Line data Source code
1 : // Protocol Buffers - Google's data interchange format
2 : // Copyright 2008 Google Inc. All rights reserved.
3 : // https://developers.google.com/protocol-buffers/
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 : #include <google/protobuf/stubs/stringpiece.h>
40 :
41 : namespace google {
42 : namespace protobuf {
43 :
44 : #ifdef _MSC_VER
45 : #define strtoll _strtoi64
46 : #define strtoull _strtoui64
47 : #elif defined(__DECCXX) && defined(__osf__)
48 : // HP C++ on Tru64 does not have strtoll, but strtol is already 64-bit.
49 : #define strtoll strtol
50 : #define strtoull strtoul
51 : #endif
52 :
53 : // ----------------------------------------------------------------------
54 : // ascii_isalnum()
55 : // Check if an ASCII character is alphanumeric. We can't use ctype's
56 : // isalnum() because it is affected by locale. This function is applied
57 : // to identifiers in the protocol buffer language, not to natural-language
58 : // strings, so locale should not be taken into account.
59 : // ascii_isdigit()
60 : // Like above, but only accepts digits.
61 : // ascii_isspace()
62 : // Check if the character is a space character.
63 : // ----------------------------------------------------------------------
64 :
65 : inline bool ascii_isalnum(char c) {
66 164522 : return ('a' <= c && c <= 'z') ||
67 174623 : ('A' <= c && c <= 'Z') ||
68 10101 : ('0' <= c && c <= '9');
69 : }
70 :
71 : inline bool ascii_isdigit(char c) {
72 128408 : return ('0' <= c && c <= '9');
73 : }
74 :
75 : inline bool ascii_isspace(char c) {
76 0 : return c == ' ' || c == '\t' || c == '\n' || c == '\v' || c == '\f' ||
77 : c == '\r';
78 : }
79 :
80 : inline bool ascii_isupper(char c) {
81 0 : return c >= 'A' && c <= 'Z';
82 : }
83 :
84 : inline bool ascii_islower(char c) {
85 0 : return c >= 'a' && c <= 'z';
86 : }
87 :
88 : inline char ascii_toupper(char c) {
89 0 : return ascii_islower(c) ? c - ('a' - 'A') : c;
90 : }
91 :
92 : inline char ascii_tolower(char c) {
93 0 : return ascii_isupper(c) ? c + ('a' - 'A') : c;
94 : }
95 :
96 : inline int hex_digit_to_int(char c) {
97 : /* Assume ASCII. */
98 0 : int x = static_cast<unsigned char>(c);
99 0 : if (x > '9') {
100 0 : x += 9;
101 : }
102 0 : return x & 0xf;
103 : }
104 :
105 : // ----------------------------------------------------------------------
106 : // HasPrefixString()
107 : // Check if a string begins with a given prefix.
108 : // StripPrefixString()
109 : // Given a string and a putative prefix, returns the string minus the
110 : // prefix string if the prefix matches, otherwise the original
111 : // string.
112 : // ----------------------------------------------------------------------
113 2342 : inline bool HasPrefixString(const string& str,
114 : const string& prefix) {
115 6269 : return str.size() >= prefix.size() &&
116 3927 : str.compare(0, prefix.size(), prefix) == 0;
117 : }
118 :
119 0 : inline string StripPrefixString(const string& str, const string& prefix) {
120 0 : if (HasPrefixString(str, prefix)) {
121 0 : return str.substr(prefix.size());
122 : } else {
123 0 : return str;
124 : }
125 : }
126 :
127 : // ----------------------------------------------------------------------
128 : // HasSuffixString()
129 : // Return true if str ends in suffix.
130 : // StripSuffixString()
131 : // Given a string and a putative suffix, returns the string minus the
132 : // suffix string if the suffix matches, otherwise the original
133 : // string.
134 : // ----------------------------------------------------------------------
135 578 : inline bool HasSuffixString(const string& str,
136 : const string& suffix) {
137 1734 : return str.size() >= suffix.size() &&
138 1156 : str.compare(str.size() - suffix.size(), suffix.size(), suffix) == 0;
139 : }
140 :
141 149 : inline string StripSuffixString(const string& str, const string& suffix) {
142 149 : if (HasSuffixString(str, suffix)) {
143 298 : return str.substr(0, str.size() - suffix.size());
144 : } else {
145 0 : return str;
146 : }
147 : }
148 :
149 : // ----------------------------------------------------------------------
150 : // StripString
151 : // Replaces any occurrence of the character 'remove' (or the characters
152 : // in 'remove') with the character 'replacewith'.
153 : // Good for keeping html characters or protocol characters (\t) out
154 : // of places where they might cause a problem.
155 : // StripWhitespace
156 : // Removes whitespaces from both ends of the given string.
157 : // ----------------------------------------------------------------------
158 : LIBPROTOBUF_EXPORT void StripString(string* s, const char* remove,
159 : char replacewith);
160 :
161 : LIBPROTOBUF_EXPORT void StripWhitespace(string* s);
162 :
163 :
164 : // ----------------------------------------------------------------------
165 : // LowerString()
166 : // UpperString()
167 : // ToUpper()
168 : // Convert the characters in "s" to lowercase or uppercase. ASCII-only:
169 : // these functions intentionally ignore locale because they are applied to
170 : // identifiers used in the Protocol Buffer language, not to natural-language
171 : // strings.
172 : // ----------------------------------------------------------------------
173 :
174 52042 : inline void LowerString(string * s) {
175 52042 : string::iterator end = s->end();
176 2718842 : for (string::iterator i = s->begin(); i != end; ++i) {
177 : // tolower() changes based on locale. We don't want this!
178 2614758 : if ('A' <= *i && *i <= 'Z') *i += 'a' - 'A';
179 : }
180 52042 : }
181 :
182 546 : inline void UpperString(string * s) {
183 546 : string::iterator end = s->end();
184 4292 : for (string::iterator i = s->begin(); i != end; ++i) {
185 : // toupper() changes based on locale. We don't want this!
186 3200 : if ('a' <= *i && *i <= 'z') *i += 'A' - 'a';
187 : }
188 546 : }
189 :
190 546 : inline string ToUpper(const string& s) {
191 546 : string out = s;
192 546 : UpperString(&out);
193 546 : return out;
194 : }
195 :
196 : // ----------------------------------------------------------------------
197 : // StringReplace()
198 : // Give me a string and two patterns "old" and "new", and I replace
199 : // the first instance of "old" in the string with "new", if it
200 : // exists. RETURN a new string, regardless of whether the replacement
201 : // happened or not.
202 : // ----------------------------------------------------------------------
203 :
204 : LIBPROTOBUF_EXPORT string StringReplace(const string& s, const string& oldsub,
205 : const string& newsub, bool replace_all);
206 :
207 : // ----------------------------------------------------------------------
208 : // SplitStringUsing()
209 : // Split a string using a character delimiter. Append the components
210 : // to 'result'. If there are consecutive delimiters, this function skips
211 : // over all of them.
212 : // ----------------------------------------------------------------------
213 : LIBPROTOBUF_EXPORT void SplitStringUsing(const string& full, const char* delim,
214 : vector<string>* res);
215 :
216 : // Split a string using one or more byte delimiters, presented
217 : // as a nul-terminated c string. Append the components to 'result'.
218 : // If there are consecutive delimiters, this function will return
219 : // corresponding empty strings. If you want to drop the empty
220 : // strings, try SplitStringUsing().
221 : //
222 : // If "full" is the empty string, yields an empty string as the only value.
223 : // ----------------------------------------------------------------------
224 : LIBPROTOBUF_EXPORT void SplitStringAllowEmpty(const string& full,
225 : const char* delim,
226 : vector<string>* result);
227 :
228 : // ----------------------------------------------------------------------
229 : // Split()
230 : // Split a string using a character delimiter.
231 : // ----------------------------------------------------------------------
232 317 : inline vector<string> Split(
233 : const string& full, const char* delim, bool skip_empty = true) {
234 : vector<string> result;
235 317 : if (skip_empty) {
236 317 : SplitStringUsing(full, delim, &result);
237 : } else {
238 0 : SplitStringAllowEmpty(full, delim, &result);
239 : }
240 317 : return result;
241 : }
242 :
243 : // ----------------------------------------------------------------------
244 : // JoinStrings()
245 : // These methods concatenate a vector of strings into a C++ string, using
246 : // the C-string "delim" as a separator between components. There are two
247 : // flavors of the function, one flavor returns the concatenated string,
248 : // another takes a pointer to the target string. In the latter case the
249 : // target string is cleared and overwritten.
250 : // ----------------------------------------------------------------------
251 : LIBPROTOBUF_EXPORT void JoinStrings(const vector<string>& components,
252 : const char* delim, string* result);
253 :
254 0 : inline string JoinStrings(const vector<string>& components,
255 : const char* delim) {
256 : string result;
257 0 : JoinStrings(components, delim, &result);
258 0 : return result;
259 : }
260 :
261 : // ----------------------------------------------------------------------
262 : // UnescapeCEscapeSequences()
263 : // Copies "source" to "dest", rewriting C-style escape sequences
264 : // -- '\n', '\r', '\\', '\ooo', etc -- to their ASCII
265 : // equivalents. "dest" must be sufficiently large to hold all
266 : // the characters in the rewritten string (i.e. at least as large
267 : // as strlen(source) + 1 should be safe, since the replacements
268 : // are always shorter than the original escaped sequences). It's
269 : // safe for source and dest to be the same. RETURNS the length
270 : // of dest.
271 : //
272 : // It allows hex sequences \xhh, or generally \xhhhhh with an
273 : // arbitrary number of hex digits, but all of them together must
274 : // specify a value of a single byte (e.g. \x0045 is equivalent
275 : // to \x45, and \x1234 is erroneous).
276 : //
277 : // It also allows escape sequences of the form \uhhhh (exactly four
278 : // hex digits, upper or lower case) or \Uhhhhhhhh (exactly eight
279 : // hex digits, upper or lower case) to specify a Unicode code
280 : // point. The dest array will contain the UTF8-encoded version of
281 : // that code-point (e.g., if source contains \u2019, then dest will
282 : // contain the three bytes 0xE2, 0x80, and 0x99).
283 : //
284 : // Errors: In the first form of the call, errors are reported with
285 : // LOG(ERROR). The same is true for the second form of the call if
286 : // the pointer to the string vector is NULL; otherwise, error
287 : // messages are stored in the vector. In either case, the effect on
288 : // the dest array is not defined, but rest of the source will be
289 : // processed.
290 : // ----------------------------------------------------------------------
291 :
292 : LIBPROTOBUF_EXPORT int UnescapeCEscapeSequences(const char* source, char* dest);
293 : LIBPROTOBUF_EXPORT int UnescapeCEscapeSequences(const char* source, char* dest,
294 : vector<string> *errors);
295 :
296 : // ----------------------------------------------------------------------
297 : // UnescapeCEscapeString()
298 : // This does the same thing as UnescapeCEscapeSequences, but creates
299 : // a new string. The caller does not need to worry about allocating
300 : // a dest buffer. This should be used for non performance critical
301 : // tasks such as printing debug messages. It is safe for src and dest
302 : // to be the same.
303 : //
304 : // The second call stores its errors in a supplied string vector.
305 : // If the string vector pointer is NULL, it reports the errors with LOG().
306 : //
307 : // In the first and second calls, the length of dest is returned. In the
308 : // the third call, the new string is returned.
309 : // ----------------------------------------------------------------------
310 :
311 : LIBPROTOBUF_EXPORT int UnescapeCEscapeString(const string& src, string* dest);
312 : LIBPROTOBUF_EXPORT int UnescapeCEscapeString(const string& src, string* dest,
313 : vector<string> *errors);
314 : LIBPROTOBUF_EXPORT string UnescapeCEscapeString(const string& src);
315 :
316 : // ----------------------------------------------------------------------
317 : // CEscapeString()
318 : // Copies 'src' to 'dest', escaping dangerous characters using
319 : // C-style escape sequences. This is very useful for preparing query
320 : // flags. 'src' and 'dest' should not overlap.
321 : // Returns the number of bytes written to 'dest' (not including the \0)
322 : // or -1 if there was insufficient space.
323 : //
324 : // Currently only \n, \r, \t, ", ', \ and !isprint() chars are escaped.
325 : // ----------------------------------------------------------------------
326 : LIBPROTOBUF_EXPORT int CEscapeString(const char* src, int src_len,
327 : char* dest, int dest_len);
328 :
329 : // ----------------------------------------------------------------------
330 : // CEscape()
331 : // More convenient form of CEscapeString: returns result as a "string".
332 : // This version is slower than CEscapeString() because it does more
333 : // allocation. However, it is much more convenient to use in
334 : // non-speed-critical code like logging messages etc.
335 : // ----------------------------------------------------------------------
336 : LIBPROTOBUF_EXPORT string CEscape(const string& src);
337 :
338 : namespace strings {
339 : // Like CEscape() but does not escape bytes with the upper bit set.
340 : LIBPROTOBUF_EXPORT string Utf8SafeCEscape(const string& src);
341 :
342 : // Like CEscape() but uses hex (\x) escapes instead of octals.
343 : LIBPROTOBUF_EXPORT string CHexEscape(const string& src);
344 : } // namespace strings
345 :
346 : // ----------------------------------------------------------------------
347 : // strto32()
348 : // strtou32()
349 : // strto64()
350 : // strtou64()
351 : // Architecture-neutral plug compatible replacements for strtol() and
352 : // strtoul(). Long's have different lengths on ILP-32 and LP-64
353 : // platforms, so using these is safer, from the point of view of
354 : // overflow behavior, than using the standard libc functions.
355 : // ----------------------------------------------------------------------
356 : LIBPROTOBUF_EXPORT int32 strto32_adaptor(const char *nptr, char **endptr,
357 : int base);
358 : LIBPROTOBUF_EXPORT uint32 strtou32_adaptor(const char *nptr, char **endptr,
359 : int base);
360 :
361 : inline int32 strto32(const char *nptr, char **endptr, int base) {
362 : if (sizeof(int32) == sizeof(long))
363 : return strtol(nptr, endptr, base);
364 : else
365 : return strto32_adaptor(nptr, endptr, base);
366 : }
367 :
368 : inline uint32 strtou32(const char *nptr, char **endptr, int base) {
369 : if (sizeof(uint32) == sizeof(unsigned long))
370 : return strtoul(nptr, endptr, base);
371 : else
372 : return strtou32_adaptor(nptr, endptr, base);
373 : }
374 :
375 : // For now, long long is 64-bit on all the platforms we care about, so these
376 : // functions can simply pass the call to strto[u]ll.
377 : inline int64 strto64(const char *nptr, char **endptr, int base) {
378 : GOOGLE_COMPILE_ASSERT(sizeof(int64) == sizeof(long long),
379 : sizeof_int64_is_not_sizeof_long_long);
380 17 : return strtoll(nptr, endptr, base);
381 : }
382 :
383 : inline uint64 strtou64(const char *nptr, char **endptr, int base) {
384 : GOOGLE_COMPILE_ASSERT(sizeof(uint64) == sizeof(unsigned long long),
385 : sizeof_uint64_is_not_sizeof_long_long);
386 11 : return strtoull(nptr, endptr, base);
387 : }
388 :
389 : // ----------------------------------------------------------------------
390 : // safe_strtob()
391 : // safe_strto32()
392 : // safe_strtou32()
393 : // safe_strto64()
394 : // safe_strtou64()
395 : // safe_strtof()
396 : // safe_strtod()
397 : // ----------------------------------------------------------------------
398 : LIBPROTOBUF_EXPORT bool safe_strtob(StringPiece str, bool* value);
399 :
400 : LIBPROTOBUF_EXPORT bool safe_strto32(const string& str, int32* value);
401 : LIBPROTOBUF_EXPORT bool safe_strtou32(const string& str, uint32* value);
402 : inline bool safe_strto32(const char* str, int32* value) {
403 : return safe_strto32(string(str), value);
404 : }
405 : inline bool safe_strto32(StringPiece str, int32* value) {
406 : return safe_strto32(str.ToString(), value);
407 : }
408 : inline bool safe_strtou32(const char* str, uint32* value) {
409 : return safe_strtou32(string(str), value);
410 : }
411 : inline bool safe_strtou32(StringPiece str, uint32* value) {
412 : return safe_strtou32(str.ToString(), value);
413 : }
414 :
415 : LIBPROTOBUF_EXPORT bool safe_strto64(const string& str, int64* value);
416 : LIBPROTOBUF_EXPORT bool safe_strtou64(const string& str, uint64* value);
417 : inline bool safe_strto64(const char* str, int64* value) {
418 : return safe_strto64(string(str), value);
419 : }
420 : inline bool safe_strto64(StringPiece str, int64* value) {
421 : return safe_strto64(str.ToString(), value);
422 : }
423 : inline bool safe_strtou64(const char* str, uint64* value) {
424 : return safe_strtou64(string(str), value);
425 : }
426 : inline bool safe_strtou64(StringPiece str, uint64* value) {
427 : return safe_strtou64(str.ToString(), value);
428 : }
429 :
430 : LIBPROTOBUF_EXPORT bool safe_strtof(const char* str, float* value);
431 : LIBPROTOBUF_EXPORT bool safe_strtod(const char* str, double* value);
432 : inline bool safe_strtof(const string& str, float* value) {
433 : return safe_strtof(str.c_str(), value);
434 : }
435 : inline bool safe_strtod(const string& str, double* value) {
436 : return safe_strtod(str.c_str(), value);
437 : }
438 : inline bool safe_strtof(StringPiece str, float* value) {
439 : return safe_strtof(str.ToString(), value);
440 : }
441 : inline bool safe_strtod(StringPiece str, double* value) {
442 : return safe_strtod(str.ToString(), value);
443 : }
444 :
445 : // ----------------------------------------------------------------------
446 : // FastIntToBuffer()
447 : // FastHexToBuffer()
448 : // FastHex64ToBuffer()
449 : // FastHex32ToBuffer()
450 : // FastTimeToBuffer()
451 : // These are intended for speed. FastIntToBuffer() assumes the
452 : // integer is non-negative. FastHexToBuffer() puts output in
453 : // hex rather than decimal. FastTimeToBuffer() puts the output
454 : // into RFC822 format.
455 : //
456 : // FastHex64ToBuffer() puts a 64-bit unsigned value in hex-format,
457 : // padded to exactly 16 bytes (plus one byte for '\0')
458 : //
459 : // FastHex32ToBuffer() puts a 32-bit unsigned value in hex-format,
460 : // padded to exactly 8 bytes (plus one byte for '\0')
461 : //
462 : // All functions take the output buffer as an arg.
463 : // They all return a pointer to the beginning of the output,
464 : // which may not be the beginning of the input buffer.
465 : // ----------------------------------------------------------------------
466 :
467 : // Suggested buffer size for FastToBuffer functions. Also works with
468 : // DoubleToBuffer() and FloatToBuffer().
469 : static const int kFastToBufferSize = 32;
470 :
471 : LIBPROTOBUF_EXPORT char* FastInt32ToBuffer(int32 i, char* buffer);
472 : LIBPROTOBUF_EXPORT char* FastInt64ToBuffer(int64 i, char* buffer);
473 : char* FastUInt32ToBuffer(uint32 i, char* buffer); // inline below
474 : char* FastUInt64ToBuffer(uint64 i, char* buffer); // inline below
475 : LIBPROTOBUF_EXPORT char* FastHexToBuffer(int i, char* buffer);
476 : LIBPROTOBUF_EXPORT char* FastHex64ToBuffer(uint64 i, char* buffer);
477 : LIBPROTOBUF_EXPORT char* FastHex32ToBuffer(uint32 i, char* buffer);
478 :
479 : // at least 22 bytes long
480 : inline char* FastIntToBuffer(int i, char* buffer) {
481 : return (sizeof(i) == 4 ?
482 : FastInt32ToBuffer(i, buffer) : FastInt64ToBuffer(i, buffer));
483 : }
484 : inline char* FastUIntToBuffer(unsigned int i, char* buffer) {
485 : return (sizeof(i) == 4 ?
486 : FastUInt32ToBuffer(i, buffer) : FastUInt64ToBuffer(i, buffer));
487 : }
488 : inline char* FastLongToBuffer(long i, char* buffer) {
489 : return (sizeof(i) == 4 ?
490 : FastInt32ToBuffer(i, buffer) : FastInt64ToBuffer(i, buffer));
491 : }
492 : inline char* FastULongToBuffer(unsigned long i, char* buffer) {
493 : return (sizeof(i) == 4 ?
494 : FastUInt32ToBuffer(i, buffer) : FastUInt64ToBuffer(i, buffer));
495 : }
496 :
497 : // ----------------------------------------------------------------------
498 : // FastInt32ToBufferLeft()
499 : // FastUInt32ToBufferLeft()
500 : // FastInt64ToBufferLeft()
501 : // FastUInt64ToBufferLeft()
502 : //
503 : // Like the Fast*ToBuffer() functions above, these are intended for speed.
504 : // Unlike the Fast*ToBuffer() functions, however, these functions write
505 : // their output to the beginning of the buffer (hence the name, as the
506 : // output is left-aligned). The caller is responsible for ensuring that
507 : // the buffer has enough space to hold the output.
508 : //
509 : // Returns a pointer to the end of the string (i.e. the null character
510 : // terminating the string).
511 : // ----------------------------------------------------------------------
512 :
513 : LIBPROTOBUF_EXPORT char* FastInt32ToBufferLeft(int32 i, char* buffer);
514 : LIBPROTOBUF_EXPORT char* FastUInt32ToBufferLeft(uint32 i, char* buffer);
515 : LIBPROTOBUF_EXPORT char* FastInt64ToBufferLeft(int64 i, char* buffer);
516 : LIBPROTOBUF_EXPORT char* FastUInt64ToBufferLeft(uint64 i, char* buffer);
517 :
518 : // Just define these in terms of the above.
519 : inline char* FastUInt32ToBuffer(uint32 i, char* buffer) {
520 : FastUInt32ToBufferLeft(i, buffer);
521 : return buffer;
522 : }
523 : inline char* FastUInt64ToBuffer(uint64 i, char* buffer) {
524 : FastUInt64ToBufferLeft(i, buffer);
525 : return buffer;
526 : }
527 :
528 : inline string SimpleBtoa(bool value) {
529 : return value ? "true" : "false";
530 : }
531 :
532 : // ----------------------------------------------------------------------
533 : // SimpleItoa()
534 : // Description: converts an integer to a string.
535 : //
536 : // Return value: string
537 : // ----------------------------------------------------------------------
538 : LIBPROTOBUF_EXPORT string SimpleItoa(int i);
539 : LIBPROTOBUF_EXPORT string SimpleItoa(unsigned int i);
540 : LIBPROTOBUF_EXPORT string SimpleItoa(long i);
541 : LIBPROTOBUF_EXPORT string SimpleItoa(unsigned long i);
542 : LIBPROTOBUF_EXPORT string SimpleItoa(long long i);
543 : LIBPROTOBUF_EXPORT string SimpleItoa(unsigned long long i);
544 :
545 : // ----------------------------------------------------------------------
546 : // SimpleDtoa()
547 : // SimpleFtoa()
548 : // DoubleToBuffer()
549 : // FloatToBuffer()
550 : // Description: converts a double or float to a string which, if
551 : // passed to NoLocaleStrtod(), will produce the exact same original double
552 : // (except in case of NaN; all NaNs are considered the same value).
553 : // We try to keep the string short but it's not guaranteed to be as
554 : // short as possible.
555 : //
556 : // DoubleToBuffer() and FloatToBuffer() write the text to the given
557 : // buffer and return it. The buffer must be at least
558 : // kDoubleToBufferSize bytes for doubles and kFloatToBufferSize
559 : // bytes for floats. kFastToBufferSize is also guaranteed to be large
560 : // enough to hold either.
561 : //
562 : // Return value: string
563 : // ----------------------------------------------------------------------
564 : LIBPROTOBUF_EXPORT string SimpleDtoa(double value);
565 : LIBPROTOBUF_EXPORT string SimpleFtoa(float value);
566 :
567 : LIBPROTOBUF_EXPORT char* DoubleToBuffer(double i, char* buffer);
568 : LIBPROTOBUF_EXPORT char* FloatToBuffer(float i, char* buffer);
569 :
570 : // In practice, doubles should never need more than 24 bytes and floats
571 : // should never need more than 14 (including null terminators), but we
572 : // overestimate to be safe.
573 : static const int kDoubleToBufferSize = 32;
574 : static const int kFloatToBufferSize = 24;
575 :
576 : namespace strings {
577 :
578 : enum PadSpec {
579 : NO_PAD = 1,
580 : ZERO_PAD_2,
581 : ZERO_PAD_3,
582 : ZERO_PAD_4,
583 : ZERO_PAD_5,
584 : ZERO_PAD_6,
585 : ZERO_PAD_7,
586 : ZERO_PAD_8,
587 : ZERO_PAD_9,
588 : ZERO_PAD_10,
589 : ZERO_PAD_11,
590 : ZERO_PAD_12,
591 : ZERO_PAD_13,
592 : ZERO_PAD_14,
593 : ZERO_PAD_15,
594 : ZERO_PAD_16,
595 : };
596 :
597 : struct Hex {
598 : uint64 value;
599 : enum PadSpec spec;
600 : template <class Int>
601 : explicit Hex(Int v, PadSpec s = NO_PAD)
602 12771 : : spec(s) {
603 : // Prevent sign-extension by casting integers to
604 : // their unsigned counterparts.
605 : #ifdef LANG_CXX11
606 : static_assert(
607 : sizeof(v) == 1 || sizeof(v) == 2 || sizeof(v) == 4 || sizeof(v) == 8,
608 : "Unknown integer type");
609 : #endif
610 12771 : value = sizeof(v) == 1 ? static_cast<uint8>(v)
611 : : sizeof(v) == 2 ? static_cast<uint16>(v)
612 : : sizeof(v) == 4 ? static_cast<uint32>(v)
613 : : static_cast<uint64>(v);
614 : }
615 : };
616 :
617 : struct LIBPROTOBUF_EXPORT AlphaNum {
618 : const char *piece_data_; // move these to string_ref eventually
619 : size_t piece_size_; // move these to string_ref eventually
620 :
621 : char digits[kFastToBufferSize];
622 :
623 : // No bool ctor -- bools convert to an integral type.
624 : // A bool ctor would also convert incoming pointers (bletch).
625 :
626 : AlphaNum(int32 i32)
627 : : piece_data_(digits),
628 4 : piece_size_(FastInt32ToBufferLeft(i32, digits) - &digits[0]) {}
629 : AlphaNum(uint32 u32)
630 : : piece_data_(digits),
631 : piece_size_(FastUInt32ToBufferLeft(u32, digits) - &digits[0]) {}
632 : AlphaNum(int64 i64)
633 : : piece_data_(digits),
634 : piece_size_(FastInt64ToBufferLeft(i64, digits) - &digits[0]) {}
635 : AlphaNum(uint64 u64)
636 : : piece_data_(digits),
637 : piece_size_(FastUInt64ToBufferLeft(u64, digits) - &digits[0]) {}
638 :
639 : AlphaNum(float f)
640 : : piece_data_(digits), piece_size_(strlen(FloatToBuffer(f, digits))) {}
641 : AlphaNum(double f)
642 : : piece_data_(digits), piece_size_(strlen(DoubleToBuffer(f, digits))) {}
643 :
644 : AlphaNum(Hex hex);
645 :
646 : AlphaNum(const char* c_str)
647 30138 : : piece_data_(c_str), piece_size_(strlen(c_str)) {}
648 : // TODO: Add a string_ref constructor, eventually
649 : // AlphaNum(const StringPiece &pc) : piece(pc) {}
650 :
651 : AlphaNum(const string& str)
652 23032 : : piece_data_(str.data()), piece_size_(str.size()) {}
653 :
654 : AlphaNum(StringPiece str)
655 : : piece_data_(str.data()), piece_size_(str.size()) {}
656 :
657 : size_t size() const { return piece_size_; }
658 : const char *data() const { return piece_data_; }
659 :
660 : private:
661 : // Use ":" not ':'
662 : AlphaNum(char c); // NOLINT(runtime/explicit)
663 :
664 : // Disallow copy and assign.
665 : AlphaNum(const AlphaNum&);
666 : void operator=(const AlphaNum&);
667 : };
668 :
669 : } // namespace strings
670 :
671 : using strings::AlphaNum;
672 :
673 : // ----------------------------------------------------------------------
674 : // StrCat()
675 : // This merges the given strings or numbers, with no delimiter. This
676 : // is designed to be the fastest possible way to construct a string out
677 : // of a mix of raw C strings, strings, bool values,
678 : // and numeric values.
679 : //
680 : // Don't use this for user-visible strings. The localization process
681 : // works poorly on strings built up out of fragments.
682 : //
683 : // For clarity and performance, don't use StrCat when appending to a
684 : // string. In particular, avoid using any of these (anti-)patterns:
685 : // str.append(StrCat(...)
686 : // str += StrCat(...)
687 : // str = StrCat(str, ...)
688 : // where the last is the worse, with the potential to change a loop
689 : // from a linear time operation with O(1) dynamic allocations into a
690 : // quadratic time operation with O(n) dynamic allocations. StrAppend
691 : // is a better choice than any of the above, subject to the restriction
692 : // of StrAppend(&str, a, b, c, ...) that none of the a, b, c, ... may
693 : // be a reference into str.
694 : // ----------------------------------------------------------------------
695 :
696 : LIBPROTOBUF_EXPORT string StrCat(const AlphaNum& a, const AlphaNum& b);
697 : LIBPROTOBUF_EXPORT string StrCat(const AlphaNum& a, const AlphaNum& b,
698 : const AlphaNum& c);
699 : LIBPROTOBUF_EXPORT string StrCat(const AlphaNum& a, const AlphaNum& b,
700 : const AlphaNum& c, const AlphaNum& d);
701 : LIBPROTOBUF_EXPORT string StrCat(const AlphaNum& a, const AlphaNum& b,
702 : const AlphaNum& c, const AlphaNum& d,
703 : const AlphaNum& e);
704 : LIBPROTOBUF_EXPORT string StrCat(const AlphaNum& a, const AlphaNum& b,
705 : const AlphaNum& c, const AlphaNum& d,
706 : const AlphaNum& e, const AlphaNum& f);
707 : LIBPROTOBUF_EXPORT string StrCat(const AlphaNum& a, const AlphaNum& b,
708 : const AlphaNum& c, const AlphaNum& d,
709 : const AlphaNum& e, const AlphaNum& f,
710 : const AlphaNum& g);
711 : LIBPROTOBUF_EXPORT string StrCat(const AlphaNum& a, const AlphaNum& b,
712 : const AlphaNum& c, const AlphaNum& d,
713 : const AlphaNum& e, const AlphaNum& f,
714 : const AlphaNum& g, const AlphaNum& h);
715 : LIBPROTOBUF_EXPORT string StrCat(const AlphaNum& a, const AlphaNum& b,
716 : const AlphaNum& c, const AlphaNum& d,
717 : const AlphaNum& e, const AlphaNum& f,
718 : const AlphaNum& g, const AlphaNum& h,
719 : const AlphaNum& i);
720 :
721 2824 : inline string StrCat(const AlphaNum& a) { return string(a.data(), a.size()); }
722 :
723 : // ----------------------------------------------------------------------
724 : // StrAppend()
725 : // Same as above, but adds the output to the given string.
726 : // WARNING: For speed, StrAppend does not try to check each of its input
727 : // arguments to be sure that they are not a subset of the string being
728 : // appended to. That is, while this will work:
729 : //
730 : // string s = "foo";
731 : // s += s;
732 : //
733 : // This will not (necessarily) work:
734 : //
735 : // string s = "foo";
736 : // StrAppend(&s, s);
737 : //
738 : // Note: while StrCat supports appending up to 9 arguments, StrAppend
739 : // is currently limited to 4. That's rarely an issue except when
740 : // automatically transforming StrCat to StrAppend, and can easily be
741 : // worked around as consecutive calls to StrAppend are quite efficient.
742 : // ----------------------------------------------------------------------
743 :
744 : LIBPROTOBUF_EXPORT void StrAppend(string* dest, const AlphaNum& a);
745 : LIBPROTOBUF_EXPORT void StrAppend(string* dest, const AlphaNum& a,
746 : const AlphaNum& b);
747 : LIBPROTOBUF_EXPORT void StrAppend(string* dest, const AlphaNum& a,
748 : const AlphaNum& b, const AlphaNum& c);
749 : LIBPROTOBUF_EXPORT void StrAppend(string* dest, const AlphaNum& a,
750 : const AlphaNum& b, const AlphaNum& c,
751 : const AlphaNum& d);
752 :
753 : // ----------------------------------------------------------------------
754 : // Join()
755 : // These methods concatenate a range of components into a C++ string, using
756 : // the C-string "delim" as a separator between components.
757 : // ----------------------------------------------------------------------
758 : template <typename Iterator>
759 1174 : void Join(Iterator start, Iterator end,
760 : const char* delim, string* result) {
761 3872 : for (Iterator it = start; it != end; ++it) {
762 1524 : if (it != start) {
763 367 : result->append(delim);
764 : }
765 3048 : StrAppend(result, *it);
766 : }
767 1174 : }
768 :
769 : template <typename Range>
770 3522 : string Join(const Range& components,
771 : const char* delim) {
772 : string result;
773 3522 : Join(components.begin(), components.end(), delim, &result);
774 1174 : return result;
775 : }
776 :
777 : // ----------------------------------------------------------------------
778 : // ToHex()
779 : // Return a lower-case hex string representation of the given integer.
780 : // ----------------------------------------------------------------------
781 : LIBPROTOBUF_EXPORT string ToHex(uint64 num);
782 :
783 : // ----------------------------------------------------------------------
784 : // GlobalReplaceSubstring()
785 : // Replaces all instances of a substring in a string. Does nothing
786 : // if 'substring' is empty. Returns the number of replacements.
787 : //
788 : // NOTE: The string pieces must not overlap s.
789 : // ----------------------------------------------------------------------
790 : LIBPROTOBUF_EXPORT int GlobalReplaceSubstring(const string& substring,
791 : const string& replacement,
792 : string* s);
793 :
794 : // ----------------------------------------------------------------------
795 : // Base64Unescape()
796 : // Converts "src" which is encoded in Base64 to its binary equivalent and
797 : // writes it to "dest". If src contains invalid characters, dest is cleared
798 : // and the function returns false. Returns true on success.
799 : // ----------------------------------------------------------------------
800 : LIBPROTOBUF_EXPORT bool Base64Unescape(StringPiece src, string* dest);
801 :
802 : // ----------------------------------------------------------------------
803 : // WebSafeBase64Unescape()
804 : // This is a variation of Base64Unescape which uses '-' instead of '+', and
805 : // '_' instead of '/'. src is not null terminated, instead specify len. I
806 : // recommend that slen<szdest, but we honor szdest anyway.
807 : // RETURNS the length of dest, or -1 if src contains invalid chars.
808 :
809 : // The variation that stores into a string clears the string first, and
810 : // returns false (with dest empty) if src contains invalid chars; for
811 : // this version src and dest must be different strings.
812 : // ----------------------------------------------------------------------
813 : LIBPROTOBUF_EXPORT int WebSafeBase64Unescape(const char* src, int slen,
814 : char* dest, int szdest);
815 : LIBPROTOBUF_EXPORT bool WebSafeBase64Unescape(StringPiece src, string* dest);
816 :
817 : // Return the length to use for the output buffer given to the base64 escape
818 : // routines. Make sure to use the same value for do_padding in both.
819 : // This function may return incorrect results if given input_len values that
820 : // are extremely high, which should happen rarely.
821 : LIBPROTOBUF_EXPORT int CalculateBase64EscapedLen(int input_len,
822 : bool do_padding);
823 : // Use this version when calling Base64Escape without a do_padding arg.
824 : LIBPROTOBUF_EXPORT int CalculateBase64EscapedLen(int input_len);
825 :
826 : // ----------------------------------------------------------------------
827 : // Base64Escape()
828 : // WebSafeBase64Escape()
829 : // Encode "src" to "dest" using base64 encoding.
830 : // src is not null terminated, instead specify len.
831 : // 'dest' should have at least CalculateBase64EscapedLen() length.
832 : // RETURNS the length of dest.
833 : // The WebSafe variation use '-' instead of '+' and '_' instead of '/'
834 : // so that we can place the out in the URL or cookies without having
835 : // to escape them. It also has an extra parameter "do_padding",
836 : // which when set to false will prevent padding with "=".
837 : // ----------------------------------------------------------------------
838 : LIBPROTOBUF_EXPORT int Base64Escape(const unsigned char* src, int slen,
839 : char* dest, int szdest);
840 : LIBPROTOBUF_EXPORT int WebSafeBase64Escape(
841 : const unsigned char* src, int slen, char* dest,
842 : int szdest, bool do_padding);
843 : // Encode src into dest with padding.
844 : LIBPROTOBUF_EXPORT void Base64Escape(StringPiece src, string* dest);
845 : // Encode src into dest web-safely without padding.
846 : LIBPROTOBUF_EXPORT void WebSafeBase64Escape(StringPiece src, string* dest);
847 : // Encode src into dest web-safely with padding.
848 : LIBPROTOBUF_EXPORT void WebSafeBase64EscapeWithPadding(StringPiece src,
849 : string* dest);
850 :
851 : LIBPROTOBUF_EXPORT void Base64Escape(const unsigned char* src, int szsrc,
852 : string* dest, bool do_padding);
853 : LIBPROTOBUF_EXPORT void WebSafeBase64Escape(const unsigned char* src, int szsrc,
854 : string* dest, bool do_padding);
855 :
856 : static const int UTFmax = 4;
857 : // ----------------------------------------------------------------------
858 : // EncodeAsUTF8Char()
859 : // Helper to append a Unicode code point to a string as UTF8, without bringing
860 : // in any external dependencies. The output buffer must be as least 4 bytes
861 : // large.
862 : // ----------------------------------------------------------------------
863 : LIBPROTOBUF_EXPORT int EncodeAsUTF8Char(uint32 code_point, char* output);
864 :
865 : // ----------------------------------------------------------------------
866 : // UTF8FirstLetterNumBytes()
867 : // Length of the first UTF-8 character.
868 : // ----------------------------------------------------------------------
869 : LIBPROTOBUF_EXPORT int UTF8FirstLetterNumBytes(const char* src, int len);
870 :
871 : } // namespace protobuf
872 : } // namespace google
873 :
874 : #endif // GOOGLE_PROTOBUF_STUBS_STRUTIL_H__
|