/[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 293 by ph10, Sat Dec 29 16:50:30 2007 UTC revision 835 by ph10, Wed Dec 28 16:10:09 2011 UTC
# Line 6  Line 6 
6  and semantics are as close as possible to those of the Perl 5 language.  and semantics are as close as possible to those of the Perl 5 language.
7    
8                         Written by Philip Hazel                         Written by Philip Hazel
9             Copyright (c) 1997-2007 University of Cambridge             Copyright (c) 1997-2009 University of Cambridge
10    
11  -----------------------------------------------------------------------------  -----------------------------------------------------------------------------
12  Redistribution and use in source and binary forms, with or without  Redistribution and use in source and binary forms, with or without
# Line 54  strings. */ Line 54  strings. */
54  *************************************************/  *************************************************/
55    
56  /* This function is called (optionally) at the start of compile or match, to  /* This function is called (optionally) at the start of compile or match, to
57  validate that a supposed UTF-8 string is actually valid. The early check means  check that a supposed UTF-8 string is actually valid. The early check means
58  that subsequent code can assume it is dealing with a valid string. The check  that subsequent code can assume it is dealing with a valid string. The check
59  can be turned off for maximum performance, but the consequences of supplying  can be turned off for maximum performance, but the consequences of supplying an
60  an invalid string are then undefined.  invalid string are then undefined.
61    
62  Originally, this function checked according to RFC 2279, allowing for values in  Originally, this function checked according to RFC 2279, allowing for values in
63  the range 0 to 0x7fffffff, up to 6 bytes long, but ensuring that they were in  the range 0 to 0x7fffffff, up to 6 bytes long, but ensuring that they were in
64  the canonical format. Once somebody had pointed out RFC 3629 to me (it  the canonical format. Once somebody had pointed out RFC 3629 to me (it
65  obsoletes 2279), additional restrictions were applied. The values are now  obsoletes 2279), additional restrictions were applied. The values are now
66  limited to be between 0 and 0x0010ffff, no more than 4 bytes long, and the  limited to be between 0 and 0x0010ffff, no more than 4 bytes long, and the
67  subrange 0xd000 to 0xdfff is excluded.  subrange 0xd000 to 0xdfff is excluded. However, the format of 5-byte and 6-byte
68    characters is still checked.
69    
70    From release 8.13 more information about the details of the error are passed
71    back in the returned value:
72    
73    PCRE_UTF8_ERR0   No error
74    PCRE_UTF8_ERR1   Missing 1 byte at the end of the string
75    PCRE_UTF8_ERR2   Missing 2 bytes at the end of the string
76    PCRE_UTF8_ERR3   Missing 3 bytes at the end of the string
77    PCRE_UTF8_ERR4   Missing 4 bytes at the end of the string
78    PCRE_UTF8_ERR5   Missing 5 bytes at the end of the string
79    PCRE_UTF8_ERR6   2nd-byte's two top bits are not 0x80
80    PCRE_UTF8_ERR7   3rd-byte's two top bits are not 0x80
81    PCRE_UTF8_ERR8   4th-byte's two top bits are not 0x80
82    PCRE_UTF8_ERR9   5th-byte's two top bits are not 0x80
83    PCRE_UTF8_ERR10  6th-byte's two top bits are not 0x80
84    PCRE_UTF8_ERR11  5-byte character is not permitted by RFC 3629
85    PCRE_UTF8_ERR12  6-byte character is not permitted by RFC 3629
86    PCRE_UTF8_ERR13  4-byte character with value > 0x10ffff is not permitted
87    PCRE_UTF8_ERR14  3-byte character with value 0xd000-0xdfff is not permitted
88    PCRE_UTF8_ERR15  Overlong 2-byte sequence
89    PCRE_UTF8_ERR16  Overlong 3-byte sequence
90    PCRE_UTF8_ERR17  Overlong 4-byte sequence
91    PCRE_UTF8_ERR18  Overlong 5-byte sequence (won't ever occur)
92    PCRE_UTF8_ERR19  Overlong 6-byte sequence (won't ever occur)
93    PCRE_UTF8_ERR20  Isolated 0x80 byte (not within UTF-8 character)
94    PCRE_UTF8_ERR21  Byte with the illegal value 0xfe or 0xff
95    
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 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 byte                 > 0    otherwise, setting the offset of the bad character
103  */  */
104    
105  int  int
106  _pcre_valid_utf8(const uschar *string, int length)  _pcre_valid_utf8(USPTR string, int length, int *erroroffset)
107  {  {
108  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF8
109  register const uschar *p;  register USPTR p;
110    
111  if (length < 0)  if (length < 0)
112    {    {
# Line 88  if (length < 0) Line 116  if (length < 0)
116    
117  for (p = string; length-- > 0; p++)  for (p = string; length-- > 0; p++)
118    {    {
119    register int ab;    register int ab, c, d;
120    register int c = *p;  
121    if (c < 128) continue;    c = *p;
122    if (c < 0xc0) return p - string;    if (c < 128) continue;                /* ASCII character */
123    
124      if (c < 0xc0)                         /* Isolated 10xx xxxx byte */
125        {
126        *erroroffset = p - string;
127        return PCRE_UTF8_ERR20;
128        }
129    
130      if (c >= 0xfe)                        /* Invalid 0xfe or 0xff bytes */
131        {
132        *erroroffset = p - string;
133        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 || ab > 3) return p - string;    if (length < ab)
138    length -= ab;      {
139        *erroroffset = p - string;          /* Missing bytes */
140        return ab - length;                 /* Codes ERR1 to ERR5 */
141        }
142      length -= ab;                         /* Length remaining */
143    
144    /* Check top bits in the second byte */    /* Check top bits in the second byte */
   if ((*(++p) & 0xc0) != 0x80) return p - string;  
145    
146    /* Check for overlong sequences for each different length, and for the    if (((d = *(++p)) & 0xc0) != 0x80)
147    excluded range 0xd000 to 0xdfff.  */      {
148        *erroroffset = p - string - 1;
149        return PCRE_UTF8_ERR6;
150        }
151    
152      /* For each length, check that the remaining bytes start with the 0x80 bit
153      set and not the 0x40 bit. Then check for an overlong sequence, and for the
154      excluded range 0xd800 to 0xdfff. */
155    
156    switch (ab)    switch (ab)
157      {      {
158      /* Check for xx00 000x (overlong sequence) */      /* 2-byte character. No further bytes to check for 0x80. Check first byte
159        for for xx00 000x (overlong sequence). */
160    
161      case 1:      case 1: if ((c & 0x3e) == 0)
162      if ((c & 0x3e) == 0) return p - string;        {
163      continue;   /* We know there aren't any more bytes to check */        *erroroffset = p - string - 1;
164          return PCRE_UTF8_ERR15;
165          }
166        break;
167    
168      /* Check for 1110 0000, xx0x xxxx (overlong sequence) or      /* 3-byte character. Check third byte for 0x80. Then check first 2 bytes
169                   1110 1101, 1010 xxxx (0xd000 - 0xdfff) */        for 1110 0000, xx0x xxxx (overlong sequence) or
170              1110 1101, 1010 xxxx (0xd800 - 0xdfff) */
171    
172      case 2:      case 2:
173      if ((c == 0xe0 && (*p & 0x20) == 0) ||      if ((*(++p) & 0xc0) != 0x80)     /* Third byte */
174          (c == 0xed && *p >= 0xa0))        {
175        return p - string;        *erroroffset = p - string - 2;
176          return PCRE_UTF8_ERR7;
177          }
178        if (c == 0xe0 && (d & 0x20) == 0)
179          {
180          *erroroffset = p - string - 2;
181          return PCRE_UTF8_ERR16;
182          }
183        if (c == 0xed && d >= 0xa0)
184          {
185          *erroroffset = p - string - 2;
186          return PCRE_UTF8_ERR14;
187          }
188      break;      break;
189    
190      /* Check for 1111 0000, xx00 xxxx (overlong sequence) or      /* 4-byte character. Check 3rd and 4th bytes for 0x80. Then check first 2
191         greater than 0x0010ffff (f4 8f bf bf) */         bytes for for 1111 0000, xx00 xxxx (overlong sequence), then check for a
192           character greater than 0x0010ffff (f4 8f bf bf) */
193    
194      case 3:      case 3:
195      if ((c == 0xf0 && (*p & 0x30) == 0) ||      if ((*(++p) & 0xc0) != 0x80)     /* Third byte */
196          (c > 0xf4 ) ||        {
197          (c == 0xf4 && *p > 0x8f))        *erroroffset = p - string - 2;
198        return p - string;        return PCRE_UTF8_ERR7;
199          }
200        if ((*(++p) & 0xc0) != 0x80)     /* Fourth byte */
201          {
202          *erroroffset = p - string - 3;
203          return PCRE_UTF8_ERR8;
204          }
205        if (c == 0xf0 && (d & 0x30) == 0)
206          {
207          *erroroffset = p - string - 3;
208          return PCRE_UTF8_ERR17;
209          }
210        if (c > 0xf4 || (c == 0xf4 && d > 0x8f))
211          {
212          *erroroffset = p - string - 3;
213          return PCRE_UTF8_ERR13;
214          }
215      break;      break;
216    
217  #if 0      /* 5-byte and 6-byte characters are not allowed by RFC 3629, and will be
218      /* These cases can no longer occur, as we restrict to a maximum of four      rejected by the length test below. However, we do the appropriate tests
219      bytes nowadays. Leave the code here in case we ever want to add an option      here so that overlong sequences get diagnosed, and also in case there is
220      for longer sequences. */      ever an option for handling these larger code points. */
221    
222        /* 5-byte character. Check 3rd, 4th, and 5th bytes for 0x80. Then check for
223        1111 1000, xx00 0xxx */
224    
     /* Check for 1111 1000, xx00 0xxx */  
225      case 4:      case 4:
226      if (c == 0xf8 && (*p & 0x38) == 0) return p - string;      if ((*(++p) & 0xc0) != 0x80)     /* Third byte */
227          {
228          *erroroffset = p - string - 2;
229          return PCRE_UTF8_ERR7;
230          }
231        if ((*(++p) & 0xc0) != 0x80)     /* Fourth byte */
232          {
233          *erroroffset = p - string - 3;
234          return PCRE_UTF8_ERR8;
235          }
236        if ((*(++p) & 0xc0) != 0x80)     /* Fifth byte */
237          {
238          *erroroffset = p - string - 4;
239          return PCRE_UTF8_ERR9;
240          }
241        if (c == 0xf8 && (d & 0x38) == 0)
242          {
243          *erroroffset = p - string - 4;
244          return PCRE_UTF8_ERR18;
245          }
246      break;      break;
247    
248      /* Check for leading 0xfe or 0xff, and then for 1111 1100, xx00 00xx */      /* 6-byte character. Check 3rd-6th bytes for 0x80. Then check for
249        1111 1100, xx00 00xx. */
250    
251      case 5:      case 5:
252      if (c == 0xfe || c == 0xff ||      if ((*(++p) & 0xc0) != 0x80)     /* Third byte */
253         (c == 0xfc && (*p & 0x3c) == 0)) return p - string;        {
254          *erroroffset = p - string - 2;
255          return PCRE_UTF8_ERR7;
256          }
257        if ((*(++p) & 0xc0) != 0x80)     /* Fourth byte */
258          {
259          *erroroffset = p - string - 3;
260          return PCRE_UTF8_ERR8;
261          }
262        if ((*(++p) & 0xc0) != 0x80)     /* Fifth byte */
263          {
264          *erroroffset = p - string - 4;
265          return PCRE_UTF8_ERR9;
266          }
267        if ((*(++p) & 0xc0) != 0x80)     /* Sixth byte */
268          {
269          *erroroffset = p - string - 5;
270          return PCRE_UTF8_ERR10;
271          }
272        if (c == 0xfc && (d & 0x3c) == 0)
273          {
274          *erroroffset = p - string - 5;
275          return PCRE_UTF8_ERR19;
276          }
277      break;      break;
 #endif  
   
278      }      }
279    
280    /* Check for valid bytes after the 2nd, if any; all must start 10 */    /* Character is valid under RFC 2279, but 4-byte and 5-byte characters are
281    while (--ab > 0)    excluded by RFC 3629. The pointer p is currently at the last byte of the
282      character. */
283    
284      if (ab > 3)
285      {      {
286      if ((*(++p) & 0xc0) != 0x80) return p - string;      *erroroffset = p - string - ab;
287        return (ab == 4)? PCRE_UTF8_ERR11 : PCRE_UTF8_ERR12;
288      }      }
289    }    }
290    
291    #else  /* SUPPORT_UTF8 */
292    (void)(string);  /* Keep picky compilers happy */
293    (void)(length);
294  #endif  #endif
295    
296  return -1;  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.293  
changed lines
  Added in v.835

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12