xref: /PHP-5.3/ext/mbstring/oniguruma/regext.c (revision 6284ef11)
1 /**********************************************************************
2   regext.c -  Oniguruma (regular expression library)
3 **********************************************************************/
4 /*-
5  * Copyright (c) 2002-2006  K.Kosako  <sndgk393 AT ybb DOT ne DOT jp>
6  * All rights reserved.
7  *
8  * Redistribution and use in source and binary forms, with or without
9  * modification, are permitted provided that the following conditions
10  * are met:
11  * 1. Redistributions of source code must retain the above copyright
12  *    notice, this list of conditions and the following disclaimer.
13  * 2. Redistributions in binary form must reproduce the above copyright
14  *    notice, this list of conditions and the following disclaimer in the
15  *    documentation and/or other materials provided with the distribution.
16  *
17  * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
18  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
19  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
20  * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
21  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
22  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
23  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
24  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
25  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
26  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
27  * SUCH DAMAGE.
28  */
29 
30 #include "regint.h"
31 
32 static void
conv_ext0be32(const UChar * s,const UChar * end,UChar * conv)33 conv_ext0be32(const UChar* s, const UChar* end, UChar* conv)
34 {
35   while (s < end) {
36     *conv++ = '\0';
37     *conv++ = '\0';
38     *conv++ = '\0';
39     *conv++ = *s++;
40   }
41 }
42 
43 static void
conv_ext0le32(const UChar * s,const UChar * end,UChar * conv)44 conv_ext0le32(const UChar* s, const UChar* end, UChar* conv)
45 {
46   while (s < end) {
47     *conv++ = *s++;
48     *conv++ = '\0';
49     *conv++ = '\0';
50     *conv++ = '\0';
51   }
52 }
53 
54 static void
conv_ext0be(const UChar * s,const UChar * end,UChar * conv)55 conv_ext0be(const UChar* s, const UChar* end, UChar* conv)
56 {
57   while (s < end) {
58     *conv++ = '\0';
59     *conv++ = *s++;
60   }
61 }
62 
63 static void
conv_ext0le(const UChar * s,const UChar * end,UChar * conv)64 conv_ext0le(const UChar* s, const UChar* end, UChar* conv)
65 {
66   while (s < end) {
67     *conv++ = *s++;
68     *conv++ = '\0';
69   }
70 }
71 
72 static void
conv_swap4bytes(const UChar * s,const UChar * end,UChar * conv)73 conv_swap4bytes(const UChar* s, const UChar* end, UChar* conv)
74 {
75   while (s < end) {
76     *conv++ = s[3];
77     *conv++ = s[2];
78     *conv++ = s[1];
79     *conv++ = s[0];
80     s += 4;
81   }
82 }
83 
84 static void
conv_swap2bytes(const UChar * s,const UChar * end,UChar * conv)85 conv_swap2bytes(const UChar* s, const UChar* end, UChar* conv)
86 {
87   while (s < end) {
88     *conv++ = s[1];
89     *conv++ = s[0];
90     s += 2;
91   }
92 }
93 
94 static int
conv_encoding(OnigEncoding from,OnigEncoding to,const UChar * s,const UChar * end,UChar ** conv,UChar ** conv_end)95 conv_encoding(OnigEncoding from, OnigEncoding to, const UChar* s, const UChar* end,
96               UChar** conv, UChar** conv_end)
97 {
98   int len = end - s;
99 
100   if (to == ONIG_ENCODING_UTF16_BE) {
101     if (from == ONIG_ENCODING_ASCII || from == ONIG_ENCODING_ISO_8859_1) {
102       *conv = (UChar* )xmalloc(len * 2);
103       CHECK_NULL_RETURN_VAL(*conv, ONIGERR_MEMORY);
104       *conv_end = *conv + (len * 2);
105       conv_ext0be(s, end, *conv);
106       return 0;
107     }
108     else if (from == ONIG_ENCODING_UTF16_LE) {
109     swap16:
110       *conv = (UChar* )xmalloc(len);
111       CHECK_NULL_RETURN_VAL(*conv, ONIGERR_MEMORY);
112       *conv_end = *conv + len;
113       conv_swap2bytes(s, end, *conv);
114       return 0;
115     }
116   }
117   else if (to == ONIG_ENCODING_UTF16_LE) {
118     if (from == ONIG_ENCODING_ASCII || from == ONIG_ENCODING_ISO_8859_1) {
119       *conv = (UChar* )xmalloc(len * 2);
120       CHECK_NULL_RETURN_VAL(*conv, ONIGERR_MEMORY);
121       *conv_end = *conv + (len * 2);
122       conv_ext0le(s, end, *conv);
123       return 0;
124     }
125     else if (from == ONIG_ENCODING_UTF16_BE) {
126       goto swap16;
127     }
128   }
129   if (to == ONIG_ENCODING_UTF32_BE) {
130     if (from == ONIG_ENCODING_ASCII || from == ONIG_ENCODING_ISO_8859_1) {
131       *conv = (UChar* )xmalloc(len * 4);
132       CHECK_NULL_RETURN_VAL(*conv, ONIGERR_MEMORY);
133       *conv_end = *conv + (len * 4);
134       conv_ext0be32(s, end, *conv);
135       return 0;
136     }
137     else if (from == ONIG_ENCODING_UTF32_LE) {
138     swap32:
139       *conv = (UChar* )xmalloc(len);
140       CHECK_NULL_RETURN_VAL(*conv, ONIGERR_MEMORY);
141       *conv_end = *conv + len;
142       conv_swap4bytes(s, end, *conv);
143       return 0;
144     }
145   }
146   else if (to == ONIG_ENCODING_UTF32_LE) {
147     if (from == ONIG_ENCODING_ASCII || from == ONIG_ENCODING_ISO_8859_1) {
148       *conv = (UChar* )xmalloc(len * 4);
149       CHECK_NULL_RETURN_VAL(*conv, ONIGERR_MEMORY);
150       *conv_end = *conv + (len * 4);
151       conv_ext0le32(s, end, *conv);
152       return 0;
153     }
154     else if (from == ONIG_ENCODING_UTF32_BE) {
155       goto swap32;
156     }
157   }
158 
159   return ONIGERR_NOT_SUPPORTED_ENCODING_COMBINATION;
160 }
161 
162 extern int
onig_new_deluxe(regex_t ** reg,const UChar * pattern,const UChar * pattern_end,OnigCompileInfo * ci,OnigErrorInfo * einfo)163 onig_new_deluxe(regex_t** reg, const UChar* pattern, const UChar* pattern_end,
164                 OnigCompileInfo* ci, OnigErrorInfo* einfo)
165 {
166   int r;
167   UChar *cpat, *cpat_end;
168 
169   if (IS_NOT_NULL(einfo)) einfo->par = (UChar* )NULL;
170 
171   if (ci->pattern_enc != ci->target_enc) {
172     r = conv_encoding(ci->pattern_enc, ci->target_enc, pattern, pattern_end,
173                       &cpat, &cpat_end);
174     if (r) return r;
175   }
176   else {
177     cpat     = (UChar* )pattern;
178     cpat_end = (UChar* )pattern_end;
179   }
180 
181   r = onig_alloc_init(reg, ci->option, ci->ambig_flag, ci->target_enc,
182                       ci->syntax);
183   if (r) goto err;
184 
185   r = onig_compile(*reg, cpat, cpat_end, einfo);
186   if (r) {
187     onig_free(*reg);
188     *reg = NULL;
189   }
190 
191  err:
192   if (cpat != pattern) xfree(cpat);
193 
194   return r;
195 }
196 
197 #ifdef USE_RECOMPILE_API
198 extern int
onig_recompile_deluxe(regex_t * reg,const UChar * pattern,const UChar * pattern_end,OnigCompileInfo * ci,OnigErrorInfo * einfo)199 onig_recompile_deluxe(regex_t* reg, const UChar* pattern, const UChar* pattern_end,
200                       OnigCompileInfo* ci, OnigErrorInfo* einfo)
201 {
202   int r;
203   regex_t *new_reg;
204 
205   r = onig_new_deluxe(&new_reg, pattern, pattern_end, ci, einfo);
206   if (r) return r;
207   if (ONIG_STATE(reg) == ONIG_STATE_NORMAL) {
208     onig_transfer(reg, new_reg);
209   }
210   else {
211     onig_chain_link_add(reg, new_reg);
212   }
213   return 0;
214 }
215 #endif
216