• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1 /*
2  *  Copyright 2004 The WebRTC Project Authors. All rights reserved.
3  *
4  *  Use of this source code is governed by a BSD-style license
5  *  that can be found in the LICENSE file in the root of the source
6  *  tree. An additional intellectual property rights grant can be found
7  *  in the file PATENTS.  All contributing project authors may
8  *  be found in the AUTHORS file in the root of the source tree.
9  */
10 
11 #include "webrtc/base/stringencode.h"
12 
13 #include <stdio.h>
14 #include <stdlib.h>
15 
16 #include "webrtc/base/basictypes.h"
17 #include "webrtc/base/common.h"
18 #include "webrtc/base/stringutils.h"
19 
20 namespace rtc {
21 
22 /////////////////////////////////////////////////////////////////////////////
23 // String Encoding Utilities
24 /////////////////////////////////////////////////////////////////////////////
25 
escape(char * buffer,size_t buflen,const char * source,size_t srclen,const char * illegal,char escape)26 size_t escape(char * buffer, size_t buflen,
27               const char * source, size_t srclen,
28               const char * illegal, char escape) {
29   ASSERT(NULL != buffer);  // TODO: estimate output size
30   if (buflen <= 0)
31     return 0;
32 
33   size_t srcpos = 0, bufpos = 0;
34   while ((srcpos < srclen) && (bufpos + 1 < buflen)) {
35     char ch = source[srcpos++];
36     if ((ch == escape) || ::strchr(illegal, ch)) {
37       if (bufpos + 2 >= buflen)
38         break;
39       buffer[bufpos++] = escape;
40     }
41     buffer[bufpos++] = ch;
42   }
43 
44   buffer[bufpos] = '\0';
45   return bufpos;
46 }
47 
unescape(char * buffer,size_t buflen,const char * source,size_t srclen,char escape)48 size_t unescape(char * buffer, size_t buflen,
49                 const char * source, size_t srclen,
50                 char escape) {
51   ASSERT(NULL != buffer);  // TODO: estimate output size
52   if (buflen <= 0)
53     return 0;
54 
55   size_t srcpos = 0, bufpos = 0;
56   while ((srcpos < srclen) && (bufpos + 1 < buflen)) {
57     char ch = source[srcpos++];
58     if ((ch == escape) && (srcpos < srclen)) {
59       ch = source[srcpos++];
60     }
61     buffer[bufpos++] = ch;
62   }
63   buffer[bufpos] = '\0';
64   return bufpos;
65 }
66 
encode(char * buffer,size_t buflen,const char * source,size_t srclen,const char * illegal,char escape)67 size_t encode(char * buffer, size_t buflen,
68               const char * source, size_t srclen,
69               const char * illegal, char escape) {
70   ASSERT(NULL != buffer);  // TODO: estimate output size
71   if (buflen <= 0)
72     return 0;
73 
74   size_t srcpos = 0, bufpos = 0;
75   while ((srcpos < srclen) && (bufpos + 1 < buflen)) {
76     char ch = source[srcpos++];
77     if ((ch != escape) && !::strchr(illegal, ch)) {
78       buffer[bufpos++] = ch;
79     } else if (bufpos + 3 >= buflen) {
80       break;
81     } else {
82       buffer[bufpos+0] = escape;
83       buffer[bufpos+1] = hex_encode((static_cast<unsigned char>(ch) >> 4) & 0xF);
84       buffer[bufpos+2] = hex_encode((static_cast<unsigned char>(ch)     ) & 0xF);
85       bufpos += 3;
86     }
87   }
88   buffer[bufpos] = '\0';
89   return bufpos;
90 }
91 
decode(char * buffer,size_t buflen,const char * source,size_t srclen,char escape)92 size_t decode(char * buffer, size_t buflen,
93               const char * source, size_t srclen,
94               char escape) {
95   if (buflen <= 0)
96     return 0;
97 
98   unsigned char h1, h2;
99   size_t srcpos = 0, bufpos = 0;
100   while ((srcpos < srclen) && (bufpos + 1 < buflen)) {
101     char ch = source[srcpos++];
102     if ((ch == escape)
103         && (srcpos + 1 < srclen)
104         && hex_decode(source[srcpos], &h1)
105         && hex_decode(source[srcpos+1], &h2)) {
106       buffer[bufpos++] = (h1 << 4) | h2;
107       srcpos += 2;
108     } else {
109       buffer[bufpos++] = ch;
110     }
111   }
112   buffer[bufpos] = '\0';
113   return bufpos;
114 }
115 
unsafe_filename_characters()116 const char* unsafe_filename_characters() {
117   // It might be better to have a single specification which is the union of
118   // all operating systems, unless one system is overly restrictive.
119 #if defined(WEBRTC_WIN)
120   return "\\/:*?\"<>|";
121 #else  // !WEBRTC_WIN
122   // TODO
123   ASSERT(false);
124   return "";
125 #endif  // !WEBRTC_WIN
126 }
127 
128 const unsigned char URL_UNSAFE  = 0x1; // 0-33 "#$%&+,/:;<=>?@[\]^`{|} 127
129 const unsigned char XML_UNSAFE  = 0x2; // "&'<>
130 const unsigned char HTML_UNSAFE = 0x2; // "&'<>
131 
132 //  ! " # $ % & ' ( ) * + , - . / 0 1 2 3 4 6 5 7 8 9 : ; < = > ?
133 //@ A B C D E F G H I J K L M N O P Q R S T U V W X Y Z [ \ ] ^ _
134 //` a b c d e f g h i j k l m n o p q r s t u v w x y z { | } ~
135 
136 const unsigned char ASCII_CLASS[128] = {
137   1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
138   1,0,3,1,1,1,3,2,0,0,0,1,1,0,0,1,0,0,0,0,0,0,0,0,0,0,1,1,3,1,3,1,
139   1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,1,1,1,0,
140   1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,1,1,0,1,
141 };
142 
url_encode(char * buffer,size_t buflen,const char * source,size_t srclen)143 size_t url_encode(char * buffer, size_t buflen,
144                   const char * source, size_t srclen) {
145   if (NULL == buffer)
146     return srclen * 3 + 1;
147   if (buflen <= 0)
148     return 0;
149 
150   size_t srcpos = 0, bufpos = 0;
151   while ((srcpos < srclen) && (bufpos + 1 < buflen)) {
152     unsigned char ch = source[srcpos++];
153     if ((ch < 128) && (ASCII_CLASS[ch] & URL_UNSAFE)) {
154       if (bufpos + 3 >= buflen) {
155         break;
156       }
157       buffer[bufpos+0] = '%';
158       buffer[bufpos+1] = hex_encode((ch >> 4) & 0xF);
159       buffer[bufpos+2] = hex_encode((ch     ) & 0xF);
160       bufpos += 3;
161     } else {
162       buffer[bufpos++] = ch;
163     }
164   }
165   buffer[bufpos] = '\0';
166   return bufpos;
167 }
168 
url_decode(char * buffer,size_t buflen,const char * source,size_t srclen)169 size_t url_decode(char * buffer, size_t buflen,
170                   const char * source, size_t srclen) {
171   if (NULL == buffer)
172     return srclen + 1;
173   if (buflen <= 0)
174     return 0;
175 
176   unsigned char h1, h2;
177   size_t srcpos = 0, bufpos = 0;
178   while ((srcpos < srclen) && (bufpos + 1 < buflen)) {
179     unsigned char ch = source[srcpos++];
180     if (ch == '+') {
181       buffer[bufpos++] = ' ';
182     } else if ((ch == '%')
183                && (srcpos + 1 < srclen)
184                && hex_decode(source[srcpos], &h1)
185                && hex_decode(source[srcpos+1], &h2))
186     {
187       buffer[bufpos++] = (h1 << 4) | h2;
188       srcpos += 2;
189     } else {
190       buffer[bufpos++] = ch;
191     }
192   }
193   buffer[bufpos] = '\0';
194   return bufpos;
195 }
196 
utf8_decode(const char * source,size_t srclen,unsigned long * value)197 size_t utf8_decode(const char* source, size_t srclen, unsigned long* value) {
198   const unsigned char* s = reinterpret_cast<const unsigned char*>(source);
199   if ((s[0] & 0x80) == 0x00) {                    // Check s[0] == 0xxxxxxx
200     *value = s[0];
201     return 1;
202   }
203   if ((srclen < 2) || ((s[1] & 0xC0) != 0x80)) {  // Check s[1] != 10xxxxxx
204     return 0;
205   }
206   // Accumulate the trailer byte values in value16, and combine it with the
207   // relevant bits from s[0], once we've determined the sequence length.
208   unsigned long value16 = (s[1] & 0x3F);
209   if ((s[0] & 0xE0) == 0xC0) {                    // Check s[0] == 110xxxxx
210     *value = ((s[0] & 0x1F) << 6) | value16;
211     return 2;
212   }
213   if ((srclen < 3) || ((s[2] & 0xC0) != 0x80)) {  // Check s[2] != 10xxxxxx
214     return 0;
215   }
216   value16 = (value16 << 6) | (s[2] & 0x3F);
217   if ((s[0] & 0xF0) == 0xE0) {                    // Check s[0] == 1110xxxx
218     *value = ((s[0] & 0x0F) << 12) | value16;
219     return 3;
220   }
221   if ((srclen < 4) || ((s[3] & 0xC0) != 0x80)) {  // Check s[3] != 10xxxxxx
222     return 0;
223   }
224   value16 = (value16 << 6) | (s[3] & 0x3F);
225   if ((s[0] & 0xF8) == 0xF0) {                    // Check s[0] == 11110xxx
226     *value = ((s[0] & 0x07) << 18) | value16;
227     return 4;
228   }
229   return 0;
230 }
231 
utf8_encode(char * buffer,size_t buflen,unsigned long value)232 size_t utf8_encode(char* buffer, size_t buflen, unsigned long value) {
233   if ((value <= 0x7F) && (buflen >= 1)) {
234     buffer[0] = static_cast<unsigned char>(value);
235     return 1;
236   }
237   if ((value <= 0x7FF) && (buflen >= 2)) {
238     buffer[0] = 0xC0 | static_cast<unsigned char>(value >> 6);
239     buffer[1] = 0x80 | static_cast<unsigned char>(value & 0x3F);
240     return 2;
241   }
242   if ((value <= 0xFFFF) && (buflen >= 3)) {
243     buffer[0] = 0xE0 | static_cast<unsigned char>(value >> 12);
244     buffer[1] = 0x80 | static_cast<unsigned char>((value >> 6) & 0x3F);
245     buffer[2] = 0x80 | static_cast<unsigned char>(value & 0x3F);
246     return 3;
247   }
248   if ((value <= 0x1FFFFF) && (buflen >= 4)) {
249     buffer[0] = 0xF0 | static_cast<unsigned char>(value >> 18);
250     buffer[1] = 0x80 | static_cast<unsigned char>((value >> 12) & 0x3F);
251     buffer[2] = 0x80 | static_cast<unsigned char>((value >> 6) & 0x3F);
252     buffer[3] = 0x80 | static_cast<unsigned char>(value & 0x3F);
253     return 4;
254   }
255   return 0;
256 }
257 
html_encode(char * buffer,size_t buflen,const char * source,size_t srclen)258 size_t html_encode(char * buffer, size_t buflen,
259                    const char * source, size_t srclen) {
260   ASSERT(NULL != buffer);  // TODO: estimate output size
261   if (buflen <= 0)
262     return 0;
263 
264   size_t srcpos = 0, bufpos = 0;
265   while ((srcpos < srclen) && (bufpos + 1 < buflen)) {
266     unsigned char ch = source[srcpos];
267     if (ch < 128) {
268       srcpos += 1;
269       if (ASCII_CLASS[ch] & HTML_UNSAFE) {
270         const char * escseq = 0;
271         size_t esclen = 0;
272         switch (ch) {
273           case '<':  escseq = "&lt;";   esclen = 4; break;
274           case '>':  escseq = "&gt;";   esclen = 4; break;
275           case '\'': escseq = "&#39;";  esclen = 5; break;
276           case '\"': escseq = "&quot;"; esclen = 6; break;
277           case '&':  escseq = "&amp;";  esclen = 5; break;
278           default: ASSERT(false);
279         }
280         if (bufpos + esclen >= buflen) {
281           break;
282         }
283         memcpy(buffer + bufpos, escseq, esclen);
284         bufpos += esclen;
285       } else {
286         buffer[bufpos++] = ch;
287       }
288     } else {
289       // Largest value is 0x1FFFFF => &#2097151;  (10 characters)
290       char escseq[11];
291       unsigned long val;
292       if (size_t vallen = utf8_decode(&source[srcpos], srclen - srcpos, &val)) {
293         srcpos += vallen;
294       } else {
295         // Not a valid utf8 sequence, just use the raw character.
296         val = static_cast<unsigned char>(source[srcpos++]);
297       }
298       size_t esclen = sprintfn(escseq, ARRAY_SIZE(escseq), "&#%lu;", val);
299       if (bufpos + esclen >= buflen) {
300         break;
301       }
302       memcpy(buffer + bufpos, escseq, esclen);
303       bufpos += esclen;
304     }
305   }
306   buffer[bufpos] = '\0';
307   return bufpos;
308 }
309 
html_decode(char * buffer,size_t buflen,const char * source,size_t srclen)310 size_t html_decode(char * buffer, size_t buflen,
311                    const char * source, size_t srclen) {
312   ASSERT(NULL != buffer);  // TODO: estimate output size
313   return xml_decode(buffer, buflen, source, srclen);
314 }
315 
xml_encode(char * buffer,size_t buflen,const char * source,size_t srclen)316 size_t xml_encode(char * buffer, size_t buflen,
317                   const char * source, size_t srclen) {
318   ASSERT(NULL != buffer);  // TODO: estimate output size
319   if (buflen <= 0)
320     return 0;
321 
322   size_t srcpos = 0, bufpos = 0;
323   while ((srcpos < srclen) && (bufpos + 1 < buflen)) {
324     unsigned char ch = source[srcpos++];
325     if ((ch < 128) && (ASCII_CLASS[ch] & XML_UNSAFE)) {
326       const char * escseq = 0;
327       size_t esclen = 0;
328       switch (ch) {
329         case '<':  escseq = "&lt;";   esclen = 4; break;
330         case '>':  escseq = "&gt;";   esclen = 4; break;
331         case '\'': escseq = "&apos;"; esclen = 6; break;
332         case '\"': escseq = "&quot;"; esclen = 6; break;
333         case '&':  escseq = "&amp;";  esclen = 5; break;
334         default: ASSERT(false);
335       }
336       if (bufpos + esclen >= buflen) {
337         break;
338       }
339       memcpy(buffer + bufpos, escseq, esclen);
340       bufpos += esclen;
341     } else {
342       buffer[bufpos++] = ch;
343     }
344   }
345   buffer[bufpos] = '\0';
346   return bufpos;
347 }
348 
xml_decode(char * buffer,size_t buflen,const char * source,size_t srclen)349 size_t xml_decode(char * buffer, size_t buflen,
350                   const char * source, size_t srclen) {
351   ASSERT(NULL != buffer);  // TODO: estimate output size
352   if (buflen <= 0)
353     return 0;
354 
355   size_t srcpos = 0, bufpos = 0;
356   while ((srcpos < srclen) && (bufpos + 1 < buflen)) {
357     unsigned char ch = source[srcpos++];
358     if (ch != '&') {
359       buffer[bufpos++] = ch;
360     } else if ((srcpos + 2 < srclen)
361                && (memcmp(source + srcpos, "lt;", 3) == 0)) {
362       buffer[bufpos++] = '<';
363       srcpos += 3;
364     } else if ((srcpos + 2 < srclen)
365                && (memcmp(source + srcpos, "gt;", 3) == 0)) {
366       buffer[bufpos++] = '>';
367       srcpos += 3;
368     } else if ((srcpos + 4 < srclen)
369                && (memcmp(source + srcpos, "apos;", 5) == 0)) {
370       buffer[bufpos++] = '\'';
371       srcpos += 5;
372     } else if ((srcpos + 4 < srclen)
373                && (memcmp(source + srcpos, "quot;", 5) == 0)) {
374       buffer[bufpos++] = '\"';
375       srcpos += 5;
376     } else if ((srcpos + 3 < srclen)
377                && (memcmp(source + srcpos, "amp;", 4) == 0)) {
378       buffer[bufpos++] = '&';
379       srcpos += 4;
380     } else if ((srcpos < srclen) && (source[srcpos] == '#')) {
381       int int_base = 10;
382       if ((srcpos + 1 < srclen) && (source[srcpos+1] == 'x')) {
383         int_base = 16;
384         srcpos += 1;
385       }
386       char * ptr;
387       // TODO: Fix hack (ptr may go past end of data)
388       unsigned long val = strtoul(source + srcpos + 1, &ptr, int_base);
389       if ((static_cast<size_t>(ptr - source) < srclen) && (*ptr == ';')) {
390         srcpos = ptr - source + 1;
391       } else {
392         // Not a valid escape sequence.
393         break;
394       }
395       if (size_t esclen = utf8_encode(buffer + bufpos, buflen - bufpos, val)) {
396         bufpos += esclen;
397       } else {
398         // Not enough room to encode the character, or illegal character
399         break;
400       }
401     } else {
402       // Unrecognized escape sequence.
403       break;
404     }
405   }
406   buffer[bufpos] = '\0';
407   return bufpos;
408 }
409 
410 static const char HEX[] = "0123456789abcdef";
411 
hex_encode(unsigned char val)412 char hex_encode(unsigned char val) {
413   ASSERT(val < 16);
414   return (val < 16) ? HEX[val] : '!';
415 }
416 
hex_decode(char ch,unsigned char * val)417 bool hex_decode(char ch, unsigned char* val) {
418   if ((ch >= '0') && (ch <= '9')) {
419     *val = ch - '0';
420   } else if ((ch >= 'A') && (ch <= 'Z')) {
421     *val = (ch - 'A') + 10;
422   } else if ((ch >= 'a') && (ch <= 'z')) {
423     *val = (ch - 'a') + 10;
424   } else {
425     return false;
426   }
427   return true;
428 }
429 
hex_encode(char * buffer,size_t buflen,const char * csource,size_t srclen)430 size_t hex_encode(char* buffer, size_t buflen,
431                   const char* csource, size_t srclen) {
432   return hex_encode_with_delimiter(buffer, buflen, csource, srclen, 0);
433 }
434 
hex_encode_with_delimiter(char * buffer,size_t buflen,const char * csource,size_t srclen,char delimiter)435 size_t hex_encode_with_delimiter(char* buffer, size_t buflen,
436                                  const char* csource, size_t srclen,
437                                  char delimiter) {
438   ASSERT(NULL != buffer);  // TODO: estimate output size
439   if (buflen == 0)
440     return 0;
441 
442   // Init and check bounds.
443   const unsigned char* bsource =
444       reinterpret_cast<const unsigned char*>(csource);
445   size_t srcpos = 0, bufpos = 0;
446   size_t needed = delimiter ? (srclen * 3) : (srclen * 2 + 1);
447   if (buflen < needed)
448     return 0;
449 
450   while (srcpos < srclen) {
451     unsigned char ch = bsource[srcpos++];
452     buffer[bufpos  ] = hex_encode((ch >> 4) & 0xF);
453     buffer[bufpos+1] = hex_encode((ch     ) & 0xF);
454     bufpos += 2;
455 
456     // Don't write a delimiter after the last byte.
457     if (delimiter && (srcpos < srclen)) {
458       buffer[bufpos] = delimiter;
459       ++bufpos;
460     }
461   }
462 
463   // Null terminate.
464   buffer[bufpos] = '\0';
465   return bufpos;
466 }
467 
hex_encode(const char * source,size_t srclen)468 std::string hex_encode(const char* source, size_t srclen) {
469   return hex_encode_with_delimiter(source, srclen, 0);
470 }
471 
hex_encode_with_delimiter(const char * source,size_t srclen,char delimiter)472 std::string hex_encode_with_delimiter(const char* source, size_t srclen,
473                                       char delimiter) {
474   const size_t kBufferSize = srclen * 3;
475   char* buffer = STACK_ARRAY(char, kBufferSize);
476   size_t length = hex_encode_with_delimiter(buffer, kBufferSize,
477                                             source, srclen, delimiter);
478   ASSERT(srclen == 0 || length > 0);
479   return std::string(buffer, length);
480 }
481 
hex_decode(char * cbuffer,size_t buflen,const char * source,size_t srclen)482 size_t hex_decode(char * cbuffer, size_t buflen,
483                   const char * source, size_t srclen) {
484   return hex_decode_with_delimiter(cbuffer, buflen, source, srclen, 0);
485 }
486 
hex_decode_with_delimiter(char * cbuffer,size_t buflen,const char * source,size_t srclen,char delimiter)487 size_t hex_decode_with_delimiter(char* cbuffer, size_t buflen,
488                                  const char* source, size_t srclen,
489                                  char delimiter) {
490   ASSERT(NULL != cbuffer);  // TODO: estimate output size
491   if (buflen == 0)
492     return 0;
493 
494   // Init and bounds check.
495   unsigned char* bbuffer = reinterpret_cast<unsigned char*>(cbuffer);
496   size_t srcpos = 0, bufpos = 0;
497   size_t needed = (delimiter) ? (srclen + 1) / 3 : srclen / 2;
498   if (buflen < needed)
499     return 0;
500 
501   while (srcpos < srclen) {
502     if ((srclen - srcpos) < 2) {
503       // This means we have an odd number of bytes.
504       return 0;
505     }
506 
507     unsigned char h1, h2;
508     if (!hex_decode(source[srcpos], &h1) ||
509         !hex_decode(source[srcpos + 1], &h2))
510       return 0;
511 
512     bbuffer[bufpos++] = (h1 << 4) | h2;
513     srcpos += 2;
514 
515     // Remove the delimiter if needed.
516     if (delimiter && (srclen - srcpos) > 1) {
517       if (source[srcpos] != delimiter)
518         return 0;
519       ++srcpos;
520     }
521   }
522 
523   return bufpos;
524 }
525 
hex_decode(char * buffer,size_t buflen,const std::string & source)526 size_t hex_decode(char* buffer, size_t buflen, const std::string& source) {
527   return hex_decode_with_delimiter(buffer, buflen, source, 0);
528 }
hex_decode_with_delimiter(char * buffer,size_t buflen,const std::string & source,char delimiter)529 size_t hex_decode_with_delimiter(char* buffer, size_t buflen,
530                                  const std::string& source, char delimiter) {
531   return hex_decode_with_delimiter(buffer, buflen,
532                                    source.c_str(), source.length(), delimiter);
533 }
534 
transform(std::string & value,size_t maxlen,const std::string & source,Transform t)535 size_t transform(std::string& value, size_t maxlen, const std::string& source,
536                  Transform t) {
537   char* buffer = STACK_ARRAY(char, maxlen + 1);
538   size_t length = t(buffer, maxlen + 1, source.data(), source.length());
539   value.assign(buffer, length);
540   return length;
541 }
542 
s_transform(const std::string & source,Transform t)543 std::string s_transform(const std::string& source, Transform t) {
544   // Ask transformation function to approximate the destination size (returns upper bound)
545   size_t maxlen = t(NULL, 0, source.data(), source.length());
546   char * buffer = STACK_ARRAY(char, maxlen);
547   size_t len = t(buffer, maxlen, source.data(), source.length());
548   std::string result(buffer, len);
549   return result;
550 }
551 
tokenize(const std::string & source,char delimiter,std::vector<std::string> * fields)552 size_t tokenize(const std::string& source, char delimiter,
553                 std::vector<std::string>* fields) {
554   ASSERT(NULL != fields);
555   fields->clear();
556   size_t last = 0;
557   for (size_t i = 0; i < source.length(); ++i) {
558     if (source[i] == delimiter) {
559       if (i != last) {
560         fields->push_back(source.substr(last, i - last));
561       }
562       last = i + 1;
563     }
564   }
565   if (last != source.length()) {
566     fields->push_back(source.substr(last, source.length() - last));
567   }
568   return fields->size();
569 }
570 
tokenize_append(const std::string & source,char delimiter,std::vector<std::string> * fields)571 size_t tokenize_append(const std::string& source, char delimiter,
572                        std::vector<std::string>* fields) {
573   if (!fields) return 0;
574 
575   std::vector<std::string> new_fields;
576   tokenize(source, delimiter, &new_fields);
577   fields->insert(fields->end(), new_fields.begin(), new_fields.end());
578   return fields->size();
579 }
580 
tokenize(const std::string & source,char delimiter,char start_mark,char end_mark,std::vector<std::string> * fields)581 size_t tokenize(const std::string& source, char delimiter, char start_mark,
582                 char end_mark, std::vector<std::string>* fields) {
583   if (!fields) return 0;
584   fields->clear();
585 
586   std::string remain_source = source;
587   while (!remain_source.empty()) {
588     size_t start_pos = remain_source.find(start_mark);
589     if (std::string::npos == start_pos) break;
590     std::string pre_mark;
591     if (start_pos > 0) {
592       pre_mark = remain_source.substr(0, start_pos - 1);
593     }
594 
595     ++start_pos;
596     size_t end_pos = remain_source.find(end_mark, start_pos);
597     if (std::string::npos == end_pos) break;
598 
599     // We have found the matching marks. First tokenize the pre-mask. Then add
600     // the marked part as a single field. Finally, loop back for the post-mark.
601     tokenize_append(pre_mark, delimiter, fields);
602     fields->push_back(remain_source.substr(start_pos, end_pos - start_pos));
603     remain_source = remain_source.substr(end_pos + 1);
604   }
605 
606   return tokenize_append(remain_source, delimiter, fields);
607 }
608 
split(const std::string & source,char delimiter,std::vector<std::string> * fields)609 size_t split(const std::string& source, char delimiter,
610              std::vector<std::string>* fields) {
611   ASSERT(NULL != fields);
612   fields->clear();
613   size_t last = 0;
614   for (size_t i = 0; i < source.length(); ++i) {
615     if (source[i] == delimiter) {
616       fields->push_back(source.substr(last, i - last));
617       last = i + 1;
618     }
619   }
620   fields->push_back(source.substr(last, source.length() - last));
621   return fields->size();
622 }
623 
make_char_safe_for_filename(char c)624 char make_char_safe_for_filename(char c) {
625   if (c < 32)
626     return '_';
627 
628   switch (c) {
629     case '<':
630     case '>':
631     case ':':
632     case '"':
633     case '/':
634     case '\\':
635     case '|':
636     case '*':
637     case '?':
638       return '_';
639 
640     default:
641       return c;
642   }
643 }
644 
645 /*
646 void sprintf(std::string& value, size_t maxlen, const char * format, ...) {
647   char * buffer = STACK_ARRAY(char, maxlen + 1);
648   va_list args;
649   va_start(args, format);
650   value.assign(buffer, vsprintfn(buffer, maxlen + 1, format, args));
651   va_end(args);
652 }
653 */
654 
655 /////////////////////////////////////////////////////////////////////////////
656 
657 }  // namespace rtc
658