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 size_t escape(char * buffer, size_t buflen,
44 const char * source, size_t srclen,
45 const char * illegal, char escape) {
46 ASSERT(NULL != buffer); // TODO: estimate output size
50 size_t srcpos = 0, bufpos = 0;
51 while ((srcpos < srclen) && (bufpos + 1 < buflen)) {
52 char ch = source[srcpos++];
53 if ((ch == escape) || ::strchr(illegal, ch)) {
54 if (bufpos + 2 >= buflen)
56 buffer[bufpos++] = escape;
58 buffer[bufpos++] = ch;
61 buffer[bufpos] = '\0';
65 size_t unescape(char * buffer, size_t buflen,
66 const char * source, size_t srclen,
68 ASSERT(NULL != buffer); // TODO: estimate output size
72 size_t srcpos = 0, bufpos = 0;
73 while ((srcpos < srclen) && (bufpos + 1 < buflen)) {
74 char ch = source[srcpos++];
75 if ((ch == escape) && (srcpos < srclen)) {
76 ch = source[srcpos++];
78 buffer[bufpos++] = ch;
80 buffer[bufpos] = '\0';
84 size_t encode(char * buffer, size_t buflen,
85 const char * source, size_t srclen,
86 const char * illegal, char escape) {
87 ASSERT(NULL != buffer); // TODO: estimate output size
91 size_t srcpos = 0, bufpos = 0;
92 while ((srcpos < srclen) && (bufpos + 1 < buflen)) {
93 char ch = source[srcpos++];
94 if ((ch != escape) && !::strchr(illegal, ch)) {
95 buffer[bufpos++] = ch;
96 } else if (bufpos + 3 >= buflen) {
99 buffer[bufpos+0] = escape;
100 buffer[bufpos+1] = hex_encode((static_cast<unsigned char>(ch) >> 4) & 0xF);
101 buffer[bufpos+2] = hex_encode((static_cast<unsigned char>(ch) ) & 0xF);
105 buffer[bufpos] = '\0';
109 size_t decode(char * buffer, size_t buflen,
110 const char * source, size_t srclen,
115 unsigned char h1, h2;
116 size_t srcpos = 0, bufpos = 0;
117 while ((srcpos < srclen) && (bufpos + 1 < buflen)) {
118 char ch = source[srcpos++];
120 && (srcpos + 1 < srclen)
121 && hex_decode(source[srcpos], &h1)
122 && hex_decode(source[srcpos+1], &h2)) {
123 buffer[bufpos++] = (h1 << 4) | h2;
126 buffer[bufpos++] = ch;
129 buffer[bufpos] = '\0';
133 const char* unsafe_filename_characters() {
134 // It might be better to have a single specification which is the union of
135 // all operating systems, unless one system is overly restrictive.
137 return "\\/:*?\"<>|";
145 const unsigned char URL_UNSAFE = 0x1; // 0-33 "#$%&+,/:;<=>?@[\]^`{|} 127
146 const unsigned char XML_UNSAFE = 0x2; // "&'<>
147 const unsigned char HTML_UNSAFE = 0x2; // "&'<>
149 // ! " # $ % & ' ( ) * + , - . / 0 1 2 3 4 6 5 7 8 9 : ; < = > ?
150 //@ 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 [ \ ] ^ _
151 //` 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 { | } ~
153 const unsigned char ASCII_CLASS[128] = {
154 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,
155 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,
156 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,
157 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,
160 size_t url_encode(char * buffer, size_t buflen,
161 const char * source, size_t srclen) {
163 return srclen * 3 + 1;
167 size_t srcpos = 0, bufpos = 0;
168 while ((srcpos < srclen) && (bufpos + 1 < buflen)) {
169 unsigned char ch = source[srcpos++];
170 if ((ch < 128) && (ASCII_CLASS[ch] & URL_UNSAFE)) {
171 if (bufpos + 3 >= buflen) {
174 buffer[bufpos+0] = '%';
175 buffer[bufpos+1] = hex_encode((ch >> 4) & 0xF);
176 buffer[bufpos+2] = hex_encode((ch ) & 0xF);
179 buffer[bufpos++] = ch;
182 buffer[bufpos] = '\0';
186 size_t url_decode(char * buffer, size_t buflen,
187 const char * source, size_t srclen) {
193 unsigned char h1, h2;
194 size_t srcpos = 0, bufpos = 0;
195 while ((srcpos < srclen) && (bufpos + 1 < buflen)) {
196 unsigned char ch = source[srcpos++];
198 buffer[bufpos++] = ' ';
199 } else if ((ch == '%')
200 && (srcpos + 1 < srclen)
201 && hex_decode(source[srcpos], &h1)
202 && hex_decode(source[srcpos+1], &h2))
204 buffer[bufpos++] = (h1 << 4) | h2;
207 buffer[bufpos++] = ch;
210 buffer[bufpos] = '\0';
214 size_t utf8_decode(const char* source, size_t srclen, unsigned long* value) {
215 const unsigned char* s = reinterpret_cast<const unsigned char*>(source);
216 if ((s[0] & 0x80) == 0x00) { // Check s[0] == 0xxxxxxx
220 if ((srclen < 2) || ((s[1] & 0xC0) != 0x80)) { // Check s[1] != 10xxxxxx
223 // Accumulate the trailer byte values in value16, and combine it with the
224 // relevant bits from s[0], once we've determined the sequence length.
225 unsigned long value16 = (s[1] & 0x3F);
226 if ((s[0] & 0xE0) == 0xC0) { // Check s[0] == 110xxxxx
227 *value = ((s[0] & 0x1F) << 6) | value16;
230 if ((srclen < 3) || ((s[2] & 0xC0) != 0x80)) { // Check s[2] != 10xxxxxx
233 value16 = (value16 << 6) | (s[2] & 0x3F);
234 if ((s[0] & 0xF0) == 0xE0) { // Check s[0] == 1110xxxx
235 *value = ((s[0] & 0x0F) << 12) | value16;
238 if ((srclen < 4) || ((s[3] & 0xC0) != 0x80)) { // Check s[3] != 10xxxxxx
241 value16 = (value16 << 6) | (s[3] & 0x3F);
242 if ((s[0] & 0xF8) == 0xF0) { // Check s[0] == 11110xxx
243 *value = ((s[0] & 0x07) << 18) | value16;
249 size_t utf8_encode(char* buffer, size_t buflen, unsigned long value) {
250 if ((value <= 0x7F) && (buflen >= 1)) {
251 buffer[0] = static_cast<unsigned char>(value);
254 if ((value <= 0x7FF) && (buflen >= 2)) {
255 buffer[0] = 0xC0 | static_cast<unsigned char>(value >> 6);
256 buffer[1] = 0x80 | static_cast<unsigned char>(value & 0x3F);
259 if ((value <= 0xFFFF) && (buflen >= 3)) {
260 buffer[0] = 0xE0 | static_cast<unsigned char>(value >> 12);
261 buffer[1] = 0x80 | static_cast<unsigned char>((value >> 6) & 0x3F);
262 buffer[2] = 0x80 | static_cast<unsigned char>(value & 0x3F);
265 if ((value <= 0x1FFFFF) && (buflen >= 4)) {
266 buffer[0] = 0xF0 | static_cast<unsigned char>(value >> 18);
267 buffer[1] = 0x80 | static_cast<unsigned char>((value >> 12) & 0x3F);
268 buffer[2] = 0x80 | static_cast<unsigned char>((value >> 6) & 0x3F);
269 buffer[3] = 0x80 | static_cast<unsigned char>(value & 0x3F);
275 size_t html_encode(char * buffer, size_t buflen,
276 const char * source, size_t srclen) {
277 ASSERT(NULL != buffer); // TODO: estimate output size
281 size_t srcpos = 0, bufpos = 0;
282 while ((srcpos < srclen) && (bufpos + 1 < buflen)) {
283 unsigned char ch = source[srcpos];
286 if (ASCII_CLASS[ch] & HTML_UNSAFE) {
287 const char * escseq = 0;
290 case '<': escseq = "<"; esclen = 4; break;
291 case '>': escseq = ">"; esclen = 4; break;
292 case '\'': escseq = "'"; esclen = 5; break;
293 case '\"': escseq = """; esclen = 6; break;
294 case '&': escseq = "&"; esclen = 5; break;
295 default: ASSERT(false);
297 if (bufpos + esclen >= buflen) {
300 memcpy(buffer + bufpos, escseq, esclen);
303 buffer[bufpos++] = ch;
306 // Largest value is 0x1FFFFF => � (10 characters)
309 if (size_t vallen = utf8_decode(&source[srcpos], srclen - srcpos, &val)) {
312 // Not a valid utf8 sequence, just use the raw character.
313 val = static_cast<unsigned char>(source[srcpos++]);
315 size_t esclen = sprintfn(escseq, ARRAY_SIZE(escseq), "&#%lu;", val);
316 if (bufpos + esclen >= buflen) {
319 memcpy(buffer + bufpos, escseq, esclen);
323 buffer[bufpos] = '\0';
327 size_t html_decode(char * buffer, size_t buflen,
328 const char * source, size_t srclen) {
329 ASSERT(NULL != buffer); // TODO: estimate output size
330 return xml_decode(buffer, buflen, source, srclen);
333 size_t xml_encode(char * buffer, size_t buflen,
334 const char * source, size_t srclen) {
335 ASSERT(NULL != buffer); // TODO: estimate output size
339 size_t srcpos = 0, bufpos = 0;
340 while ((srcpos < srclen) && (bufpos + 1 < buflen)) {
341 unsigned char ch = source[srcpos++];
342 if ((ch < 128) && (ASCII_CLASS[ch] & XML_UNSAFE)) {
343 const char * escseq = 0;
346 case '<': escseq = "<"; esclen = 4; break;
347 case '>': escseq = ">"; esclen = 4; break;
348 case '\'': escseq = "'"; esclen = 6; break;
349 case '\"': escseq = """; esclen = 6; break;
350 case '&': escseq = "&"; esclen = 5; break;
351 default: ASSERT(false);
353 if (bufpos + esclen >= buflen) {
356 memcpy(buffer + bufpos, escseq, esclen);
359 buffer[bufpos++] = ch;
362 buffer[bufpos] = '\0';
366 size_t xml_decode(char * buffer, size_t buflen,
367 const char * source, size_t srclen) {
368 ASSERT(NULL != buffer); // TODO: estimate output size
372 size_t srcpos = 0, bufpos = 0;
373 while ((srcpos < srclen) && (bufpos + 1 < buflen)) {
374 unsigned char ch = source[srcpos++];
376 buffer[bufpos++] = ch;
377 } else if ((srcpos + 2 < srclen)
378 && (memcmp(source + srcpos, "lt;", 3) == 0)) {
379 buffer[bufpos++] = '<';
381 } else if ((srcpos + 2 < srclen)
382 && (memcmp(source + srcpos, "gt;", 3) == 0)) {
383 buffer[bufpos++] = '>';
385 } else if ((srcpos + 4 < srclen)
386 && (memcmp(source + srcpos, "apos;", 5) == 0)) {
387 buffer[bufpos++] = '\'';
389 } else if ((srcpos + 4 < srclen)
390 && (memcmp(source + srcpos, "quot;", 5) == 0)) {
391 buffer[bufpos++] = '\"';
393 } else if ((srcpos + 3 < srclen)
394 && (memcmp(source + srcpos, "amp;", 4) == 0)) {
395 buffer[bufpos++] = '&';
397 } else if ((srcpos < srclen) && (source[srcpos] == '#')) {
399 if ((srcpos + 1 < srclen) && (source[srcpos+1] == 'x')) {
404 // TODO: Fix hack (ptr may go past end of data)
405 unsigned long val = strtoul(source + srcpos + 1, &ptr, int_base);
406 if ((static_cast<size_t>(ptr - source) < srclen) && (*ptr == ';')) {
407 srcpos = ptr - source + 1;
409 // Not a valid escape sequence.
412 if (size_t esclen = utf8_encode(buffer + bufpos, buflen - bufpos, val)) {
415 // Not enough room to encode the character, or illegal character
419 // Unrecognized escape sequence.
423 buffer[bufpos] = '\0';
427 static const char HEX[] = "0123456789abcdef";
429 char hex_encode(unsigned char val) {
431 return (val < 16) ? HEX[val] : '!';
434 bool hex_decode(char ch, unsigned char* val) {
435 if ((ch >= '0') && (ch <= '9')) {
437 } else if ((ch >= 'A') && (ch <= 'Z')) {
438 *val = (ch - 'A') + 10;
439 } else if ((ch >= 'a') && (ch <= 'z')) {
440 *val = (ch - 'a') + 10;
447 size_t hex_encode(char* buffer, size_t buflen,
448 const char* csource, size_t srclen) {
449 return hex_encode_with_delimiter(buffer, buflen, csource, srclen, 0);
452 size_t hex_encode_with_delimiter(char* buffer, size_t buflen,
453 const char* csource, size_t srclen,
455 ASSERT(NULL != buffer); // TODO: estimate output size
459 // Init and check bounds.
460 const unsigned char* bsource =
461 reinterpret_cast<const unsigned char*>(csource);
462 size_t srcpos = 0, bufpos = 0;
463 size_t needed = delimiter ? (srclen * 3) : (srclen * 2 + 1);
467 while (srcpos < srclen) {
468 unsigned char ch = bsource[srcpos++];
469 buffer[bufpos ] = hex_encode((ch >> 4) & 0xF);
470 buffer[bufpos+1] = hex_encode((ch ) & 0xF);
473 // Don't write a delimiter after the last byte.
474 if (delimiter && (srcpos < srclen)) {
475 buffer[bufpos] = delimiter;
481 buffer[bufpos] = '\0';
485 std::string hex_encode(const char* source, size_t srclen) {
486 return hex_encode_with_delimiter(source, srclen, 0);
489 std::string hex_encode_with_delimiter(const char* source, size_t srclen,
491 const size_t kBufferSize = srclen * 3;
492 char* buffer = STACK_ARRAY(char, kBufferSize);
493 size_t length = hex_encode_with_delimiter(buffer, kBufferSize,
494 source, srclen, delimiter);
495 ASSERT(srclen == 0 || length > 0);
496 return std::string(buffer, length);
499 size_t hex_decode(char * cbuffer, size_t buflen,
500 const char * source, size_t srclen) {
501 return hex_decode_with_delimiter(cbuffer, buflen, source, srclen, 0);
504 size_t hex_decode_with_delimiter(char* cbuffer, size_t buflen,
505 const char* source, size_t srclen,
507 ASSERT(NULL != cbuffer); // TODO: estimate output size
511 // Init and bounds check.
512 unsigned char* bbuffer = reinterpret_cast<unsigned char*>(cbuffer);
513 size_t srcpos = 0, bufpos = 0;
514 size_t needed = (delimiter) ? (srclen + 1) / 3 : srclen / 2;
518 while (srcpos < srclen) {
519 if ((srclen - srcpos) < 2) {
520 // This means we have an odd number of bytes.
524 unsigned char h1, h2;
525 if (!hex_decode(source[srcpos], &h1) ||
526 !hex_decode(source[srcpos + 1], &h2))
529 bbuffer[bufpos++] = (h1 << 4) | h2;
532 // Remove the delimiter if needed.
533 if (delimiter && (srclen - srcpos) > 1) {
534 if (source[srcpos] != delimiter)
543 size_t hex_decode(char* buffer, size_t buflen, const std::string& source) {
544 return hex_decode_with_delimiter(buffer, buflen, source, 0);
546 size_t hex_decode_with_delimiter(char* buffer, size_t buflen,
547 const std::string& source, char delimiter) {
548 return hex_decode_with_delimiter(buffer, buflen,
549 source.c_str(), source.length(), delimiter);
552 size_t transform(std::string& value, size_t maxlen, const std::string& source,
554 char* buffer = STACK_ARRAY(char, maxlen + 1);
555 size_t length = t(buffer, maxlen + 1, source.data(), source.length());
556 value.assign(buffer, length);
560 std::string s_transform(const std::string& source, Transform t) {
561 // Ask transformation function to approximate the destination size (returns upper bound)
562 size_t maxlen = t(NULL, 0, source.data(), source.length());
563 char * buffer = STACK_ARRAY(char, maxlen);
564 size_t len = t(buffer, maxlen, source.data(), source.length());
565 std::string result(buffer, len);
569 size_t tokenize(const std::string& source, char delimiter,
570 std::vector<std::string>* fields) {
571 ASSERT(NULL != fields);
574 for (size_t i = 0; i < source.length(); ++i) {
575 if (source[i] == delimiter) {
577 fields->push_back(source.substr(last, i - last));
582 if (last != source.length()) {
583 fields->push_back(source.substr(last, source.length() - last));
585 return fields->size();
588 size_t tokenize_append(const std::string& source, char delimiter,
589 std::vector<std::string>* fields) {
590 if (!fields) return 0;
592 std::vector<std::string> new_fields;
593 tokenize(source, delimiter, &new_fields);
594 fields->insert(fields->end(), new_fields.begin(), new_fields.end());
595 return fields->size();
598 size_t tokenize(const std::string& source, char delimiter, char start_mark,
599 char end_mark, std::vector<std::string>* fields) {
600 if (!fields) return 0;
603 std::string remain_source = source;
604 while (!remain_source.empty()) {
605 size_t start_pos = remain_source.find(start_mark);
606 if (std::string::npos == start_pos) break;
607 std::string pre_mark;
609 pre_mark = remain_source.substr(0, start_pos - 1);
613 size_t end_pos = remain_source.find(end_mark, start_pos);
614 if (std::string::npos == end_pos) break;
616 // We have found the matching marks. First tokenize the pre-mask. Then add
617 // the marked part as a single field. Finally, loop back for the post-mark.
618 tokenize_append(pre_mark, delimiter, fields);
619 fields->push_back(remain_source.substr(start_pos, end_pos - start_pos));
620 remain_source = remain_source.substr(end_pos + 1);
623 return tokenize_append(remain_source, delimiter, fields);
626 size_t split(const std::string& source, char delimiter,
627 std::vector<std::string>* fields) {
628 ASSERT(NULL != fields);
631 for (size_t i = 0; i < source.length(); ++i) {
632 if (source[i] == delimiter) {
633 fields->push_back(source.substr(last, i - last));
637 fields->push_back(source.substr(last, source.length() - last));
638 return fields->size();
641 char make_char_safe_for_filename(char c) {
663 void sprintf(std::string& value, size_t maxlen, const char * format, ...) {
664 char * buffer = STACK_ARRAY(char, maxlen + 1);
666 va_start(args, format);
667 value.assign(buffer, vsprintfn(buffer, maxlen + 1, format, args));
672 /////////////////////////////////////////////////////////////////////////////
674 } // namespace talk_base