1 /*
2 * libjingle
3 * Copyright 2004--2005, Google Inc.
4 *
5 * Redistribution and use in source and binary forms, with or without
6 * modification, are permitted provided that the following conditions are met:
7 *
8 * 1. Redistributions of source code must retain the above copyright notice,
9 * this list of conditions and the following disclaimer.
10 * 2. Redistributions in binary form must reproduce the above copyright notice,
11 * this list of conditions and the following disclaimer in the documentation
12 * and/or other materials provided with the distribution.
13 * 3. The name of the author may not be used to endorse or promote products
14 * derived from this software without specific prior written permission.
15 *
16 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR IMPLIED
17 * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
18 * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO
19 * EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
20 * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
21 * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
22 * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
23 * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
24 * OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
25 * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
26 */
27
28 #ifndef TALK_BASE_STRINGUTILS_H__
29 #define TALK_BASE_STRINGUTILS_H__
30
31 #include <ctype.h>
32 #include <stdarg.h>
33 #include <stdio.h>
34
35 #ifdef WIN32
36 #include <malloc.h>
37 #include <wchar.h>
38 #define alloca _alloca
39 #endif // WIN32
40
41 #ifdef POSIX
42 #ifdef BSD
43 #include <stdlib.h>
44 #else // BSD
45 #include <alloca.h>
46 #endif // !BSD
47 #endif // POSIX
48
49 #include <cstring>
50 #include <string>
51
52 #include "talk/base/basictypes.h"
53
54 ///////////////////////////////////////////////////////////////////////////////
55 // Generic string/memory utilities
56 ///////////////////////////////////////////////////////////////////////////////
57
58 #define STACK_ARRAY(TYPE, LEN) static_cast<TYPE*>(::alloca((LEN)*sizeof(TYPE)))
59
60 namespace talk_base {
61
62 // Complement to memset. Verifies memory consists of count bytes of value c.
63 bool memory_check(const void* memory, int c, size_t count);
64
65 // Determines whether the simple wildcard pattern matches target.
66 // Alpha characters in pattern match case-insensitively.
67 // Asterisks in pattern match 0 or more characters.
68 // Ex: string_match("www.TEST.GOOGLE.COM", "www.*.com") -> true
69 bool string_match(const char* target, const char* pattern);
70
71 } // namespace talk_base
72
73 ///////////////////////////////////////////////////////////////////////////////
74 // Rename a bunch of common string functions so they are consistent across
75 // platforms and between char and wchar_t variants.
76 // Here is the full list of functions that are unified:
77 // strlen, strcmp, stricmp, strncmp, strnicmp
78 // strchr, vsnprintf, strtoul, tolowercase
79 // tolowercase is like tolower, but not compatible with end-of-file value
80 //
81 // It's not clear if we will ever use wchar_t strings on unix. In theory,
82 // all strings should be Utf8 all the time, except when interfacing with Win32
83 // APIs that require Utf16.
84 ///////////////////////////////////////////////////////////////////////////////
85
tolowercase(char c)86 inline char tolowercase(char c) {
87 return static_cast<char>(tolower(c));
88 }
89
90 #ifdef WIN32
91
strlen(const wchar_t * s)92 inline size_t strlen(const wchar_t* s) {
93 return wcslen(s);
94 }
strcmp(const wchar_t * s1,const wchar_t * s2)95 inline int strcmp(const wchar_t* s1, const wchar_t* s2) {
96 return wcscmp(s1, s2);
97 }
stricmp(const wchar_t * s1,const wchar_t * s2)98 inline int stricmp(const wchar_t* s1, const wchar_t* s2) {
99 return _wcsicmp(s1, s2);
100 }
strncmp(const wchar_t * s1,const wchar_t * s2,size_t n)101 inline int strncmp(const wchar_t* s1, const wchar_t* s2, size_t n) {
102 return wcsncmp(s1, s2, n);
103 }
strnicmp(const wchar_t * s1,const wchar_t * s2,size_t n)104 inline int strnicmp(const wchar_t* s1, const wchar_t* s2, size_t n) {
105 return _wcsnicmp(s1, s2, n);
106 }
strchr(const wchar_t * s,wchar_t c)107 inline const wchar_t* strchr(const wchar_t* s, wchar_t c) {
108 return wcschr(s, c);
109 }
strstr(const wchar_t * haystack,const wchar_t * needle)110 inline const wchar_t* strstr(const wchar_t* haystack, const wchar_t* needle) {
111 return wcsstr(haystack, needle);
112 }
113 #ifndef vsnprintf
vsnprintf(char * buf,size_t n,const char * fmt,va_list args)114 inline int vsnprintf(char* buf, size_t n, const char* fmt, va_list args) {
115 return _vsnprintf(buf, n, fmt, args);
116 }
vsnprintf(wchar_t * buf,size_t n,const wchar_t * fmt,va_list args)117 inline int vsnprintf(wchar_t* buf, size_t n, const wchar_t* fmt, va_list args) {
118 return _vsnwprintf(buf, n, fmt, args);
119 }
120 #endif // !vsnprintf
strtoul(const wchar_t * snum,wchar_t ** end,int base)121 inline unsigned long strtoul(const wchar_t* snum, wchar_t** end, int base) {
122 return wcstoul(snum, end, base);
123 }
tolowercase(wchar_t c)124 inline wchar_t tolowercase(wchar_t c) {
125 return static_cast<wchar_t>(towlower(c));
126 }
127
128 #endif // WIN32
129
130 #ifdef POSIX
131
_stricmp(const char * s1,const char * s2)132 inline int _stricmp(const char* s1, const char* s2) {
133 return strcasecmp(s1, s2);
134 }
_strnicmp(const char * s1,const char * s2,size_t n)135 inline int _strnicmp(const char* s1, const char* s2, size_t n) {
136 return strncasecmp(s1, s2, n);
137 }
138
139 #endif // POSIX
140
141 ///////////////////////////////////////////////////////////////////////////////
142 // Traits simplifies porting string functions to be CTYPE-agnostic
143 ///////////////////////////////////////////////////////////////////////////////
144
145 namespace talk_base {
146
147 const size_t SIZE_UNKNOWN = static_cast<size_t>(-1);
148
149 template<class CTYPE>
150 struct Traits {
151 // STL string type
152 //typedef XXX string;
153 // Null-terminated string
154 //inline static const CTYPE* empty_str();
155 };
156
157 ///////////////////////////////////////////////////////////////////////////////
158 // String utilities which work with char or wchar_t
159 ///////////////////////////////////////////////////////////////////////////////
160
161 template<class CTYPE>
162 inline const CTYPE* nonnull(const CTYPE* str, const CTYPE* def_str = NULL) {
163 return str ? str : (def_str ? def_str : Traits<CTYPE>::empty_str());
164 }
165
166 template<class CTYPE>
strchr(const CTYPE * str,const CTYPE * chs)167 const CTYPE* strchr(const CTYPE* str, const CTYPE* chs) {
168 for (size_t i=0; str[i]; ++i) {
169 for (size_t j=0; chs[j]; ++j) {
170 if (str[i] == chs[j]) {
171 return str + i;
172 }
173 }
174 }
175 return 0;
176 }
177
178 template<class CTYPE>
strchrn(const CTYPE * str,size_t slen,CTYPE ch)179 const CTYPE* strchrn(const CTYPE* str, size_t slen, CTYPE ch) {
180 for (size_t i=0; i<slen && str[i]; ++i) {
181 if (str[i] == ch) {
182 return str + i;
183 }
184 }
185 return 0;
186 }
187
188 template<class CTYPE>
strlenn(const CTYPE * buffer,size_t buflen)189 size_t strlenn(const CTYPE* buffer, size_t buflen) {
190 size_t bufpos = 0;
191 while (buffer[bufpos] && (bufpos < buflen)) {
192 ++bufpos;
193 }
194 return bufpos;
195 }
196
197 // Safe versions of strncpy, strncat, snprintf and vsnprintf that always
198 // null-terminate.
199
200 template<class CTYPE>
201 size_t strcpyn(CTYPE* buffer, size_t buflen,
202 const CTYPE* source, size_t srclen = SIZE_UNKNOWN) {
203 if (buflen <= 0)
204 return 0;
205
206 if (srclen == SIZE_UNKNOWN) {
207 srclen = strlenn(source, buflen - 1);
208 } else if (srclen >= buflen) {
209 srclen = buflen - 1;
210 }
211 memcpy(buffer, source, srclen * sizeof(CTYPE));
212 buffer[srclen] = 0;
213 return srclen;
214 }
215
216 template<class CTYPE>
217 size_t strcatn(CTYPE* buffer, size_t buflen,
218 const CTYPE* source, size_t srclen = SIZE_UNKNOWN) {
219 if (buflen <= 0)
220 return 0;
221
222 size_t bufpos = strlenn(buffer, buflen - 1);
223 return bufpos + strcpyn(buffer + bufpos, buflen - bufpos, source, srclen);
224 }
225
226 // Some compilers (clang specifically) require vsprintfn be defined before
227 // sprintfn.
228 template<class CTYPE>
vsprintfn(CTYPE * buffer,size_t buflen,const CTYPE * format,va_list args)229 size_t vsprintfn(CTYPE* buffer, size_t buflen, const CTYPE* format,
230 va_list args) {
231 int len = vsnprintf(buffer, buflen, format, args);
232 if ((len < 0) || (static_cast<size_t>(len) >= buflen)) {
233 len = static_cast<int>(buflen - 1);
234 buffer[len] = 0;
235 }
236 return len;
237 }
238
239 template<class CTYPE>
240 size_t sprintfn(CTYPE* buffer, size_t buflen, const CTYPE* format, ...);
241 /* This works to get GCC to notice printf argument mismatches, but then complains of missing implementation of sprintfn<char>
242 template<>
243 size_t sprintfn(char* buffer, size_t buflen, const char* format, ...)
244 GCC_ATTR(format(printf,3,4));
245 */
246 template<class CTYPE>
sprintfn(CTYPE * buffer,size_t buflen,const CTYPE * format,...)247 size_t sprintfn(CTYPE* buffer, size_t buflen, const CTYPE* format, ...) {
248 va_list args;
249 va_start(args, format);
250 size_t len = vsprintfn(buffer, buflen, format, args);
251 va_end(args);
252 return len;
253 }
254
255 ///////////////////////////////////////////////////////////////////////////////
256 // Allow safe comparing and copying ascii (not UTF-8) with both wide and
257 // non-wide character strings.
258 ///////////////////////////////////////////////////////////////////////////////
259
asccmp(const char * s1,const char * s2)260 inline int asccmp(const char* s1, const char* s2) {
261 return strcmp(s1, s2);
262 }
ascicmp(const char * s1,const char * s2)263 inline int ascicmp(const char* s1, const char* s2) {
264 return _stricmp(s1, s2);
265 }
ascncmp(const char * s1,const char * s2,size_t n)266 inline int ascncmp(const char* s1, const char* s2, size_t n) {
267 return strncmp(s1, s2, n);
268 }
ascnicmp(const char * s1,const char * s2,size_t n)269 inline int ascnicmp(const char* s1, const char* s2, size_t n) {
270 return _strnicmp(s1, s2, n);
271 }
272 inline size_t asccpyn(char* buffer, size_t buflen,
273 const char* source, size_t srclen = SIZE_UNKNOWN) {
274 return strcpyn(buffer, buflen, source, srclen);
275 }
276
277 #ifdef WIN32
278
279 typedef wchar_t(*CharacterTransformation)(wchar_t);
identity(wchar_t c)280 inline wchar_t identity(wchar_t c) { return c; }
281 int ascii_string_compare(const wchar_t* s1, const char* s2, size_t n,
282 CharacterTransformation transformation);
283
asccmp(const wchar_t * s1,const char * s2)284 inline int asccmp(const wchar_t* s1, const char* s2) {
285 return ascii_string_compare(s1, s2, static_cast<size_t>(-1), identity);
286 }
ascicmp(const wchar_t * s1,const char * s2)287 inline int ascicmp(const wchar_t* s1, const char* s2) {
288 return ascii_string_compare(s1, s2, static_cast<size_t>(-1), tolowercase);
289 }
ascncmp(const wchar_t * s1,const char * s2,size_t n)290 inline int ascncmp(const wchar_t* s1, const char* s2, size_t n) {
291 return ascii_string_compare(s1, s2, n, identity);
292 }
ascnicmp(const wchar_t * s1,const char * s2,size_t n)293 inline int ascnicmp(const wchar_t* s1, const char* s2, size_t n) {
294 return ascii_string_compare(s1, s2, n, tolowercase);
295 }
296 size_t asccpyn(wchar_t* buffer, size_t buflen,
297 const char* source, size_t srclen = SIZE_UNKNOWN);
298
299 #endif // WIN32
300
301 ///////////////////////////////////////////////////////////////////////////////
302 // Traits<char> specializations
303 ///////////////////////////////////////////////////////////////////////////////
304
305 template<>
306 struct Traits<char> {
307 typedef std::string string;
308 inline static const char* empty_str() { return ""; }
309 };
310
311 ///////////////////////////////////////////////////////////////////////////////
312 // Traits<wchar_t> specializations (Windows only, currently)
313 ///////////////////////////////////////////////////////////////////////////////
314
315 #ifdef WIN32
316
317 template<>
318 struct Traits<wchar_t> {
319 typedef std::wstring string;
320 inline static const wchar_t* Traits<wchar_t>::empty_str() { return L""; }
321 };
322
323 #endif // WIN32
324
325 // Replaces all occurrences of "search" with "replace".
326 void replace_substrs(const char *search,
327 size_t search_len,
328 const char *replace,
329 size_t replace_len,
330 std::string *s);
331
332 // True iff s1 starts with s2.
333 bool starts_with(const char *s1, const char *s2);
334
335 // Remove leading and trailing whitespaces.
336 std::string string_trim(const std::string& s);
337
338 } // namespace talk_base
339
340 #endif // TALK_BASE_STRINGUTILS_H__
341