_DISABLE_WARNINGS \
62  _Pragma("clang diagnostic push") \
63  _Pragma("clang diagnostic ignored \"-Wimplicit-fallthrough\"")
64 # define ConvertUTF_RESTORE_WARNINGS \
65  _Pragma("clang diagnostic pop")
66 # endif
67 #elif defined(__GNUC__) && __GNUC__ > 6
68 # define ConvertUTF_DISABLE_WARNINGS \
69  _Pragma("GCC diagnostic push") \
70  _Pragma("GCC diagnostic ignored \"-Wimplicit-fallthrough\"")
71 # define ConvertUTF_RESTORE_WARNINGS \
72  _Pragma("GCC diagnostic pop")
73 #endif
74 #ifndef ConvertUTF_DISABLE_WARNINGS
75 # define ConvertUTF_DISABLE_WARNINGS
76 #endif
77 #ifndef ConvertUTF_RESTORE_WARNINGS
78 # define ConvertUTF_RESTORE_WARNINGS
79 #endif
80 
82 
83 namespace llvm {
84 
85 static const int halfShift = 10; /* used for shifting by 10 bits */
86 
87 static const UTF32 halfBase = 0x0010000UL;
88 static const UTF32 halfMask = 0x3FFUL;
89 
90 #define UNI_SUR_HIGH_START (UTF32)0xD800
91 #define UNI_SUR_HIGH_END (UTF32)0xDBFF
92 #define UNI_SUR_LOW_START (UTF32)0xDC00
93 #define UNI_SUR_LOW_END (UTF32)0xDFFF
94 
95 /* --------------------------------------------------------------------- */
96 
97 /*
98  * Index into the table below with the first byte of a UTF-8 sequence to
99  * get the number of trailing bytes that are supposed to follow it.
100  * Note that *legal* UTF-8 values can't have 4 or 5-bytes. The table is
101  * left as-is for anyone who may want to do such conversion, which was
102  * allowed in earlier algorithms.
103  */
104 static const char trailingBytesForUTF8[256] = {
105  0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
106  0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
107  0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
108  0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
109  0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
110  0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
111  1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
112  2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2, 3,3,3,3,3,3,3,3,4,4,4,4,5,5,5,5
113 };
114 
115 /*
116  * Magic values subtracted from a buffer value during UTF8 conversion.
117  * This table contains as many values as there might be trailing bytes
118  * in a UTF-8 sequence.
119  */
120 static const UTF32 offsetsFromUTF8[6] = { 0x00000000UL, 0x00003080UL, 0x000E2080UL,
121  0x03C82080UL, 0xFA082080UL, 0x82082080UL };
122 
123 /*
124  * Once the bits are split out into bytes of UTF-8, this is a mask OR-ed
125  * into the first byte, depending on how many bytes follow. There are
126  * as many entries in this table as there are UTF-8 sequence types.
127  * (I.e., one byte sequence, two byte... etc.). Remember that sequencs
128  * for *legal* UTF-8 will be 4 or fewer bytes total.
129  */
130 static const UTF8 firstByteMark[7] = { 0x00, 0x00, 0xC0, 0xE0, 0xF0, 0xF8, 0xFC };
131 
132 /* --------------------------------------------------------------------- */
133 
134 /* The interface converts a whole buffer to avoid function-call overhead.
135  * Constants have been gathered. Loops & conditionals have been removed as
136  * much as possible for efficiency, in favor of drop-through switches.
137  * (See "Note A" at the bottom of the file for equivalent code.)
138  * If your compiler supports it, the "isLegalUTF8" call can be turned
139  * into an inline function.
140  */
141 
142 
143 /* --------------------------------------------------------------------- */
144 
146  const UTF32** sourceStart, const UTF32* sourceEnd,
147  UTF16** targetStart, UTF16* targetEnd, ConversionFlags flags) {
149  const UTF32* source = *sourceStart;
150  UTF16* target = *targetStart;
151  while (source < sourceEnd) {
152  UTF32 ch;
153  if (target >= targetEnd) {
154  result = targetExhausted; break;
155  }
156  ch = *source++;
157  if (ch <= UNI_MAX_BMP) { /* Target is a character <= 0xFFFF */
158  /* UTF-16 surrogate values are illegal in UTF-32; 0xffff or 0xfffe are both reserved values */
159  if (ch >= UNI_SUR_HIGH_START && ch <= UNI_SUR_LOW_END) {
160  if (flags == strictConversion) {
161  --source; /* return to the illegal value itself */
162  result = sourceIllegal;
163  break;
164  } else {
165  *target++ = UNI_REPLACEMENT_CHAR;
166  }
167  } else {
168  *target++ = (UTF16)ch; /* normal case */
169  }
170  } else if (ch > UNI_MAX_LEGAL_UTF32) {
171  if (flags == strictConversion) {
172  result = sourceIllegal;
173  } else {
174  *target++ = UNI_REPLACEMENT_CHAR;
175  }
176  } else {
177  /* target is a character in range 0xFFFF - 0x10FFFF. */
178  if (target + 1 >= targetEnd) {
179  --source; /* Back up source pointer! */
180  result = targetExhausted; break;
181  }
182  ch -= halfBase;
183  *target++ = (UTF16)((ch >> halfShift) + UNI_SUR_HIGH_START);
184  *target++ = (UTF16)((ch & halfMask) + UNI_SUR_LOW_START);
185  }
186  }
187  *sourceStart = source;
188  *targetStart = target;
189  return result;
190 }
191 
192 /* --------------------------------------------------------------------- */
193 
195  const UTF16** sourceStart, const UTF16* sourceEnd,
196  UTF32** targetStart, UTF32* targetEnd, ConversionFlags flags) {
198  const UTF16* source = *sourceStart;
199  UTF32* target = *targetStart;
200  UTF32 ch, ch2;
201  while (source < sourceEnd) {
202  const UTF16* oldSource = source; /* In case we have to back up because of target overflow. */
203  ch = *source++;
204  /* If we have a surrogate pair, convert to UTF32 first. */
205  if (ch >= UNI_SUR_HIGH_START && ch <= UNI_SUR_HIGH_END) {
206  /* If the 16 bits following the high surrogate are in the source buffer... */
207  if (source < sourceEnd) {
208  ch2 = *source;
209  /* If it's a low surrogate, convert to UTF32. */
210  if (ch2 >= UNI_SUR_LOW_START && ch2 <= UNI_SUR_LOW_END) {
211  ch = ((ch - UNI_SUR_HIGH_START) << halfShift)
212  + (ch2 - UNI_SUR_LOW_START) + halfBase;
213  ++source;
214  } else if (flags == strictConversion) { /* it's an unpaired high surrogate */
215  --source; /* return to the illegal value itself */
216  result = sourceIllegal;
217  break;
218  }
219  } else { /* We don't have the 16 bits following the high surrogate. */
220  --source; /* return to the high surrogate */
221  result = sourceExhausted;
222  break;
223  }
224  } else if (flags == strictConversion) {
225  /* UTF-16 surrogate values are illegal in UTF-32 */
226  if (ch >= UNI_SUR_LOW_START && ch <= UNI_SUR_LOW_END) {
227  --source; /* return to the illegal value itself */
228  result = sourceIllegal;
229  break;
230  }
231  }
232  if (target >= targetEnd) {
233  source = oldSource; /* Back up source pointer! */
234  result = targetExhausted; break;
235  }
236  *target++ = ch;
237  }
238  *sourceStart = source;
239  *targetStart = target;
240 #ifdef CVTUTF_DEBUG
241 if (result == sourceIllegal) {
242  fprintf(stderr, "ConvertUTF16toUTF32 illegal seq 0x%04x,%04x\n", ch, ch2);
243  fflush(stderr);
244 }
245 #endif
246  return result;
247 }
249  const UTF16** sourceStart, const UTF16* sourceEnd,
250  UTF8** targetStart, UTF8* targetEnd, ConversionFlags flags) {
252  const UTF16* source = *sourceStart;
253  UTF8* target = *targetStart;
254  while (source < sourceEnd) {
255  UTF32 ch;
256  unsigned short bytesToWrite = 0;
257  const UTF32 byteMask = 0xBF;
258  const UTF32 byteMark = 0x80;
259  const UTF16* oldSource = source; /* In case we have to back up because of target overflow. */
260  ch = *source++;
261  /* If we have a surrogate pair, convert to UTF32 first. */
262  if (ch >= UNI_SUR_HIGH_START && ch <= UNI_SUR_HIGH_END) {
263  /* If the 16 bits following the high surrogate are in the source buffer... */
264  if (source < sourceEnd) {
265  UTF32 ch2 = *source;
266  /* If it's a low surrogate, convert to UTF32. */
267  if (ch2 >= UNI_SUR_LOW_START && ch2 <= UNI_SUR_LOW_END) {
268  ch = ((ch - UNI_SUR_HIGH_START) << halfShift)
269  + (ch2 - UNI_SUR_LOW_START) + halfBase;
270  ++source;
271  } else if (flags == strictConversion) { /* it's an unpaired high surrogate */
272  --source; /* return to the illegal value itself */
273  result = sourceIllegal;
274  break;
275  }
276  } else { /* We don't have the 16 bits following the high surrogate. */
277  --source; /* return to the high surrogate */
278  result = sourceExhausted;
279  break;
280  }
281  } else if (flags == strictConversion) {
282  /* UTF-16 surrogate values are illegal in UTF-32 */
283  if (ch >= UNI_SUR_LOW_START && ch <= UNI_SUR_LOW_END) {
284  --source; /* return to the illegal value itself */
285  result = sourceIllegal;
286  break;
287  }
288  }
289  /* Figure out how many bytes the result will require */
290  if (ch < (UTF32)0x80) { bytesToWrite = 1;
291  } else if (ch < (UTF32)0x800) { bytesToWrite = 2;
292  } else if (ch < (UTF32)0x10000) { bytesToWrite = 3;
293  } else if (ch < (UTF32)0x110000) { bytesToWrite = 4;
294  } else { bytesToWrite = 3;
296  }
297 
298  target += bytesToWrite;
299  if (target > targetEnd) {
300  source = oldSource; /* Back up source pointer! */
301  target -= bytesToWrite; result = targetExhausted; break;
302  }
303  switch (bytesToWrite) { /* note: everything falls through. */
304  case 4: *--target = (UTF8)((ch | byteMark) & byteMask); ch >>= 6;
305  case 3: *--target = (UTF8)((ch | byteMark) & byteMask); ch >>= 6;
306  case 2: *--target = (UTF8)((ch | byteMark) & byteMask); ch >>= 6;
307  case 1: *--target = (UTF8)(ch | firstByteMark[bytesToWrite]);
308  }
309  target += bytesToWrite;
310  }
311  *sourceStart = source;
312  *targetStart = target;
313  return result;
314 }
315 
316 /* --------------------------------------------------------------------- */
317 
319  const UTF32** sourceStart, const UTF32* sourceEnd,
320  UTF8** targetStart, UTF8* targetEnd, ConversionFlags flags) {
322  const UTF32* source = *sourceStart;
323  UTF8* target = *targetStart;
324  while (source < sourceEnd) {
325  UTF32 ch;
326  unsigned short bytesToWrite = 0;
327  const UTF32 byteMask = 0xBF;
328  const UTF32 byteMark = 0x80;
329  ch = *source++;
330  if (flags == strictConversion ) {
331  /* UTF-16 surrogate values are illegal in UTF-32 */
332  if (ch >= UNI_SUR_HIGH_START && ch <= UNI_SUR_LOW_END) {
333  --source; /* return to the illegal value itself */
334  result = sourceIllegal;
335  break;
336  }
337  }
338  /*
339  * Figure out how many bytes the result will require. Turn any
340  * illegally large UTF32 things (> Plane 17) into replacement chars.
341  */
342  if (ch < (UTF32)0x80) { bytesToWrite = 1;
343  } else if (ch < (UTF32)0x800) { bytesToWrite = 2;
344  } else if (ch < (UTF32)0x10000) { bytesToWrite = 3;
345  } else if (ch <= UNI_MAX_LEGAL_UTF32) { bytesToWrite = 4;
346  } else { bytesToWrite = 3;
348  result = sourceIllegal;
349  }
350 
351  target += bytesToWrite;
352  if (target > targetEnd) {
353  --source; /* Back up source pointer! */
354  target -= bytesToWrite; result = targetExhausted; break;
355  }
356  switch (bytesToWrite) { /* note: everything falls through. */
357  case 4: *--target = (UTF8)((ch | byteMark) & byteMask); ch >>= 6;
358  case 3: *--target = (UTF8)((ch | byteMark) & byteMask); ch >>= 6;
359  case 2: *--target = (UTF8)((ch | byteMark) & byteMask); ch >>= 6;
360  case 1: *--target = (UTF8) (ch | firstByteMark[bytesToWrite]);
361  }
362  target += bytesToWrite;
363  }
364  *sourceStart = source;
365  *targetStart = target;
366  return result;
367 }
368 
369 /* --------------------------------------------------------------------- */
370 
371 /*
372  * Utility routine to tell whether a sequence of bytes is legal UTF-8.
373  * This must be called with the length pre-determined by the first byte.
374  * If not calling this from ConvertUTF8to*, then the length can be set by:
375  * length = trailingBytesForUTF8[*source]+1;
376  * and the sequence is illegal right away if there aren't that many bytes
377  * available.
378  * If presented with a length > 4, this returns false. The Unicode
379  * definition of UTF-8 goes up to 4-byte sequences.
380  */
381 
382 static Boolean isLegalUTF8(const UTF8 *source, int length) {
383  UTF8 a;
384  const UTF8 *srcptr = source+length;
385  switch (length) {
386  default: return false;
387  /* Everything else falls through when "true"... */
388  case 4: if ((a = (*--srcptr)) < 0x80 || a > 0xBF) return false;
389  case 3: if ((a = (*--srcptr)) < 0x80 || a > 0xBF) return false;
390  case 2: if ((a = (*--srcptr)) < 0x80 || a > 0xBF) return false;
391 
392  switch (*source) {
393  /* no fall-through in this inner switch */
394  case 0xE0: if (a < 0xA0) return false; break;
395  case 0xED: if (a > 0x9F) return false; break;
396  case 0xF0: if (a < 0x90) return false; break;
397  case 0xF4: if (a > 0x8F) return false; break;
398  default: if (a < 0x80) return false;
399  }
400 
401  case 1: if (*source >= 0x80 && *source < 0xC2) return false;
402  }
403  if (*source > 0xF4) return false;
404  return true;
405 }
406 
407 /* --------------------------------------------------------------------- */
408 
409 /*
410  * Exported function to return whether a UTF-8 sequence is legal or not.
411  * This is not used here; it's just exported.
412  */
413 Boolean isLegalUTF8Sequence(const UTF8 *source, const UTF8 *sourceEnd) {
414  int length = trailingBytesForUTF8[*source]+1;
415  if (length > sourceEnd - source) {
416  return false;
417  }
418  return isLegalUTF8(source, length);
419 }
420 
421 /* --------------------------------------------------------------------- */
422 
423 static unsigned
425  const UTF8 *sourceEnd) {
426  UTF8 b1, b2, b3;
427 
428  assert(!isLegalUTF8Sequence(source, sourceEnd));
429 
430  /*
431  * Unicode 6.3.0, D93b:
432  *
433  * Maximal subpart of an ill-formed subsequence: The longest code unit
434  * subsequence starting at an unconvertible offset that is either:
435  * a. the initial subsequence of a well-formed code unit sequence, or
436  * b. a subsequence of length one.
437  */
438 
439  if (source == sourceEnd)
440  return 0;
441 
442  /*
443  * Perform case analysis. See Unicode 6.3.0, Table 3-7. Well-Formed UTF-8
444  * Byte Sequences.
445  */
446 
447  b1 = *source;
448  ++source;
449  if (b1 >= 0xC2 && b1 <= 0xDF) {
450  /*
451  * First byte is valid, but we know that this code unit sequence is
452  * invalid, so the maximal subpart has to end after the first byte.
453  */
454  return 1;
455  }
456 
457  if (source == sourceEnd)
458  return 1;
459 
460  b2 = *source;
461  ++source;
462 
463  if (b1 == 0xE0) {
464  return (b2 >= 0xA0 && b2 <= 0xBF) ? 2 : 1;
465  }
466  if (b1 >= 0xE1 && b1 <= 0xEC) {
467  return (b2 >= 0x80 && b2 <= 0xBF) ? 2 : 1;
468  }
469  if (b1 == 0xED) {
470  return (b2 >= 0x80 && b2 <= 0x9F) ? 2 : 1;
471  }
472  if (b1 >= 0xEE && b1 <= 0xEF) {
473  return (b2 >= 0x80 && b2 <= 0xBF) ? 2 : 1;
474  }
475  if (b1 == 0xF0) {
476  if (b2 >= 0x90 && b2 <= 0xBF) {
477  if (source == sourceEnd)
478  return 2;
479 
480  b3 = *source;
481  return (b3 >= 0x80 && b3 <= 0xBF) ? 3 : 2;
482  }
483  return 1;
484  }
485  if (b1 >= 0xF1 && b1 <= 0xF3) {
486  if (b2 >= 0x80 && b2 <= 0xBF) {
487  if (source == sourceEnd)
488  return 2;
489 
490  b3 = *source;
491  return (b3 >= 0x80 && b3 <= 0xBF) ? 3 : 2;
492  }
493  return 1;
494  }
495  if (b1 == 0xF4) {
496  if (b2 >= 0x80 && b2 <= 0x8F) {
497  if (source == sourceEnd)
498  return 2;
499 
500  b3 = *source;
501  return (b3 >= 0x80 && b3 <= 0xBF) ? 3 : 2;
502  }
503  return 1;
504  }
505 
506  assert((b1 >= 0x80 && b1 <= 0xC1) || b1 >= 0xF5);
507  /*
508  * There are no valid sequences that start with these bytes. Maximal subpart
509  * is defined to have length 1 in these cases.
510  */
511  return 1;
512 }
513 
514 /* --------------------------------------------------------------------- */
515 
516 /*
517  * Exported function to return the total number of bytes in a codepoint
518  * represented in UTF-8, given the value of the first byte.
519  */
521  return trailingBytesForUTF8[first] + 1;
522 }
523 
524 /* --------------------------------------------------------------------- */
525 
526 /*
527  * Exported function to return whether a UTF-8 string is legal or not.
528  * This is not used here; it's just exported.
529  */
530 Boolean isLegalUTF8String(const UTF8 **source, const UTF8 *sourceEnd) {
531  while (*source != sourceEnd) {
532  int length = trailingBytesForUTF8[**source] + 1;
533  if (length > sourceEnd - *source || !isLegalUTF8(*source, length))
534  return false;
535  *source += length;
536  }
537  return true;
538 }
539 
540 /* --------------------------------------------------------------------- */
541 
543  const UTF8** sourceStart, const UTF8* sourceEnd,
544  UTF16** targetStart, UTF16* targetEnd, ConversionFlags flags) {
546  const UTF8* source = *sourceStart;
547  UTF16* target = *targetStart;
548  while (source < sourceEnd) {
549  UTF32 ch = 0;
550  unsigned short extraBytesToRead = trailingBytesForUTF8[*source];
551  if (extraBytesToRead >= sourceEnd - source) {
552  result = sourceExhausted; break;
553  }
554  /* Do this check whether lenient or strict */
555  if (!isLegalUTF8(source, extraBytesToRead+1)) {
556  result = sourceIllegal;
557  break;
558  }
559  /*
560  * The cases all fall through. See "Note A" below.
561  */
562  switch (extraBytesToRead) {
563  case 5: ch += *source++; ch <<= 6; /* remember, illegal UTF-8 */
564  case 4: ch += *source++; ch <<= 6; /* remember, illegal UTF-8 */
565  case 3: ch += *source++; ch <<= 6;
566  case 2: ch += *source++; ch <<= 6;
567  case 1: ch += *source++; ch <<= 6;
568  case 0: ch += *source++;
569  }
570  ch -= offsetsFromUTF8[extraBytesToRead];
571 
572  if (target >= targetEnd) {
573  source -= (extraBytesToRead+1); /* Back up source pointer! */
574  result = targetExhausted; break;
575  }
576  if (ch <= UNI_MAX_BMP) { /* Target is a character <= 0xFFFF */
577  /* UTF-16 surrogate values are illegal in UTF-32 */
578  if (ch >= UNI_SUR_HIGH_START && ch <= UNI_SUR_LOW_END) {
579  if (flags == strictConversion) {
580  source -= (extraBytesToRead+1); /* return to the illegal value itself */
581  result = sourceIllegal;
582  break;
583  } else {
584  *target++ = UNI_REPLACEMENT_CHAR;
585  }
586  } else {
587  *target++ = (UTF16)ch; /* normal case */
588  }
589  } else if (ch > UNI_MAX_UTF16) {
590  if (flags == strictConversion) {
591  result = sourceIllegal;
592  source -= (extraBytesToRead+1); /* return to the start */
593  break; /* Bail out; shouldn't continue */
594  } else {
595  *target++ = UNI_REPLACEMENT_CHAR;
596  }
597  } else {
598  /* target is a character in range 0xFFFF - 0x10FFFF. */
599  if (target + 1 >= targetEnd) {
600  source -= (extraBytesToRead+1); /* Back up source pointer! */
601  result = targetExhausted; break;
602  }
603  ch -= halfBase;
604  *target++ = (UTF16)((ch >> halfShift) + UNI_SUR_HIGH_START);
605  *target++ = (UTF16)((ch & halfMask) + UNI_SUR_LOW_START);
606  }
607  }
608  *sourceStart = source;
609  *targetStart = target;
610  return result;
611 }
612 
613 /* --------------------------------------------------------------------- */
614 
616  const UTF8** sourceStart, const UTF8* sourceEnd,
617  UTF32** targetStart, UTF32* targetEnd, ConversionFlags flags,
618  Boolean InputIsPartial) {
620  const UTF8* source = *sourceStart;
621  UTF32* target = *targetStart;
622  while (source < sourceEnd) {
623  UTF32 ch = 0;
624  unsigned short extraBytesToRead = trailingBytesForUTF8[*source];
625  if (extraBytesToRead >= sourceEnd - source) {
626  if (flags == strictConversion || InputIsPartial) {
627  result = sourceExhausted;
628  break;
629  } else {
630  result = sourceIllegal;
631 
632  /*
633  * Replace the maximal subpart of ill-formed sequence with
634  * replacement character.
635  */
637  sourceEnd);
638  *target++ = UNI_REPLACEMENT_CHAR;
639  continue;
640  }
641  }
642  if (target >= targetEnd) {
643  result = targetExhausted; break;
644  }
645 
646  /* Do this check whether lenient or strict */
647  if (!isLegalUTF8(source, extraBytesToRead+1)) {
648  result = sourceIllegal;
649  if (flags == strictConversion) {
650  /* Abort conversion. */
651  break;
652  } else {
653  /*
654  * Replace the maximal subpart of ill-formed sequence with
655  * replacement character.
656  */
658  sourceEnd);
659  *target++ = UNI_REPLACEMENT_CHAR;
660  continue;
661  }
662  }
663  /*
664  * The cases all fall through. See "Note A" below.
665  */
666  switch (extraBytesToRead) {
667  case 5: ch += *source++; ch <<= 6;
668  case 4: ch += *source++; ch <<= 6;
669  case 3: ch += *source++; ch <<= 6;
670  case 2: ch += *source++; ch <<= 6;
671  case 1: ch += *source++; ch <<= 6;
672  case 0: ch += *source++;
673  }
674  ch -= offsetsFromUTF8[extraBytesToRead];
675 
676  if (ch <= UNI_MAX_LEGAL_UTF32) {
677  /*
678  * UTF-16 surrogate values are illegal in UTF-32, and anything
679  * over Plane 17 (> 0x10FFFF) is illegal.
680  */
681  if (ch >= UNI_SUR_HIGH_START && ch <= UNI_SUR_LOW_END) {
682  if (flags == strictConversion) {
683  source -= (extraBytesToRead+1); /* return to the illegal value itself */
684  result = sourceIllegal;
685  break;
686  } else {
687  *target++ = UNI_REPLACEMENT_CHAR;
688  }
689  } else {
690  *target++ = ch;
691  }
692  } else { /* i.e., ch > UNI_MAX_LEGAL_UTF32 */
693  result = sourceIllegal;
694  *target++ = UNI_REPLACEMENT_CHAR;
695  }
696  }
697  *sourceStart = source;
698  *targetStart = target;
699  return result;
700 }
701 
703  const UTF8 *sourceEnd,
704  UTF32 **targetStart,
705  UTF32 *targetEnd,
706  ConversionFlags flags) {
707  return ConvertUTF8toUTF32Impl(sourceStart, sourceEnd, targetStart, targetEnd,
708  flags, /*InputIsPartial=*/true);
709 }
710 
712  const UTF8 *sourceEnd, UTF32 **targetStart,
713  UTF32 *targetEnd, ConversionFlags flags) {
714  return ConvertUTF8toUTF32Impl(sourceStart, sourceEnd, targetStart, targetEnd,
715  flags, /*InputIsPartial=*/false);
716 }
717 
718 /* ---------------------------------------------------------------------
719 
720  Note A.
721  The fall-through switches in UTF-8 reading code save a
722  temp variable, some decrements & conditionals. The switches
723  are equivalent to the following loop:
724  {
725  int tmpBytesToRead = extraBytesToRead+1;
726  do {
727  ch += *source++;
728  --tmpBytesToRead;
729  if (tmpBytesToRead) ch <<= 6;
730  } while (tmpBytesToRead > 0);
731  }
732  In UTF-8 writing code, the switches on "bytesToWrite" are
733  similarly unrolled loops.
734 
735  --------------------------------------------------------------------- */
736 
737 } // namespace llvm
738 
unsigned int UTF32
Definition: ConvertUTF.h:109
static const UTF32 halfBase
Definition: ConvertUTF.cpp:87
Compute iterated dominance frontiers using a linear time algorithm.
Definition: AllocatorList.h:24
#define UNI_MAX_BMP
Definition: ConvertUTF.h:116
ConversionResult
Definition: ConvertUTF.h:126
static const char trailingBytesForUTF8[256]
Definition: ConvertUTF.cpp:104
#define UNI_MAX_UTF16
Definition: ConvertUTF.h:117
unsigned short UTF16
Definition: ConvertUTF.h:110
ConversionResult ConvertUTF8toUTF32(const UTF8 **sourceStart, const UTF8 *sourceEnd, UTF32 **targetStart, UTF32 *targetEnd, ConversionFlags flags)
Convert a partial UTF8 sequence to UTF32.
Definition: ConvertUTF.cpp:711
#define UNI_MAX_LEGAL_UTF32
Definition: ConvertUTF.h:119
#define UNI_REPLACEMENT_CHAR
Definition: ConvertUTF.h:115
ConversionResult ConvertUTF16toUTF8(const UTF16 **sourceStart, const UTF16 *sourceEnd, UTF8 **targetStart, UTF8 *targetEnd, ConversionFlags flags)
Definition: ConvertUTF.cpp:248
static ConversionResult ConvertUTF8toUTF32Impl(const UTF8 **sourceStart, const UTF8 *sourceEnd, UTF32 **targetStart, UTF32 *targetEnd, ConversionFlags flags, Boolean InputIsPartial)
Definition: ConvertUTF.cpp:615
ConversionResult ConvertUTF32toUTF16(const UTF32 **sourceStart, const UTF32 *sourceEnd, UTF16 **targetStart, UTF16 *targetEnd, ConversionFlags flags)
Definition: ConvertUTF.cpp:145
unsigned char UTF8
Definition: ConvertUTF.h:111
unsigned getNumBytesForUTF8(UTF8 firstByte)
Definition: ConvertUTF.cpp:520
#define UNI_SUR_HIGH_END
Definition: ConvertUTF.cpp:91
static Boolean isLegalUTF8(const UTF8 *source, int length)
Definition: ConvertUTF.cpp:382
#define ConvertUTF_DISABLE_WARNINGS
Definition: ConvertUTF.cpp:75
ConversionResult ConvertUTF8toUTF16(const UTF8 **sourceStart, const UTF8 *sourceEnd, UTF16 **targetStart, UTF16 *targetEnd, ConversionFlags flags)
Definition: ConvertUTF.cpp:542
ConversionResult ConvertUTF32toUTF8(const UTF32 **sourceStart, const UTF32 *sourceEnd, UTF8 **targetStart, UTF8 *targetEnd, ConversionFlags flags)
Definition: ConvertUTF.cpp:318
Boolean isLegalUTF8Sequence(const UTF8 *source, const UTF8 *sourceEnd)
Definition: ConvertUTF.cpp:413
Boolean isLegalUTF8String(const UTF8 **source, const UTF8 *sourceEnd)
Definition: ConvertUTF.cpp:530
ConversionResult ConvertUTF8toUTF32Partial(const UTF8 **sourceStart, const UTF8 *sourceEnd, UTF32 **targetStart, UTF32 *targetEnd, ConversionFlags flags)
Convert a partial UTF8 sequence to UTF32.
Definition: ConvertUTF.cpp:702
static const int halfShift
Definition: ConvertUTF.cpp:85
unsigned char Boolean
Definition: ConvertUTF.h:112
static const UTF32 offsetsFromUTF8[6]
Definition: ConvertUTF.cpp:120
unsigned first
ConversionFlags
Definition: ConvertUTF.h:133
#define UNI_SUR_LOW_START
Definition: ConvertUTF.cpp:92
ConversionResult ConvertUTF16toUTF32(const UTF16 **sourceStart, const UTF16 *sourceEnd, UTF32 **targetStart, UTF32 *targetEnd, ConversionFlags flags)
Definition: ConvertUTF.cpp:194
static unsigned findMaximalSubpartOfIllFormedUTF8Sequence(const UTF8 *source, const UTF8 *sourceEnd)
Definition: ConvertUTF.cpp:424
static const UTF8 firstByteMark[7]
Definition: ConvertUTF.cpp:130
#define ConvertUTF_RESTORE_WARNINGS
Definition: ConvertUTF.cpp:78
assert(ImpDefSCC.getReg()==AMDGPU::SCC &&ImpDefSCC.isDef())
#define UNI_SUR_HIGH_START
Definition: ConvertUTF.cpp:90
#define UNI_SUR_LOW_END
Definition: ConvertUTF.cpp:93
static const UTF32 halfMask
Definition: ConvertUTF.cpp:88