3
* Copyright 2004--2011, Google Inc.
5
* Redistribution and use in source and binary forms, with or without
6
* modification, are permitted provided that the following conditions are met:
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.
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.
28
#include "talk/base/stringencode.h"
33
#include "talk/base/basictypes.h"
34
#include "talk/base/common.h"
35
#include "talk/base/stringutils.h"
39
/////////////////////////////////////////////////////////////////////////////
40
// String Encoding Utilities
41
/////////////////////////////////////////////////////////////////////////////
43
static const char HEX[] = "0123456789abcdef";
45
char hex_encode(unsigned char val) {
47
return (val < 16) ? HEX[val] : '!';
50
bool hex_decode(char ch, unsigned char* val) {
51
if ((ch >= '0') && (ch <= '9')) {
53
} else if ((ch >= 'A') && (ch <= 'Z')) {
54
*val = (ch - 'A') + 10;
55
} else if ((ch >= 'a') && (ch <= 'z')) {
56
*val = (ch - 'a') + 10;
63
size_t escape(char * buffer, size_t buflen,
64
const char * source, size_t srclen,
65
const char * illegal, char escape) {
66
ASSERT(NULL != buffer); // TODO: estimate output size
70
size_t srcpos = 0, bufpos = 0;
71
while ((srcpos < srclen) && (bufpos + 1 < buflen)) {
72
char ch = source[srcpos++];
73
if ((ch == escape) || ::strchr(illegal, ch)) {
74
if (bufpos + 2 >= buflen)
76
buffer[bufpos++] = escape;
78
buffer[bufpos++] = ch;
81
buffer[bufpos] = '\0';
85
size_t unescape(char * buffer, size_t buflen,
86
const char * source, size_t srclen,
88
ASSERT(NULL != buffer); // TODO: estimate output size
92
size_t srcpos = 0, bufpos = 0;
93
while ((srcpos < srclen) && (bufpos + 1 < buflen)) {
94
char ch = source[srcpos++];
95
if ((ch == escape) && (srcpos < srclen)) {
96
ch = source[srcpos++];
98
buffer[bufpos++] = ch;
100
buffer[bufpos] = '\0';
104
size_t encode(char * buffer, size_t buflen,
105
const char * source, size_t srclen,
106
const char * illegal, char escape) {
107
ASSERT(NULL != buffer); // TODO: estimate output size
111
size_t srcpos = 0, bufpos = 0;
112
while ((srcpos < srclen) && (bufpos + 1 < buflen)) {
113
char ch = source[srcpos++];
114
if ((ch != escape) && !::strchr(illegal, ch)) {
115
buffer[bufpos++] = ch;
116
} else if (bufpos + 3 >= buflen) {
119
buffer[bufpos+0] = escape;
120
buffer[bufpos+1] = hex_encode((static_cast<unsigned char>(ch) >> 4) & 0xF);
121
buffer[bufpos+2] = hex_encode((static_cast<unsigned char>(ch) ) & 0xF);
125
buffer[bufpos] = '\0';
129
size_t decode(char * buffer, size_t buflen,
130
const char * source, size_t srclen,
135
unsigned char h1, h2;
136
size_t srcpos = 0, bufpos = 0;
137
while ((srcpos < srclen) && (bufpos + 1 < buflen)) {
138
char ch = source[srcpos++];
140
&& (srcpos + 1 < srclen)
141
&& hex_decode(source[srcpos], &h1)
142
&& hex_decode(source[srcpos+1], &h2)) {
143
buffer[bufpos++] = (h1 << 4) | h2;
146
buffer[bufpos++] = ch;
149
buffer[bufpos] = '\0';
153
const char* unsafe_filename_characters() {
154
// It might be better to have a single specification which is the union of
155
// all operating systems, unless one system is overly restrictive.
157
return "\\/:*?\"<>|";
165
const unsigned char URL_UNSAFE = 0x1; // 0-33 "#$%&+,/:;<=>?@[\]^`{|} 127
166
const unsigned char XML_UNSAFE = 0x2; // "&'<>
167
const unsigned char HTML_UNSAFE = 0x2; // "&'<>
169
// ! " # $ % & ' ( ) * + , - . / 0 1 2 3 4 6 5 7 8 9 : ; < = > ?
170
//@ 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 [ \ ] ^ _
171
//` 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 { | } ~
173
const unsigned char ASCII_CLASS[128] = {
174
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,
175
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,
176
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,
177
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,
180
size_t url_encode(char * buffer, size_t buflen,
181
const char * source, size_t srclen) {
183
return srclen * 3 + 1;
187
size_t srcpos = 0, bufpos = 0;
188
while ((srcpos < srclen) && (bufpos + 1 < buflen)) {
189
unsigned char ch = source[srcpos++];
190
if ((ch < 128) && (ASCII_CLASS[ch] & URL_UNSAFE)) {
191
if (bufpos + 3 >= buflen) {
194
buffer[bufpos+0] = '%';
195
buffer[bufpos+1] = hex_encode((ch >> 4) & 0xF);
196
buffer[bufpos+2] = hex_encode((ch ) & 0xF);
199
buffer[bufpos++] = ch;
202
buffer[bufpos] = '\0';
206
size_t url_decode(char * buffer, size_t buflen,
207
const char * source, size_t srclen) {
213
unsigned char h1, h2;
214
size_t srcpos = 0, bufpos = 0;
215
while ((srcpos < srclen) && (bufpos + 1 < buflen)) {
216
unsigned char ch = source[srcpos++];
218
buffer[bufpos++] = ' ';
219
} else if ((ch == '%')
220
&& (srcpos + 1 < srclen)
221
&& hex_decode(source[srcpos], &h1)
222
&& hex_decode(source[srcpos+1], &h2))
224
buffer[bufpos++] = (h1 << 4) | h2;
227
buffer[bufpos++] = ch;
230
buffer[bufpos] = '\0';
234
size_t utf8_decode(const char* source, size_t srclen, unsigned long* value) {
235
const unsigned char* s = reinterpret_cast<const unsigned char*>(source);
236
if ((s[0] & 0x80) == 0x00) { // Check s[0] == 0xxxxxxx
240
if ((srclen < 2) || ((s[1] & 0xC0) != 0x80)) { // Check s[1] != 10xxxxxx
243
// Accumulate the trailer byte values in value16, and combine it with the
244
// relevant bits from s[0], once we've determined the sequence length.
245
unsigned long value16 = (s[1] & 0x3F);
246
if ((s[0] & 0xE0) == 0xC0) { // Check s[0] == 110xxxxx
247
*value = ((s[0] & 0x1F) << 6) | value16;
250
if ((srclen < 3) || ((s[2] & 0xC0) != 0x80)) { // Check s[2] != 10xxxxxx
253
value16 = (value16 << 6) | (s[2] & 0x3F);
254
if ((s[0] & 0xF0) == 0xE0) { // Check s[0] == 1110xxxx
255
*value = ((s[0] & 0x0F) << 12) | value16;
258
if ((srclen < 4) || ((s[3] & 0xC0) != 0x80)) { // Check s[3] != 10xxxxxx
261
value16 = (value16 << 6) | (s[3] & 0x3F);
262
if ((s[0] & 0xF8) == 0xF0) { // Check s[0] == 11110xxx
263
*value = ((s[0] & 0x07) << 18) | value16;
269
size_t utf8_encode(char* buffer, size_t buflen, unsigned long value) {
270
if ((value <= 0x7F) && (buflen >= 1)) {
271
buffer[0] = static_cast<unsigned char>(value);
274
if ((value <= 0x7FF) && (buflen >= 2)) {
275
buffer[0] = 0xC0 | static_cast<unsigned char>(value >> 6);
276
buffer[1] = 0x80 | static_cast<unsigned char>(value & 0x3F);
279
if ((value <= 0xFFFF) && (buflen >= 3)) {
280
buffer[0] = 0xE0 | static_cast<unsigned char>(value >> 12);
281
buffer[1] = 0x80 | static_cast<unsigned char>((value >> 6) & 0x3F);
282
buffer[2] = 0x80 | static_cast<unsigned char>(value & 0x3F);
285
if ((value <= 0x1FFFFF) && (buflen >= 4)) {
286
buffer[0] = 0xF0 | static_cast<unsigned char>(value >> 18);
287
buffer[1] = 0x80 | static_cast<unsigned char>((value >> 12) & 0x3F);
288
buffer[2] = 0x80 | static_cast<unsigned char>((value >> 6) & 0x3F);
289
buffer[3] = 0x80 | static_cast<unsigned char>(value & 0x3F);
295
size_t html_encode(char * buffer, size_t buflen,
296
const char * source, size_t srclen) {
297
ASSERT(NULL != buffer); // TODO: estimate output size
301
size_t srcpos = 0, bufpos = 0;
302
while ((srcpos < srclen) && (bufpos + 1 < buflen)) {
303
unsigned char ch = source[srcpos];
306
if (ASCII_CLASS[ch] & HTML_UNSAFE) {
307
const char * escseq = 0;
310
case '<': escseq = "<"; esclen = 4; break;
311
case '>': escseq = ">"; esclen = 4; break;
312
case '\'': escseq = "'"; esclen = 5; break;
313
case '\"': escseq = """; esclen = 6; break;
314
case '&': escseq = "&"; esclen = 5; break;
315
default: ASSERT(false);
317
if (bufpos + esclen >= buflen) {
320
memcpy(buffer + bufpos, escseq, esclen);
323
buffer[bufpos++] = ch;
326
// Largest value is 0x1FFFFF => � (10 characters)
329
if (size_t vallen = utf8_decode(&source[srcpos], srclen - srcpos, &val)) {
332
// Not a valid utf8 sequence, just use the raw character.
333
val = static_cast<unsigned char>(source[srcpos++]);
335
size_t esclen = sprintfn(escseq, ARRAY_SIZE(escseq), "&#%lu;", val);
336
if (bufpos + esclen >= buflen) {
339
memcpy(buffer + bufpos, escseq, esclen);
343
buffer[bufpos] = '\0';
347
size_t html_decode(char * buffer, size_t buflen,
348
const char * source, size_t srclen) {
349
ASSERT(NULL != buffer); // TODO: estimate output size
350
return xml_decode(buffer, buflen, source, srclen);
353
size_t xml_encode(char * buffer, size_t buflen,
354
const char * source, size_t srclen) {
355
ASSERT(NULL != buffer); // TODO: estimate output size
359
size_t srcpos = 0, bufpos = 0;
360
while ((srcpos < srclen) && (bufpos + 1 < buflen)) {
361
unsigned char ch = source[srcpos++];
362
if ((ch < 128) && (ASCII_CLASS[ch] & XML_UNSAFE)) {
363
const char * escseq = 0;
366
case '<': escseq = "<"; esclen = 4; break;
367
case '>': escseq = ">"; esclen = 4; break;
368
case '\'': escseq = "'"; esclen = 6; break;
369
case '\"': escseq = """; esclen = 6; break;
370
case '&': escseq = "&"; esclen = 5; break;
371
default: ASSERT(false);
373
if (bufpos + esclen >= buflen) {
376
memcpy(buffer + bufpos, escseq, esclen);
379
buffer[bufpos++] = ch;
382
buffer[bufpos] = '\0';
386
size_t xml_decode(char * buffer, size_t buflen,
387
const char * source, size_t srclen) {
388
ASSERT(NULL != buffer); // TODO: estimate output size
392
size_t srcpos = 0, bufpos = 0;
393
while ((srcpos < srclen) && (bufpos + 1 < buflen)) {
394
unsigned char ch = source[srcpos++];
396
buffer[bufpos++] = ch;
397
} else if ((srcpos + 2 < srclen)
398
&& (memcmp(source + srcpos, "lt;", 3) == 0)) {
399
buffer[bufpos++] = '<';
401
} else if ((srcpos + 2 < srclen)
402
&& (memcmp(source + srcpos, "gt;", 3) == 0)) {
403
buffer[bufpos++] = '>';
405
} else if ((srcpos + 4 < srclen)
406
&& (memcmp(source + srcpos, "apos;", 5) == 0)) {
407
buffer[bufpos++] = '\'';
409
} else if ((srcpos + 4 < srclen)
410
&& (memcmp(source + srcpos, "quot;", 5) == 0)) {
411
buffer[bufpos++] = '\"';
413
} else if ((srcpos + 3 < srclen)
414
&& (memcmp(source + srcpos, "amp;", 4) == 0)) {
415
buffer[bufpos++] = '&';
417
} else if ((srcpos < srclen) && (source[srcpos] == '#')) {
419
if ((srcpos + 1 < srclen) && (source[srcpos+1] == 'x')) {
424
// TODO: Fix hack (ptr may go past end of data)
425
unsigned long val = strtoul(source + srcpos + 1, &ptr, int_base);
426
if ((static_cast<size_t>(ptr - source) < srclen) && (*ptr == ';')) {
427
srcpos = ptr - source + 1;
429
// Not a valid escape sequence.
432
if (size_t esclen = utf8_encode(buffer + bufpos, buflen - bufpos, val)) {
435
// Not enough room to encode the character, or illegal character
439
// Unrecognized escape sequence.
443
buffer[bufpos] = '\0';
447
std::string hex_encode(const char * source, size_t srclen) {
448
const size_t kBufferSize = srclen * 2 + 1;
449
char* buffer = STACK_ARRAY(char, kBufferSize);
450
size_t length = hex_encode(buffer, kBufferSize, source, srclen);
451
return std::string(buffer, length);
454
size_t hex_encode(char * buffer, size_t buflen,
455
const char * csource, size_t srclen) {
456
ASSERT(NULL != buffer); // TODO: estimate output size
460
const unsigned char * bsource =
461
reinterpret_cast<const unsigned char *>(csource);
463
size_t srcpos = 0, bufpos = 0;
464
srclen = _min(srclen, (buflen - 1) / 2);
465
while (srcpos < srclen) {
466
unsigned char ch = bsource[srcpos++];
467
buffer[bufpos ] = hex_encode((ch >> 4) & 0xF);
468
buffer[bufpos+1] = hex_encode((ch ) & 0xF);
471
buffer[bufpos] = '\0';
475
size_t hex_decode(char * cbuffer, size_t buflen,
476
const char * source, size_t srclen) {
477
ASSERT(NULL != cbuffer); // TODO: estimate output size
481
unsigned char * bbuffer = reinterpret_cast<unsigned char *>(cbuffer);
483
unsigned char h1, h2;
484
size_t srcpos = 0, bufpos = 0;
485
while ((srcpos + 1 < srclen)
486
&& (bufpos + 1 < buflen)
487
&& hex_decode(source[srcpos], &h1)
488
&& hex_decode(source[srcpos+1], &h2))
490
bbuffer[bufpos++] = (h1 << 4) | h2;
493
bbuffer[bufpos] = '\0';
497
size_t transform(std::string& value, size_t maxlen, const std::string& source,
499
char* buffer = STACK_ARRAY(char, maxlen + 1);
500
size_t length = t(buffer, maxlen + 1, source.data(), source.length());
501
value.assign(buffer, length);
505
std::string s_transform(const std::string& source, Transform t) {
506
// Ask transformation function to approximate the destination size (returns upper bound)
507
size_t maxlen = t(NULL, 0, source.data(), source.length());
508
char * buffer = STACK_ARRAY(char, maxlen);
509
size_t len = t(buffer, maxlen, source.data(), source.length());
510
std::string result(buffer, len);
514
size_t tokenize(const std::string& source, char delimiter,
515
std::vector<std::string>* fields) {
516
ASSERT(NULL != fields);
519
for (size_t i = 0; i < source.length(); ++i) {
520
if (source[i] == delimiter) {
522
fields->push_back(source.substr(last, i - last));
527
if (last != source.length()) {
528
fields->push_back(source.substr(last, source.length() - last));
530
return fields->size();
533
size_t tokenize_append(const std::string& source, char delimiter,
534
std::vector<std::string>* fields) {
535
if (!fields) return 0;
537
std::vector<std::string> new_fields;
538
tokenize(source, delimiter, &new_fields);
539
fields->insert(fields->end(), new_fields.begin(), new_fields.end());
540
return fields->size();
543
size_t tokenize(const std::string& source, char delimiter, char start_mark,
544
char end_mark, std::vector<std::string>* fields) {
545
if (!fields) return 0;
548
std::string remain_source = source;
549
while (!remain_source.empty()) {
550
size_t start_pos = remain_source.find(start_mark);
551
if (std::string::npos == start_pos) break;
552
std::string pre_mark;
554
pre_mark = remain_source.substr(0, start_pos - 1);
558
size_t end_pos = remain_source.find(end_mark, start_pos);
559
if (std::string::npos == end_pos) break;
561
// We have found the matching marks. First tokenize the pre-mask. Then add
562
// the marked part as a single field. Finally, loop back for the post-mark.
563
tokenize_append(pre_mark, delimiter, fields);
564
fields->push_back(remain_source.substr(start_pos, end_pos - start_pos));
565
remain_source = remain_source.substr(end_pos + 1);
568
return tokenize_append(remain_source, delimiter, fields);
571
size_t split(const std::string& source, char delimiter,
572
std::vector<std::string>* fields) {
573
ASSERT(NULL != fields);
576
for (size_t i = 0; i < source.length(); ++i) {
577
if (source[i] == delimiter) {
578
fields->push_back(source.substr(last, i - last));
582
fields->push_back(source.substr(last, source.length() - last));
583
return fields->size();
586
char make_char_safe_for_filename(char c) {
608
void sprintf(std::string& value, size_t maxlen, const char * format, ...) {
609
char * buffer = STACK_ARRAY(char, maxlen + 1);
611
va_start(args, format);
612
value.assign(buffer, vsprintfn(buffer, maxlen + 1, format, args));
617
/////////////////////////////////////////////////////////////////////////////
619
} // namespace talk_base