-
Notifications
You must be signed in to change notification settings - Fork 8
/
iconvstream.cpp
367 lines (332 loc) · 9.14 KB
/
iconvstream.cpp
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
/*
* Copyright 2020-2023 Fabian Groffen <grobian@gentoo.org>
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License in the file COPYING for more details.
*/
#include <stdlib.h>
#include <stdio.h>
#include <string.h>
#include <unistd.h>
#include <ctype.h>
#include <errno.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <fcntl.h>
#include <errno.h>
#include <iconv.h>
#include <iostream>
#include "iconvstream.h"
void
iconvstream::open_is(const char *file_name, const char *encoding_in)
{
close_is();
encoding = encoding_in;
fd_is =
strcmp(file_name, "-") == 0 ? ::dup(0) : ::open(file_name, O_RDONLY);
if (fd_is == -1)
open_err = strerror(errno);
readbufsze = 1024; /* matches w3c's req for content type declaration */
rutf8bufsze = readbufsze * 4;
readbuflen = 0;
rutf8buflen = 0;
readbufpos = 0;
rutf8bufpos = 0;
readbuf = new unsigned char[readbufsze];
rutf8buf = new unsigned char[rutf8bufsze];
/* trigger charset detection, and reset the pointer afterwards,
* doing this now generates an error if the charset is invalid,
* which is not expected to be set during reading */
if (!open_err) {
get();
rutf8bufpos = 0;
if (open_err) {
close_is();
fd_is = -1;
}
}
}
void
iconvstream::open_is(const string &url, const char *encoding)
{
open_is(url.c_str(), encoding);
}
void
iconvstream::close_is(void)
{
if (is_open()) {
::close(fd_is);
if (iconv_handle_is != nullptr)
iconv_close(iconv_handle_is);
delete[] readbuf;
delete[] rutf8buf;
}
}
void
iconvstream::open_os(const char *file_name, const char *encoding)
{
close_os();
/* internal processing uses UTF-8 */
iconv_handle_os = iconv_open(encoding, "UTF-8");
if (iconv_handle_os == iconv_t(-1)) {
open_err = "invalid to_encoding";
return;
}
fd_os = strcmp(file_name, "-") == 0 ?
::dup(1) :
::open(file_name, O_WRONLY | O_CREAT,
S_IRUSR | S_IWUSR | S_IRGRP | S_IROTH);
if (fd_os == -1)
open_err = strerror(errno);
writebufsze = 1024;
wutf8bufsze = writebufsze * 4; /* worst case scenario UTF-32 */
writebuflen = 0;
wutf8buflen = 0;
writebufpos = 0;
wutf8bufpos = 0;
writebuf = new unsigned char[writebufsze];
wutf8buf = new unsigned char[wutf8bufsze];
}
void
iconvstream::open_os(const string &url, const char *encoding)
{
open_os(url.c_str(), encoding);
}
void
iconvstream::close_os(void)
{
if (os_open()) {
*this << flush;
::close(fd_os);
if (iconv_handle_os != nullptr)
iconv_close(iconv_handle_os);
delete[] writebuf;
delete[] wutf8buf;
}
}
const char *
iconvstream::find_tokens(char *buf, size_t len, const char **tokens)
{
char *startp = NULL;
char *curp;
const char **curtoken = tokens;
char startfound = 0;
for (curp = buf; (size_t)(curp - buf) < len; curp++) {
/* [a-zA-Z0-9/-]+ */
if ((*curp >= 'a' && *curp <= 'z') ||
(*curp >= 'A' && *curp <= 'Z') ||
(*curp >= '0' && *curp <= '9') ||
*curp == '/' || *curp == '-')
{
if (startp == NULL)
startp = curp;
}
else
{
if (startp != NULL) {
if (*curtoken == NULL) {
if (find_tokens_c_str_buf != nullptr)
delete[] find_tokens_c_str_buf;
find_tokens_c_str_buf = new char[(curp - startp) + 1];
memcpy(find_tokens_c_str_buf, startp, curp - startp);
find_tokens_c_str_buf[curp - startp] = '\0';
return find_tokens_c_str_buf;
} else if (strlen(*curtoken) == (size_t)(curp - startp) &&
strncasecmp(*curtoken, startp, curp - startp) == 0)
{
if (startfound != 0 || startp[-1] == '<') {
startfound = 1;
curtoken++;
}
} else if (startfound == 1) {
startfound = 0;
curtoken = tokens;
}
}
startp = NULL;
}
}
return NULL;
}
const char *
iconvstream::open_error_msg() const
{
return open_err ? open_err : "No error";
}
int
iconvstream::get()
{
if (rutf8bufpos == rutf8buflen) {
char *procinp = (char *)readbuf;
char *procout = (char *)rutf8buf;
size_t iconvret;
readbuflen = read(fd_is, readbuf + readbufpos, readbufsze - readbufpos);
if (readbuflen <= 0)
return EOF;
if (rutf8buflen == 0) {
/* on first read, figure out what encoding this is, unless a
* specific override is in place */
if (strcmp(encoding, "auto") == 0) {
/* look for UTF-BOM, this should override any meta
* declaration (feels like a safe way for M$ to screw
* this up, but let's go with this for now)
* https://www.w3.org/International/questions/qa-html-encoding-declarations#bom */
if (readbuflen >= 2 && memcmp(readbuf, "\ufeff", 2) == 0) {
encoding = "UTF-8";
} else {
const char *tokens_charset[] = {
"meta", "charset", NULL
};
const char *tokens_contenttype[] = {
"meta", "http-equiv", "content-type",
"content", "text/html", "charset", NULL
};
/* hunt down meta, which can be two forms
* - <meta charset="utf-8"/>
* - <meta http-equiv="Content-Type"
* content="text/html; charset=utf-8"/>
* https://www.w3.org/International/questions/qa-html-encoding-declarations#quickanswer
* now there is the correct way, which would be to
* parse whatever xml, and the quick 'n' dirty way,
* which is to simply do some lame-@$$ parsing */
encoding = find_tokens((char *)readbuf, readbuflen,
tokens_charset);
if (encoding == NULL) {
encoding = find_tokens((char *)readbuf, readbuflen,
tokens_contenttype);
}
/* fall back to lame historical default */
if (encoding == NULL)
encoding = "ISO-8859-1";
}
}
/* we always encode to UTF-8 for internal processing */
iconv_handle_is = iconv_open("UTF-8", encoding);
if (iconv_handle_is == iconv_t(-1)) {
open_err = "invalid from_encoding";
return 0;
}
}
readbuflen += readbufpos;
readbufpos = 0;
rutf8buflen = rutf8bufsze;
/* iconv updates readbuflen and utf8buflen */
do {
iconvret = iconv(iconv_handle_is, &procinp, &readbuflen,
&procout, &rutf8buflen);
if (iconvret == (size_t)-1) {
switch (errno) {
case EILSEQ:
/* byte is invalid, try to step over it */
*procout++ = '?';
procinp++;
readbuflen--;
rutf8buflen--;
break;
case EINVAL:
/* this typically means we stopped reading halfway,
* e.g. this is fine, resume next time */
memmove(readbuf, procinp, readbuflen);
readbufpos = readbuflen;
iconvret = 0;
break;
case E2BIG:
/* output buffer is not large enough, this is
* impossible since we allocate 4x */
default:
return EOF;
}
}
} while (iconvret == (size_t)-1);
rutf8bufpos = 0;
rutf8buflen = procout - (char *)rutf8buf;
}
return rutf8bufpos < rutf8buflen ? rutf8buf[rutf8bufpos++] : EOF;
}
int
iconvstream::write(const char *inp, size_t len)
{
size_t outlen = len;
size_t copylen;
size_t copypos = 0;
int ret = -1;
do {
copylen = writebufsze - writebufpos;
if (copylen > outlen)
copylen = outlen;
memcpy(writebuf + writebufpos, inp + copypos, copylen);
outlen -= copylen;
copypos += copylen;
writebufpos += copylen;
/* flush if the buffer is full, or when explicitly requested */
if (writebufpos == writebufsze || len == 0) {
size_t iconvret;
size_t inplen = writebufpos;
char *procinp = (char *)writebuf;
char *procout = (char *)wutf8buf;
writebufpos = 0;
wutf8buflen = wutf8bufsze;
do {
/* iconv updates len and writebuflen */
iconvret = iconv(iconv_handle_os, &procinp, &inplen,
&procout, &wutf8buflen);
if (iconvret == (size_t)-1) {
switch (errno) {
case EILSEQ:
/* byte is invalid, try to step over it,
* this shouldn't happen for the input is
* generated by iconv itself during input */
*procout++ = '?';
procinp++;
inplen--;
wutf8buflen--;
break;
case EINVAL:
/* the only valid problem should be the end
* of the input being truncated */
if (inplen < 4 && len != 0) {
/* shift this so a next attempt can
* retry the completion */
memmove(writebuf, procinp, inplen);
writebufpos = inplen;
iconvret = 0;
break;
}
/* fall through */
case E2BIG:
/* output buffer is not large enough, this is
* impossible since we allocate 4x */
default:
return EOF;
}
}
} while (iconvret == (size_t)-1);
ret = ::write(fd_os, wutf8buf, procout - (char *)wutf8buf);
} else {
ret = len;
}
} while (outlen > 0);
return ret;
}
iconvstream &iconvstream::operator<<(const char *inp)
{
(void)write(inp, strlen(inp));
return *this;
}
iconvstream &iconvstream::operator<<(const string &inp)
{
(void)write(inp.c_str(), inp.length());
return *this;
}
iconvstream &iconvstream::operator<<(char inp)
{
(void)write(&inp, inp == '\0' ? 0 : 1);
return *this;
}