/[pcre]/code/trunk/pcre_valid_utf8.c
ViewVC logotype

Diff of /code/trunk/pcre_valid_utf8.c

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 605 by ph10, Sat May 7 15:37:31 2011 UTC revision 606 by ph10, Mon Jun 6 17:46:22 2011 UTC
# Line 68  subrange 0xd000 to 0xdfff is excluded. H Line 68  subrange 0xd000 to 0xdfff is excluded. H
68  characters is still checked.  characters is still checked.
69    
70  From release 8.13 more information about the details of the error are passed  From release 8.13 more information about the details of the error are passed
71  back in the error code:  back in the returned value:
72    
73  PCRE_UTF8_ERR0   No error  PCRE_UTF8_ERR0   No error
74  PCRE_UTF8_ERR1   Missing 1 byte at the end of the string  PCRE_UTF8_ERR1   Missing 1 byte at the end of the string
# Line 96  PCRE_UTF8_ERR21 Byte with the illegal v Line 96  PCRE_UTF8_ERR21 Byte with the illegal v
96  Arguments:  Arguments:
97    string       points to the string    string       points to the string
98    length       length of string, or -1 if the string is zero-terminated    length       length of string, or -1 if the string is zero-terminated
99    errp         pointer to an error code variable    errp         pointer to an error position offset variable
100    
101  Returns:       < 0    if the string is a valid UTF-8 string  Returns:       = 0    if the string is a valid UTF-8 string
102                 >= 0   otherwise; the value is the offset of the bad character                 > 0    otherwise, setting the offset of the bad character
103  */  */
104    
105  int  int
106  _pcre_valid_utf8(USPTR string, int length, int *errorcode)  _pcre_valid_utf8(USPTR string, int length, int *erroroffset)
107  {  {
108  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF8
109  register USPTR p;  register USPTR p;
# Line 114  if (length < 0) Line 114  if (length < 0)
114    length = p - string;    length = p - string;
115    }    }
116    
 *errorcode = PCRE_UTF8_ERR0;  
   
117  for (p = string; length-- > 0; p++)  for (p = string; length-- > 0; p++)
118    {    {
119    register int ab, c, d;    register int ab, c, d;
# Line 125  for (p = string; length-- > 0; p++) Line 123  for (p = string; length-- > 0; p++)
123    
124    if (c < 0xc0)                         /* Isolated 10xx xxxx byte */    if (c < 0xc0)                         /* Isolated 10xx xxxx byte */
125      {      {
126      *errorcode = PCRE_UTF8_ERR20;      *erroroffset = p - string;
127      return p - string;      return PCRE_UTF8_ERR20;
128      }      }
129    
130    if (c >= 0xfe)                        /* Invalid 0xfe or 0xff bytes */    if (c >= 0xfe)                        /* Invalid 0xfe or 0xff bytes */
131      {      {
132      *errorcode = PCRE_UTF8_ERR21;      *erroroffset = p - string;
133      return p - string;      return PCRE_UTF8_ERR21;
134      }      }
135    
136    ab = _pcre_utf8_table4[c & 0x3f];     /* Number of additional bytes */    ab = _pcre_utf8_table4[c & 0x3f];     /* Number of additional bytes */
137    if (length < ab)    if (length < ab)
138      {      {
139      *errorcode = ab - length;           /* Codes ERR1 to ERR5 */      *erroroffset = p - string;          /* Missing bytes */
140      return p - string;                  /* Missing bytes */      return ab - length;                 /* Codes ERR1 to ERR5 */
141      }      }
142    length -= ab;                         /* Length remaining */    length -= ab;                         /* Length remaining */
143    
# Line 147  for (p = string; length-- > 0; p++) Line 145  for (p = string; length-- > 0; p++)
145    
146    if (((d = *(++p)) & 0xc0) != 0x80)    if (((d = *(++p)) & 0xc0) != 0x80)
147      {      {
148      *errorcode = PCRE_UTF8_ERR6;      *erroroffset = p - string - 1;
149      return p - string - 1;      return PCRE_UTF8_ERR6;
150      }      }
151    
152    /* For each length, check that the remaining bytes start with the 0x80 bit    /* For each length, check that the remaining bytes start with the 0x80 bit
# Line 162  for (p = string; length-- > 0; p++) Line 160  for (p = string; length-- > 0; p++)
160    
161      case 1: if ((c & 0x3e) == 0)      case 1: if ((c & 0x3e) == 0)
162        {        {
163        *errorcode = PCRE_UTF8_ERR15;        *erroroffset = p - string - 1;
164        return p - string - 1;        return PCRE_UTF8_ERR15;
165        }        }
166      break;      break;
167    
# Line 174  for (p = string; length-- > 0; p++) Line 172  for (p = string; length-- > 0; p++)
172      case 2:      case 2:
173      if ((*(++p) & 0xc0) != 0x80)     /* Third byte */      if ((*(++p) & 0xc0) != 0x80)     /* Third byte */
174        {        {
175        *errorcode = PCRE_UTF8_ERR7;        *erroroffset = p - string - 2;
176        return p - string - 2;        return PCRE_UTF8_ERR7;
177        }        }
178      if (c == 0xe0 && (d & 0x20) == 0)      if (c == 0xe0 && (d & 0x20) == 0)
179        {        {
180        *errorcode = PCRE_UTF8_ERR16;        *erroroffset = p - string - 2;
181        return p - string - 2;        return PCRE_UTF8_ERR16;
182        }        }
183      if (c == 0xed && d >= 0xa0)      if (c == 0xed && d >= 0xa0)
184        {        {
185        *errorcode = PCRE_UTF8_ERR14;        *erroroffset = p - string - 2;
186        return p - string - 2;        return PCRE_UTF8_ERR14;
187        }        }
188      break;      break;
189    
# Line 196  for (p = string; length-- > 0; p++) Line 194  for (p = string; length-- > 0; p++)
194      case 3:      case 3:
195      if ((*(++p) & 0xc0) != 0x80)     /* Third byte */      if ((*(++p) & 0xc0) != 0x80)     /* Third byte */
196        {        {
197        *errorcode = PCRE_UTF8_ERR7;        *erroroffset = p - string - 2;
198        return p - string - 2;        return PCRE_UTF8_ERR7;
199        }        }
200      if ((*(++p) & 0xc0) != 0x80)     /* Fourth byte */      if ((*(++p) & 0xc0) != 0x80)     /* Fourth byte */
201        {        {
202        *errorcode = PCRE_UTF8_ERR8;        *erroroffset = p - string - 3;
203        return p - string - 3;        return PCRE_UTF8_ERR8;
204        }        }
205      if (c == 0xf0 && (d & 0x30) == 0)      if (c == 0xf0 && (d & 0x30) == 0)
206        {        {
207        *errorcode = PCRE_UTF8_ERR17;        *erroroffset = p - string - 3;
208        return p - string - 3;        return PCRE_UTF8_ERR17;
209        }        }
210      if (c > 0xf4 || (c == 0xf4 && d > 0x8f))      if (c > 0xf4 || (c == 0xf4 && d > 0x8f))
211        {        {
212        *errorcode = PCRE_UTF8_ERR13;        *erroroffset = p - string - 3;
213        return p - string - 3;        return PCRE_UTF8_ERR13;
214        }        }
215      break;      break;
216    
# Line 227  for (p = string; length-- > 0; p++) Line 225  for (p = string; length-- > 0; p++)
225      case 4:      case 4:
226      if ((*(++p) & 0xc0) != 0x80)     /* Third byte */      if ((*(++p) & 0xc0) != 0x80)     /* Third byte */
227        {        {
228        *errorcode = PCRE_UTF8_ERR7;        *erroroffset = p - string - 2;
229        return p - string - 2;        return PCRE_UTF8_ERR7;
230        }        }
231      if ((*(++p) & 0xc0) != 0x80)     /* Fourth byte */      if ((*(++p) & 0xc0) != 0x80)     /* Fourth byte */
232        {        {
233        *errorcode = PCRE_UTF8_ERR8;        *erroroffset = p - string - 3;
234        return p - string - 3;        return PCRE_UTF8_ERR8;
235        }        }
236      if ((*(++p) & 0xc0) != 0x80)     /* Fifth byte */      if ((*(++p) & 0xc0) != 0x80)     /* Fifth byte */
237        {        {
238        *errorcode = PCRE_UTF8_ERR9;        *erroroffset = p - string - 4;
239        return p - string - 4;        return PCRE_UTF8_ERR9;
240        }        }
241      if (c == 0xf8 && (d & 0x38) == 0)      if (c == 0xf8 && (d & 0x38) == 0)
242        {        {
243        *errorcode = PCRE_UTF8_ERR18;        *erroroffset = p - string - 4;
244        return p - string - 4;        return PCRE_UTF8_ERR18;
245        }        }
246      break;      break;
247    
# Line 253  for (p = string; length-- > 0; p++) Line 251  for (p = string; length-- > 0; p++)
251      case 5:      case 5:
252      if ((*(++p) & 0xc0) != 0x80)     /* Third byte */      if ((*(++p) & 0xc0) != 0x80)     /* Third byte */
253        {        {
254        *errorcode = PCRE_UTF8_ERR7;        *erroroffset = p - string - 2;
255        return p - string - 2;        return PCRE_UTF8_ERR7;
256        }        }
257      if ((*(++p) & 0xc0) != 0x80)     /* Fourth byte */      if ((*(++p) & 0xc0) != 0x80)     /* Fourth byte */
258        {        {
259        *errorcode = PCRE_UTF8_ERR8;        *erroroffset = p - string - 3;
260        return p - string - 3;        return PCRE_UTF8_ERR8;
261        }        }
262      if ((*(++p) & 0xc0) != 0x80)     /* Fifth byte */      if ((*(++p) & 0xc0) != 0x80)     /* Fifth byte */
263        {        {
264        *errorcode = PCRE_UTF8_ERR9;        *erroroffset = p - string - 4;
265        return p - string - 4;        return PCRE_UTF8_ERR9;
266        }        }
267      if ((*(++p) & 0xc0) != 0x80)     /* Sixth byte */      if ((*(++p) & 0xc0) != 0x80)     /* Sixth byte */
268        {        {
269        *errorcode = PCRE_UTF8_ERR10;        *erroroffset = p - string - 5;
270        return p - string - 5;        return PCRE_UTF8_ERR10;
271        }        }
272      if (c == 0xfc && (d & 0x3c) == 0)      if (c == 0xfc && (d & 0x3c) == 0)
273        {        {
274        *errorcode = PCRE_UTF8_ERR19;        *erroroffset = p - string - 5;
275        return p - string - 5;        return PCRE_UTF8_ERR19;
276        }        }
277      break;      break;
278      }      }
# Line 285  for (p = string; length-- > 0; p++) Line 283  for (p = string; length-- > 0; p++)
283    
284    if (ab > 3)    if (ab > 3)
285      {      {
286      *errorcode = (ab == 4)? PCRE_UTF8_ERR11 : PCRE_UTF8_ERR12;      *erroroffset = p - string - ab;
287      return p - string - ab;      return (ab == 4)? PCRE_UTF8_ERR11 : PCRE_UTF8_ERR12;
288      }      }
289    }    }
290    
# Line 295  for (p = string; length-- > 0; p++) Line 293  for (p = string; length-- > 0; p++)
293  (void)(length);  (void)(length);
294  #endif  #endif
295    
296  return -1;   /* This indicates success */  return PCRE_UTF8_ERR0;   /* This indicates success */
297  }  }
298    
299  /* End of pcre_valid_utf8.c */  /* End of pcre_valid_utf8.c */

Legend:
Removed from v.605  
changed lines
  Added in v.606

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12