1 // Protocol Buffers - Google's data interchange format
2 // Copyright 2008 Google Inc. All rights reserved.
3 // http://code.google.com/p/protobuf/
4 //
5 // Redistribution and use in source and binary forms, with or without
6 // modification, are permitted provided that the following conditions are
7 // met:
8 //
9 // * Redistributions of source code must retain the above copyright
10 // notice, this list of conditions and the following disclaimer.
11 // * Redistributions in binary form must reproduce the above
12 // copyright notice, this list of conditions and the following disclaimer
13 // in the documentation and/or other materials provided with the
14 // distribution.
15 // * Neither the name of Google Inc. nor the names of its
16 // contributors may be used to endorse or promote products derived from
17 // this software without specific prior written permission.
18 //
19 // THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
20 // "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
21 // LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
22 // A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
23 // OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
24 // SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
25 // LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
26 // DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
27 // THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
28 // (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
29 // OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
30
31 // from google3/strings/strutil.h
32
33 #ifndef GOOGLE_PROTOBUF_STUBS_STRUTIL_H__
34 #define GOOGLE_PROTOBUF_STUBS_STRUTIL_H__
35
36 #include <stdlib.h>
37 #include <vector>
38 #include <google/protobuf/stubs/common.h>
39
40 namespace google {
41 namespace protobuf {
42
43 #ifdef _MSC_VER
44 #define strtoll _strtoi64
45 #define strtoull _strtoui64
46 #elif defined(__DECCXX) && defined(__osf__)
47 // HP C++ on Tru64 does not have strtoll, but strtol is already 64-bit.
48 #define strtoll strtol
49 #define strtoull strtoul
50 #endif
51
52 // ----------------------------------------------------------------------
53 // ascii_isalnum()
54 // Check if an ASCII character is alphanumeric. We can't use ctype's
55 // isalnum() because it is affected by locale. This function is applied
56 // to identifiers in the protocol buffer language, not to natural-language
57 // strings, so locale should not be taken into account.
58 // ascii_isdigit()
59 // Like above, but only accepts digits.
60 // ----------------------------------------------------------------------
61
ascii_isalnum(char c)62 inline bool ascii_isalnum(char c) {
63 return ('a' <= c && c <= 'z') ||
64 ('A' <= c && c <= 'Z') ||
65 ('0' <= c && c <= '9');
66 }
67
ascii_isdigit(char c)68 inline bool ascii_isdigit(char c) {
69 return ('0' <= c && c <= '9');
70 }
71
72 // ----------------------------------------------------------------------
73 // HasPrefixString()
74 // Check if a string begins with a given prefix.
75 // StripPrefixString()
76 // Given a string and a putative prefix, returns the string minus the
77 // prefix string if the prefix matches, otherwise the original
78 // string.
79 // ----------------------------------------------------------------------
HasPrefixString(const string & str,const string & prefix)80 inline bool HasPrefixString(const string& str,
81 const string& prefix) {
82 return str.size() >= prefix.size() &&
83 str.compare(0, prefix.size(), prefix) == 0;
84 }
85
StripPrefixString(const string & str,const string & prefix)86 inline string StripPrefixString(const string& str, const string& prefix) {
87 if (HasPrefixString(str, prefix)) {
88 return str.substr(prefix.size());
89 } else {
90 return str;
91 }
92 }
93
94 // ----------------------------------------------------------------------
95 // HasSuffixString()
96 // Return true if str ends in suffix.
97 // StripSuffixString()
98 // Given a string and a putative suffix, returns the string minus the
99 // suffix string if the suffix matches, otherwise the original
100 // string.
101 // ----------------------------------------------------------------------
HasSuffixString(const string & str,const string & suffix)102 inline bool HasSuffixString(const string& str,
103 const string& suffix) {
104 return str.size() >= suffix.size() &&
105 str.compare(str.size() - suffix.size(), suffix.size(), suffix) == 0;
106 }
107
StripSuffixString(const string & str,const string & suffix)108 inline string StripSuffixString(const string& str, const string& suffix) {
109 if (HasSuffixString(str, suffix)) {
110 return str.substr(0, str.size() - suffix.size());
111 } else {
112 return str;
113 }
114 }
115
116 // ----------------------------------------------------------------------
117 // StripString
118 // Replaces any occurrence of the character 'remove' (or the characters
119 // in 'remove') with the character 'replacewith'.
120 // Good for keeping html characters or protocol characters (\t) out
121 // of places where they might cause a problem.
122 // ----------------------------------------------------------------------
123 LIBPROTOBUF_EXPORT void StripString(string* s, const char* remove,
124 char replacewith);
125
126 // ----------------------------------------------------------------------
127 // LowerString()
128 // UpperString()
129 // Convert the characters in "s" to lowercase or uppercase. ASCII-only:
130 // these functions intentionally ignore locale because they are applied to
131 // identifiers used in the Protocol Buffer language, not to natural-language
132 // strings.
133 // ----------------------------------------------------------------------
134
LowerString(string * s)135 inline void LowerString(string * s) {
136 string::iterator end = s->end();
137 for (string::iterator i = s->begin(); i != end; ++i) {
138 // tolower() changes based on locale. We don't want this!
139 if ('A' <= *i && *i <= 'Z') *i += 'a' - 'A';
140 }
141 }
142
UpperString(string * s)143 inline void UpperString(string * s) {
144 string::iterator end = s->end();
145 for (string::iterator i = s->begin(); i != end; ++i) {
146 // toupper() changes based on locale. We don't want this!
147 if ('a' <= *i && *i <= 'z') *i += 'A' - 'a';
148 }
149 }
150
151 // ----------------------------------------------------------------------
152 // StringReplace()
153 // Give me a string and two patterns "old" and "new", and I replace
154 // the first instance of "old" in the string with "new", if it
155 // exists. RETURN a new string, regardless of whether the replacement
156 // happened or not.
157 // ----------------------------------------------------------------------
158
159 LIBPROTOBUF_EXPORT string StringReplace(const string& s, const string& oldsub,
160 const string& newsub, bool replace_all);
161
162 // ----------------------------------------------------------------------
163 // SplitStringUsing()
164 // Split a string using a character delimiter. Append the components
165 // to 'result'. If there are consecutive delimiters, this function skips
166 // over all of them.
167 // ----------------------------------------------------------------------
168 LIBPROTOBUF_EXPORT void SplitStringUsing(const string& full, const char* delim,
169 vector<string>* res);
170
171 // ----------------------------------------------------------------------
172 // JoinStrings()
173 // These methods concatenate a vector of strings into a C++ string, using
174 // the C-string "delim" as a separator between components. There are two
175 // flavors of the function, one flavor returns the concatenated string,
176 // another takes a pointer to the target string. In the latter case the
177 // target string is cleared and overwritten.
178 // ----------------------------------------------------------------------
179 LIBPROTOBUF_EXPORT void JoinStrings(const vector<string>& components,
180 const char* delim, string* result);
181
JoinStrings(const vector<string> & components,const char * delim)182 inline string JoinStrings(const vector<string>& components,
183 const char* delim) {
184 string result;
185 JoinStrings(components, delim, &result);
186 return result;
187 }
188
189 // ----------------------------------------------------------------------
190 // UnescapeCEscapeSequences()
191 // Copies "source" to "dest", rewriting C-style escape sequences
192 // -- '\n', '\r', '\\', '\ooo', etc -- to their ASCII
193 // equivalents. "dest" must be sufficiently large to hold all
194 // the characters in the rewritten string (i.e. at least as large
195 // as strlen(source) + 1 should be safe, since the replacements
196 // are always shorter than the original escaped sequences). It's
197 // safe for source and dest to be the same. RETURNS the length
198 // of dest.
199 //
200 // It allows hex sequences \xhh, or generally \xhhhhh with an
201 // arbitrary number of hex digits, but all of them together must
202 // specify a value of a single byte (e.g. \x0045 is equivalent
203 // to \x45, and \x1234 is erroneous).
204 //
205 // It also allows escape sequences of the form \uhhhh (exactly four
206 // hex digits, upper or lower case) or \Uhhhhhhhh (exactly eight
207 // hex digits, upper or lower case) to specify a Unicode code
208 // point. The dest array will contain the UTF8-encoded version of
209 // that code-point (e.g., if source contains \u2019, then dest will
210 // contain the three bytes 0xE2, 0x80, and 0x99). For the inverse
211 // transformation, use UniLib::UTF8EscapeString
212 // (util/utf8/unilib.h), not CEscapeString.
213 //
214 // Errors: In the first form of the call, errors are reported with
215 // LOG(ERROR). The same is true for the second form of the call if
216 // the pointer to the string vector is NULL; otherwise, error
217 // messages are stored in the vector. In either case, the effect on
218 // the dest array is not defined, but rest of the source will be
219 // processed.
220 // ----------------------------------------------------------------------
221
222 LIBPROTOBUF_EXPORT int UnescapeCEscapeSequences(const char* source, char* dest);
223 LIBPROTOBUF_EXPORT int UnescapeCEscapeSequences(const char* source, char* dest,
224 vector<string> *errors);
225
226 // ----------------------------------------------------------------------
227 // UnescapeCEscapeString()
228 // This does the same thing as UnescapeCEscapeSequences, but creates
229 // a new string. The caller does not need to worry about allocating
230 // a dest buffer. This should be used for non performance critical
231 // tasks such as printing debug messages. It is safe for src and dest
232 // to be the same.
233 //
234 // The second call stores its errors in a supplied string vector.
235 // If the string vector pointer is NULL, it reports the errors with LOG().
236 //
237 // In the first and second calls, the length of dest is returned. In the
238 // the third call, the new string is returned.
239 // ----------------------------------------------------------------------
240
241 LIBPROTOBUF_EXPORT int UnescapeCEscapeString(const string& src, string* dest);
242 LIBPROTOBUF_EXPORT int UnescapeCEscapeString(const string& src, string* dest,
243 vector<string> *errors);
244 LIBPROTOBUF_EXPORT string UnescapeCEscapeString(const string& src);
245
246 // ----------------------------------------------------------------------
247 // CEscapeString()
248 // Copies 'src' to 'dest', escaping dangerous characters using
249 // C-style escape sequences. This is very useful for preparing query
250 // flags. 'src' and 'dest' should not overlap.
251 // Returns the number of bytes written to 'dest' (not including the \0)
252 // or -1 if there was insufficient space.
253 //
254 // Currently only \n, \r, \t, ", ', \ and !isprint() chars are escaped.
255 // ----------------------------------------------------------------------
256 LIBPROTOBUF_EXPORT int CEscapeString(const char* src, int src_len,
257 char* dest, int dest_len);
258
259 // ----------------------------------------------------------------------
260 // CEscape()
261 // More convenient form of CEscapeString: returns result as a "string".
262 // This version is slower than CEscapeString() because it does more
263 // allocation. However, it is much more convenient to use in
264 // non-speed-critical code like logging messages etc.
265 // ----------------------------------------------------------------------
266 LIBPROTOBUF_EXPORT string CEscape(const string& src);
267
268 namespace strings {
269 // Like CEscape() but does not escape bytes with the upper bit set.
270 LIBPROTOBUF_EXPORT string Utf8SafeCEscape(const string& src);
271
272 // Like CEscape() but uses hex (\x) escapes instead of octals.
273 LIBPROTOBUF_EXPORT string CHexEscape(const string& src);
274 } // namespace strings
275
276 // ----------------------------------------------------------------------
277 // strto32()
278 // strtou32()
279 // strto64()
280 // strtou64()
281 // Architecture-neutral plug compatible replacements for strtol() and
282 // strtoul(). Long's have different lengths on ILP-32 and LP-64
283 // platforms, so using these is safer, from the point of view of
284 // overflow behavior, than using the standard libc functions.
285 // ----------------------------------------------------------------------
286 LIBPROTOBUF_EXPORT int32 strto32_adaptor(const char *nptr, char **endptr,
287 int base);
288 LIBPROTOBUF_EXPORT uint32 strtou32_adaptor(const char *nptr, char **endptr,
289 int base);
290
strto32(const char * nptr,char ** endptr,int base)291 inline int32 strto32(const char *nptr, char **endptr, int base) {
292 if (sizeof(int32) == sizeof(long))
293 return strtol(nptr, endptr, base);
294 else
295 return strto32_adaptor(nptr, endptr, base);
296 }
297
strtou32(const char * nptr,char ** endptr,int base)298 inline uint32 strtou32(const char *nptr, char **endptr, int base) {
299 if (sizeof(uint32) == sizeof(unsigned long))
300 return strtoul(nptr, endptr, base);
301 else
302 return strtou32_adaptor(nptr, endptr, base);
303 }
304
305 // For now, long long is 64-bit on all the platforms we care about, so these
306 // functions can simply pass the call to strto[u]ll.
strto64(const char * nptr,char ** endptr,int base)307 inline int64 strto64(const char *nptr, char **endptr, int base) {
308 GOOGLE_COMPILE_ASSERT(sizeof(int64) == sizeof(long long),
309 sizeof_int64_is_not_sizeof_long_long);
310 return strtoll(nptr, endptr, base);
311 }
312
strtou64(const char * nptr,char ** endptr,int base)313 inline uint64 strtou64(const char *nptr, char **endptr, int base) {
314 GOOGLE_COMPILE_ASSERT(sizeof(uint64) == sizeof(unsigned long long),
315 sizeof_uint64_is_not_sizeof_long_long);
316 return strtoull(nptr, endptr, base);
317 }
318
319 // ----------------------------------------------------------------------
320 // FastIntToBuffer()
321 // FastHexToBuffer()
322 // FastHex64ToBuffer()
323 // FastHex32ToBuffer()
324 // FastTimeToBuffer()
325 // These are intended for speed. FastIntToBuffer() assumes the
326 // integer is non-negative. FastHexToBuffer() puts output in
327 // hex rather than decimal. FastTimeToBuffer() puts the output
328 // into RFC822 format.
329 //
330 // FastHex64ToBuffer() puts a 64-bit unsigned value in hex-format,
331 // padded to exactly 16 bytes (plus one byte for '\0')
332 //
333 // FastHex32ToBuffer() puts a 32-bit unsigned value in hex-format,
334 // padded to exactly 8 bytes (plus one byte for '\0')
335 //
336 // All functions take the output buffer as an arg.
337 // They all return a pointer to the beginning of the output,
338 // which may not be the beginning of the input buffer.
339 // ----------------------------------------------------------------------
340
341 // Suggested buffer size for FastToBuffer functions. Also works with
342 // DoubleToBuffer() and FloatToBuffer().
343 static const int kFastToBufferSize = 32;
344
345 LIBPROTOBUF_EXPORT char* FastInt32ToBuffer(int32 i, char* buffer);
346 LIBPROTOBUF_EXPORT char* FastInt64ToBuffer(int64 i, char* buffer);
347 char* FastUInt32ToBuffer(uint32 i, char* buffer); // inline below
348 char* FastUInt64ToBuffer(uint64 i, char* buffer); // inline below
349 LIBPROTOBUF_EXPORT char* FastHexToBuffer(int i, char* buffer);
350 LIBPROTOBUF_EXPORT char* FastHex64ToBuffer(uint64 i, char* buffer);
351 LIBPROTOBUF_EXPORT char* FastHex32ToBuffer(uint32 i, char* buffer);
352
353 // at least 22 bytes long
FastIntToBuffer(int i,char * buffer)354 inline char* FastIntToBuffer(int i, char* buffer) {
355 return (sizeof(i) == 4 ?
356 FastInt32ToBuffer(i, buffer) : FastInt64ToBuffer(i, buffer));
357 }
FastUIntToBuffer(unsigned int i,char * buffer)358 inline char* FastUIntToBuffer(unsigned int i, char* buffer) {
359 return (sizeof(i) == 4 ?
360 FastUInt32ToBuffer(i, buffer) : FastUInt64ToBuffer(i, buffer));
361 }
FastLongToBuffer(long i,char * buffer)362 inline char* FastLongToBuffer(long i, char* buffer) {
363 return (sizeof(i) == 4 ?
364 FastInt32ToBuffer(i, buffer) : FastInt64ToBuffer(i, buffer));
365 }
FastULongToBuffer(unsigned long i,char * buffer)366 inline char* FastULongToBuffer(unsigned long i, char* buffer) {
367 return (sizeof(i) == 4 ?
368 FastUInt32ToBuffer(i, buffer) : FastUInt64ToBuffer(i, buffer));
369 }
370
371 // ----------------------------------------------------------------------
372 // FastInt32ToBufferLeft()
373 // FastUInt32ToBufferLeft()
374 // FastInt64ToBufferLeft()
375 // FastUInt64ToBufferLeft()
376 //
377 // Like the Fast*ToBuffer() functions above, these are intended for speed.
378 // Unlike the Fast*ToBuffer() functions, however, these functions write
379 // their output to the beginning of the buffer (hence the name, as the
380 // output is left-aligned). The caller is responsible for ensuring that
381 // the buffer has enough space to hold the output.
382 //
383 // Returns a pointer to the end of the string (i.e. the null character
384 // terminating the string).
385 // ----------------------------------------------------------------------
386
387 LIBPROTOBUF_EXPORT char* FastInt32ToBufferLeft(int32 i, char* buffer);
388 LIBPROTOBUF_EXPORT char* FastUInt32ToBufferLeft(uint32 i, char* buffer);
389 LIBPROTOBUF_EXPORT char* FastInt64ToBufferLeft(int64 i, char* buffer);
390 LIBPROTOBUF_EXPORT char* FastUInt64ToBufferLeft(uint64 i, char* buffer);
391
392 // Just define these in terms of the above.
FastUInt32ToBuffer(uint32 i,char * buffer)393 inline char* FastUInt32ToBuffer(uint32 i, char* buffer) {
394 FastUInt32ToBufferLeft(i, buffer);
395 return buffer;
396 }
FastUInt64ToBuffer(uint64 i,char * buffer)397 inline char* FastUInt64ToBuffer(uint64 i, char* buffer) {
398 FastUInt64ToBufferLeft(i, buffer);
399 return buffer;
400 }
401
402 // ----------------------------------------------------------------------
403 // SimpleItoa()
404 // Description: converts an integer to a string.
405 //
406 // Return value: string
407 // ----------------------------------------------------------------------
408 LIBPROTOBUF_EXPORT string SimpleItoa(int i);
409 LIBPROTOBUF_EXPORT string SimpleItoa(unsigned int i);
410 LIBPROTOBUF_EXPORT string SimpleItoa(long i);
411 LIBPROTOBUF_EXPORT string SimpleItoa(unsigned long i);
412 LIBPROTOBUF_EXPORT string SimpleItoa(long long i);
413 LIBPROTOBUF_EXPORT string SimpleItoa(unsigned long long i);
414
415 // ----------------------------------------------------------------------
416 // SimpleDtoa()
417 // SimpleFtoa()
418 // DoubleToBuffer()
419 // FloatToBuffer()
420 // Description: converts a double or float to a string which, if
421 // passed to NoLocaleStrtod(), will produce the exact same original double
422 // (except in case of NaN; all NaNs are considered the same value).
423 // We try to keep the string short but it's not guaranteed to be as
424 // short as possible.
425 //
426 // DoubleToBuffer() and FloatToBuffer() write the text to the given
427 // buffer and return it. The buffer must be at least
428 // kDoubleToBufferSize bytes for doubles and kFloatToBufferSize
429 // bytes for floats. kFastToBufferSize is also guaranteed to be large
430 // enough to hold either.
431 //
432 // Return value: string
433 // ----------------------------------------------------------------------
434 LIBPROTOBUF_EXPORT string SimpleDtoa(double value);
435 LIBPROTOBUF_EXPORT string SimpleFtoa(float value);
436
437 LIBPROTOBUF_EXPORT char* DoubleToBuffer(double i, char* buffer);
438 LIBPROTOBUF_EXPORT char* FloatToBuffer(float i, char* buffer);
439
440 // In practice, doubles should never need more than 24 bytes and floats
441 // should never need more than 14 (including null terminators), but we
442 // overestimate to be safe.
443 static const int kDoubleToBufferSize = 32;
444 static const int kFloatToBufferSize = 24;
445
446 // ----------------------------------------------------------------------
447 // NoLocaleStrtod()
448 // Exactly like strtod(), except it always behaves as if in the "C"
449 // locale (i.e. decimal points must be '.'s).
450 // ----------------------------------------------------------------------
451
452 LIBPROTOBUF_EXPORT double NoLocaleStrtod(const char* text, char** endptr);
453
454 } // namespace protobuf
455 } // namespace google
456
457 #endif // GOOGLE_PROTOBUF_STUBS_STRUTIL_H__
458
459
460