|
a |
|
b/src/bincimapmime/mime-parsefull.cc |
|
|
1 |
/* -*- Mode: c++; -*- */
|
|
|
2 |
/* --------------------------------------------------------------------
|
|
|
3 |
* Filename:
|
|
|
4 |
* mime-parsefull.cc
|
|
|
5 |
*
|
|
|
6 |
* Description:
|
|
|
7 |
* Implementation of main mime parser components
|
|
|
8 |
* --------------------------------------------------------------------
|
|
|
9 |
* Copyright 2002-2004 Andreas Aardal Hanssen
|
|
|
10 |
*
|
|
|
11 |
* This program is free software; you can redistribute it and/or modify
|
|
|
12 |
* it under the terms of the GNU General Public License as published by
|
|
|
13 |
* the Free Software Foundation; either version 2 of the License, or
|
|
|
14 |
* (at your option) any later version.
|
|
|
15 |
*
|
|
|
16 |
* This program is distributed in the hope that it will be useful,
|
|
|
17 |
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
18 |
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
|
19 |
* GNU General Public License for more details.
|
|
|
20 |
*
|
|
|
21 |
* You should have received a copy of the GNU General Public License
|
|
|
22 |
* along with this program; if not, write to the Free Software
|
|
|
23 |
* Foundation, Inc., 59 Temple Street #330, Boston, MA 02111-1307, USA.
|
|
|
24 |
* --------------------------------------------------------------------
|
|
|
25 |
*/
|
|
|
26 |
#ifdef HAVE_CONFIG_H
|
|
|
27 |
#include <config.h>
|
|
|
28 |
#endif
|
|
|
29 |
|
|
|
30 |
#include "mime.h"
|
|
|
31 |
#include "mime-utils.h"
|
|
|
32 |
#include "mime-inputsource.h"
|
|
|
33 |
#include "convert.h"
|
|
|
34 |
#include <string>
|
|
|
35 |
#include <vector>
|
|
|
36 |
#include <map>
|
|
|
37 |
#include <exception>
|
|
|
38 |
#include <iostream>
|
|
|
39 |
|
|
|
40 |
#include <string.h>
|
|
|
41 |
#include <ctype.h>
|
|
|
42 |
#include <stdio.h>
|
|
|
43 |
#include <errno.h>
|
|
|
44 |
|
|
|
45 |
Binc::MimeInputSource *mimeSource = 0;
|
|
|
46 |
|
|
|
47 |
using namespace ::std;
|
|
|
48 |
|
|
|
49 |
//------------------------------------------------------------------------
|
|
|
50 |
void Binc::MimeDocument::parseFull(int fd) const
|
|
|
51 |
{
|
|
|
52 |
if (allIsParsed)
|
|
|
53 |
return;
|
|
|
54 |
|
|
|
55 |
allIsParsed = true;
|
|
|
56 |
|
|
|
57 |
if (!mimeSource || mimeSource->getFileDescriptor() != fd) {
|
|
|
58 |
delete mimeSource;
|
|
|
59 |
mimeSource = new MimeInputSource(fd);
|
|
|
60 |
} else {
|
|
|
61 |
mimeSource->reset();
|
|
|
62 |
}
|
|
|
63 |
|
|
|
64 |
headerstartoffsetcrlf = 0;
|
|
|
65 |
headerlength = 0;
|
|
|
66 |
bodystartoffsetcrlf = 0;
|
|
|
67 |
bodylength = 0;
|
|
|
68 |
size = 0;
|
|
|
69 |
messagerfc822 = false;
|
|
|
70 |
multipart = false;
|
|
|
71 |
|
|
|
72 |
int bsize = 0;
|
|
|
73 |
string bound;
|
|
|
74 |
MimePart::parseFull(bound, bsize);
|
|
|
75 |
|
|
|
76 |
// eat any trailing junk to get the correct size
|
|
|
77 |
char c;
|
|
|
78 |
while (mimeSource->getChar(&c));
|
|
|
79 |
|
|
|
80 |
size = mimeSource->getOffset();
|
|
|
81 |
}
|
|
|
82 |
|
|
|
83 |
//------------------------------------------------------------------------
|
|
|
84 |
static bool parseOneHeaderLine(Binc::Header *header, unsigned int *nlines)
|
|
|
85 |
{
|
|
|
86 |
using namespace ::Binc;
|
|
|
87 |
char c;
|
|
|
88 |
bool eof = false;
|
|
|
89 |
char cqueue[4];
|
|
|
90 |
string name;
|
|
|
91 |
string content;
|
|
|
92 |
|
|
|
93 |
while (mimeSource->getChar(&c)) {
|
|
|
94 |
// If we encounter a \r before we got to the first ':', then
|
|
|
95 |
// rewind back to the start of the line and assume we're at the
|
|
|
96 |
// start of the body.
|
|
|
97 |
if (c == '\r') {
|
|
|
98 |
for (int i = 0; i < (int) name.length() + 1; ++i)
|
|
|
99 |
mimeSource->ungetChar();
|
|
|
100 |
return false;
|
|
|
101 |
}
|
|
|
102 |
|
|
|
103 |
// A colon marks the end of the header name
|
|
|
104 |
if (c == ':') break;
|
|
|
105 |
|
|
|
106 |
// Otherwise add to the header name
|
|
|
107 |
name += c;
|
|
|
108 |
}
|
|
|
109 |
|
|
|
110 |
cqueue[0] = '\0';
|
|
|
111 |
cqueue[1] = '\0';
|
|
|
112 |
cqueue[2] = '\0';
|
|
|
113 |
cqueue[3] = '\0';
|
|
|
114 |
|
|
|
115 |
// Read until the end of the header.
|
|
|
116 |
bool endOfHeaders = false;
|
|
|
117 |
while (!endOfHeaders) {
|
|
|
118 |
if (!mimeSource->getChar(&c)) {
|
|
|
119 |
eof = true;
|
|
|
120 |
break;
|
|
|
121 |
}
|
|
|
122 |
|
|
|
123 |
if (c == '\n') ++*nlines;
|
|
|
124 |
|
|
|
125 |
for (int i = 0; i < 3; ++i)
|
|
|
126 |
cqueue[i] = cqueue[i + 1];
|
|
|
127 |
cqueue[3] = c;
|
|
|
128 |
|
|
|
129 |
if (strncmp(cqueue, "\r\n\r\n", 4) == 0) {
|
|
|
130 |
endOfHeaders = true;
|
|
|
131 |
break;
|
|
|
132 |
}
|
|
|
133 |
|
|
|
134 |
// If the last character was a newline, and the first now is not
|
|
|
135 |
// whitespace, then rewind one character and store the current
|
|
|
136 |
// key,value pair.
|
|
|
137 |
if (cqueue[2] == '\n' && c != ' ' && c != '\t') {
|
|
|
138 |
if (content.length() > 2)
|
|
|
139 |
content.resize(content.length() - 2);
|
|
|
140 |
|
|
|
141 |
trim(content);
|
|
|
142 |
header->add(name, content);
|
|
|
143 |
|
|
|
144 |
if (c != '\r') {
|
|
|
145 |
mimeSource->ungetChar();
|
|
|
146 |
if (c == '\n') --*nlines;
|
|
|
147 |
return true;
|
|
|
148 |
}
|
|
|
149 |
|
|
|
150 |
mimeSource->getChar(&c);
|
|
|
151 |
return false;
|
|
|
152 |
}
|
|
|
153 |
|
|
|
154 |
content += c;
|
|
|
155 |
}
|
|
|
156 |
|
|
|
157 |
if (name != "") {
|
|
|
158 |
if (content.length() > 2)
|
|
|
159 |
content.resize(content.length() - 2);
|
|
|
160 |
header->add(name, content);
|
|
|
161 |
}
|
|
|
162 |
|
|
|
163 |
return !(eof || endOfHeaders);
|
|
|
164 |
}
|
|
|
165 |
|
|
|
166 |
//------------------------------------------------------------------------
|
|
|
167 |
static void parseHeader(Binc::Header *header, unsigned int *nlines)
|
|
|
168 |
{
|
|
|
169 |
while (parseOneHeaderLine(header, nlines))
|
|
|
170 |
{ }
|
|
|
171 |
}
|
|
|
172 |
|
|
|
173 |
//------------------------------------------------------------------------
|
|
|
174 |
static void analyzeHeader(Binc::Header *header, bool *multipart,
|
|
|
175 |
bool *messagerfc822, string *subtype,
|
|
|
176 |
string *boundary)
|
|
|
177 |
{
|
|
|
178 |
using namespace ::Binc;
|
|
|
179 |
|
|
|
180 |
// Do simple parsing of headers to determine the
|
|
|
181 |
// type of message (multipart,messagerfc822 etc)
|
|
|
182 |
HeaderItem ctype;
|
|
|
183 |
if (header->getFirstHeader("content-type", ctype)) {
|
|
|
184 |
vector<string> types;
|
|
|
185 |
split(ctype.getValue(), ";", types);
|
|
|
186 |
|
|
|
187 |
if (types.size() > 0) {
|
|
|
188 |
// first element should describe content type
|
|
|
189 |
string tmp = types[0];
|
|
|
190 |
trim(tmp);
|
|
|
191 |
vector<string> v;
|
|
|
192 |
split(tmp, "/", v);
|
|
|
193 |
string key, value;
|
|
|
194 |
|
|
|
195 |
key = (v.size() > 0) ? v[0] : "text";
|
|
|
196 |
value = (v.size() > 1) ? v[1] : "plain";
|
|
|
197 |
lowercase(key);
|
|
|
198 |
|
|
|
199 |
if (key == "multipart") {
|
|
|
200 |
*multipart = true;
|
|
|
201 |
lowercase(value);
|
|
|
202 |
*subtype = value;
|
|
|
203 |
} else if (key == "message") {
|
|
|
204 |
lowercase(value);
|
|
|
205 |
if (value == "rfc822")
|
|
|
206 |
*messagerfc822 = true;
|
|
|
207 |
}
|
|
|
208 |
}
|
|
|
209 |
|
|
|
210 |
for (vector<string>::const_iterator i = types.begin();
|
|
|
211 |
i != types.end(); ++i) {
|
|
|
212 |
string element = *i;
|
|
|
213 |
trim(element);
|
|
|
214 |
|
|
|
215 |
if (element.find("=") != string::npos) {
|
|
|
216 |
string::size_type pos = element.find('=');
|
|
|
217 |
string key = element.substr(0, pos);
|
|
|
218 |
string value = element.substr(pos + 1);
|
|
|
219 |
|
|
|
220 |
lowercase(key);
|
|
|
221 |
trim(key);
|
|
|
222 |
|
|
|
223 |
if (key == "boundary") {
|
|
|
224 |
trim(value, " \"");
|
|
|
225 |
*boundary = value;
|
|
|
226 |
}
|
|
|
227 |
}
|
|
|
228 |
}
|
|
|
229 |
}
|
|
|
230 |
}
|
|
|
231 |
|
|
|
232 |
static void parseMessageRFC822(vector<Binc::MimePart> *members,
|
|
|
233 |
bool *foundendofpart,
|
|
|
234 |
unsigned int *bodylength,
|
|
|
235 |
unsigned int *nbodylines,
|
|
|
236 |
const string &toboundary)
|
|
|
237 |
{
|
|
|
238 |
using namespace ::Binc;
|
|
|
239 |
|
|
|
240 |
// message rfc822 means a completely enclosed mime document. we
|
|
|
241 |
// call the parser recursively, and pass on the boundary string
|
|
|
242 |
// that we got. when parse() finds this boundary, it returns 0. if
|
|
|
243 |
// it finds the end boundary (boundary + "--"), it returns != 0.
|
|
|
244 |
MimePart m;
|
|
|
245 |
|
|
|
246 |
unsigned int bodystartoffsetcrlf = mimeSource->getOffset();
|
|
|
247 |
|
|
|
248 |
// parsefull returns the number of bytes that need to be removed
|
|
|
249 |
// from the body because of the terminating boundary string.
|
|
|
250 |
int bsize = 0;
|
|
|
251 |
if (m.parseFull(toboundary, bsize))
|
|
|
252 |
*foundendofpart = true;
|
|
|
253 |
|
|
|
254 |
// make sure bodylength doesn't overflow
|
|
|
255 |
*bodylength = mimeSource->getOffset();
|
|
|
256 |
if (*bodylength >= bodystartoffsetcrlf) {
|
|
|
257 |
*bodylength -= bodystartoffsetcrlf;
|
|
|
258 |
if (*bodylength >= (unsigned int) bsize) {
|
|
|
259 |
*bodylength -= (unsigned int) bsize;
|
|
|
260 |
} else {
|
|
|
261 |
*bodylength = 0;
|
|
|
262 |
}
|
|
|
263 |
} else {
|
|
|
264 |
*bodylength = 0;
|
|
|
265 |
}
|
|
|
266 |
|
|
|
267 |
*nbodylines += m.getNofLines();
|
|
|
268 |
|
|
|
269 |
members->push_back(m);
|
|
|
270 |
}
|
|
|
271 |
|
|
|
272 |
static bool skipUntilBoundary(const string &delimiter,
|
|
|
273 |
unsigned int *nlines, bool *eof)
|
|
|
274 |
{
|
|
|
275 |
int endpos = delimiter.length();
|
|
|
276 |
char *delimiterqueue = 0;
|
|
|
277 |
int delimiterpos = 0;
|
|
|
278 |
const char *delimiterStr = delimiter.c_str();
|
|
|
279 |
if (delimiter != "") {
|
|
|
280 |
delimiterqueue = new char[endpos];
|
|
|
281 |
memset(delimiterqueue, 0, endpos);
|
|
|
282 |
}
|
|
|
283 |
|
|
|
284 |
// first, skip to the first delimiter string. Anything between the
|
|
|
285 |
// header and the first delimiter string is simply ignored (it's
|
|
|
286 |
// usually a text message intended for non-mime clients)
|
|
|
287 |
char c;
|
|
|
288 |
|
|
|
289 |
bool foundBoundary = false;
|
|
|
290 |
for (;;) {
|
|
|
291 |
if (!mimeSource->getChar(&c)) {
|
|
|
292 |
*eof = true;
|
|
|
293 |
break;
|
|
|
294 |
}
|
|
|
295 |
|
|
|
296 |
if (c == '\n')
|
|
|
297 |
++*nlines;
|
|
|
298 |
|
|
|
299 |
// if there is no delimiter, we just read until the end of the
|
|
|
300 |
// file.
|
|
|
301 |
if (!delimiterqueue)
|
|
|
302 |
continue;
|
|
|
303 |
|
|
|
304 |
delimiterqueue[delimiterpos++ % endpos] = c;
|
|
|
305 |
|
|
|
306 |
if (compareStringToQueue(delimiterStr, delimiterqueue,
|
|
|
307 |
delimiterpos, endpos)) {
|
|
|
308 |
foundBoundary = true;
|
|
|
309 |
break;
|
|
|
310 |
}
|
|
|
311 |
}
|
|
|
312 |
|
|
|
313 |
delete [] delimiterqueue;
|
|
|
314 |
delimiterqueue = 0;
|
|
|
315 |
|
|
|
316 |
return foundBoundary;
|
|
|
317 |
}
|
|
|
318 |
|
|
|
319 |
|
|
|
320 |
static void parseMultipart(const string &boundary,
|
|
|
321 |
const string &toboundary,
|
|
|
322 |
bool *eof,
|
|
|
323 |
unsigned int *nlines,
|
|
|
324 |
int *boundarysize,
|
|
|
325 |
bool *foundendofpart,
|
|
|
326 |
unsigned int *bodylength,
|
|
|
327 |
vector<Binc::MimePart> *members)
|
|
|
328 |
{
|
|
|
329 |
using namespace ::Binc;
|
|
|
330 |
unsigned int bodystartoffsetcrlf = mimeSource->getOffset();
|
|
|
331 |
|
|
|
332 |
// multipart parsing starts with skipping to the first
|
|
|
333 |
// boundary. then we call parse() for all parts. the last parse()
|
|
|
334 |
// command will return a code indicating that it found the last
|
|
|
335 |
// boundary of this multipart. Note that the first boundary does
|
|
|
336 |
// not have to start with CRLF.
|
|
|
337 |
string delimiter = "--" + boundary;
|
|
|
338 |
|
|
|
339 |
skipUntilBoundary(delimiter, nlines, eof);
|
|
|
340 |
|
|
|
341 |
if (!eof)
|
|
|
342 |
*boundarysize = delimiter.size();
|
|
|
343 |
|
|
|
344 |
// Read two more characters. This may be CRLF, it may be "--" and
|
|
|
345 |
// it may be any other two characters.
|
|
|
346 |
char a;
|
|
|
347 |
if (!mimeSource->getChar(&a))
|
|
|
348 |
*eof = true;
|
|
|
349 |
|
|
|
350 |
if (a == '\n')
|
|
|
351 |
++*nlines;
|
|
|
352 |
|
|
|
353 |
char b;
|
|
|
354 |
if (!mimeSource->getChar(&b))
|
|
|
355 |
*eof = true;
|
|
|
356 |
|
|
|
357 |
if (b == '\n')
|
|
|
358 |
++*nlines;
|
|
|
359 |
|
|
|
360 |
// If we find two dashes after the boundary, then this is the end
|
|
|
361 |
// of boundary marker.
|
|
|
362 |
if (!*eof) {
|
|
|
363 |
if (a == '-' && b == '-') {
|
|
|
364 |
*foundendofpart = true;
|
|
|
365 |
*boundarysize += 2;
|
|
|
366 |
|
|
|
367 |
if (!mimeSource->getChar(&a))
|
|
|
368 |
*eof = true;
|
|
|
369 |
|
|
|
370 |
if (a == '\n')
|
|
|
371 |
++*nlines;
|
|
|
372 |
|
|
|
373 |
if (!mimeSource->getChar(&b))
|
|
|
374 |
*eof = true;
|
|
|
375 |
|
|
|
376 |
if (b == '\n')
|
|
|
377 |
++*nlines;
|
|
|
378 |
}
|
|
|
379 |
|
|
|
380 |
if (a == '\r' && b == '\n') {
|
|
|
381 |
// This exception is to handle a special case where the
|
|
|
382 |
// delimiter of one part is not followed by CRLF, but
|
|
|
383 |
// immediately followed by a CRLF prefixed delimiter.
|
|
|
384 |
if (!mimeSource->getChar(&a) || !mimeSource->getChar(&b))
|
|
|
385 |
*eof = true;
|
|
|
386 |
else if (a == '-' && b == '-') {
|
|
|
387 |
mimeSource->ungetChar();
|
|
|
388 |
mimeSource->ungetChar();
|
|
|
389 |
mimeSource->ungetChar();
|
|
|
390 |
mimeSource->ungetChar();
|
|
|
391 |
} else {
|
|
|
392 |
mimeSource->ungetChar();
|
|
|
393 |
mimeSource->ungetChar();
|
|
|
394 |
}
|
|
|
395 |
|
|
|
396 |
*boundarysize += 2;
|
|
|
397 |
} else {
|
|
|
398 |
mimeSource->ungetChar();
|
|
|
399 |
mimeSource->ungetChar();
|
|
|
400 |
}
|
|
|
401 |
}
|
|
|
402 |
|
|
|
403 |
// read all mime parts.
|
|
|
404 |
if (!*foundendofpart && !*eof) {
|
|
|
405 |
bool quit = false;
|
|
|
406 |
do {
|
|
|
407 |
MimePart m;
|
|
|
408 |
|
|
|
409 |
// If parseFull returns != 0, then it encountered the multipart's
|
|
|
410 |
// final boundary.
|
|
|
411 |
int bsize = 0;
|
|
|
412 |
if (m.parseFull(boundary, bsize)) {
|
|
|
413 |
quit = true;
|
|
|
414 |
*boundarysize = bsize;
|
|
|
415 |
}
|
|
|
416 |
|
|
|
417 |
members->push_back(m);
|
|
|
418 |
|
|
|
419 |
} while (!quit);
|
|
|
420 |
}
|
|
|
421 |
|
|
|
422 |
if (!*foundendofpart && !*eof) {
|
|
|
423 |
// multipart parsing starts with skipping to the first
|
|
|
424 |
// boundary. then we call parse() for all parts. the last parse()
|
|
|
425 |
// command will return a code indicating that it found the last
|
|
|
426 |
// boundary of this multipart. Note that the first boundary does
|
|
|
427 |
// not have to start with CRLF.
|
|
|
428 |
string delimiter = "\r\n--" + toboundary;
|
|
|
429 |
|
|
|
430 |
skipUntilBoundary(delimiter, nlines, eof);
|
|
|
431 |
|
|
|
432 |
if (!*eof)
|
|
|
433 |
*boundarysize = delimiter.size();
|
|
|
434 |
|
|
|
435 |
// Read two more characters. This may be CRLF, it may be "--" and
|
|
|
436 |
// it may be any other two characters.
|
|
|
437 |
char a = '\0';
|
|
|
438 |
if (!mimeSource->getChar(&a))
|
|
|
439 |
*eof = true;
|
|
|
440 |
|
|
|
441 |
if (a == '\n')
|
|
|
442 |
++*nlines;
|
|
|
443 |
|
|
|
444 |
char b = '\0';
|
|
|
445 |
if (!mimeSource->getChar(&b))
|
|
|
446 |
*eof = true;
|
|
|
447 |
|
|
|
448 |
if (b == '\n')
|
|
|
449 |
++*nlines;
|
|
|
450 |
|
|
|
451 |
// If we find two dashes after the boundary, then this is the end
|
|
|
452 |
// of boundary marker.
|
|
|
453 |
if (!*eof) {
|
|
|
454 |
if (a == '-' && b == '-') {
|
|
|
455 |
*foundendofpart = true;
|
|
|
456 |
*boundarysize += 2;
|
|
|
457 |
|
|
|
458 |
if (!mimeSource->getChar(&a))
|
|
|
459 |
*eof = true;
|
|
|
460 |
|
|
|
461 |
if (a == '\n')
|
|
|
462 |
++*nlines;
|
|
|
463 |
|
|
|
464 |
if (!mimeSource->getChar(&b))
|
|
|
465 |
*eof = true;
|
|
|
466 |
|
|
|
467 |
if (b == '\n')
|
|
|
468 |
++*nlines;
|
|
|
469 |
}
|
|
|
470 |
|
|
|
471 |
if (a == '\r' && b == '\n') {
|
|
|
472 |
// This exception is to handle a special case where the
|
|
|
473 |
// delimiter of one part is not followed by CRLF, but
|
|
|
474 |
// immediately followed by a CRLF prefixed delimiter.
|
|
|
475 |
if (!mimeSource->getChar(&a) || !mimeSource->getChar(&b))
|
|
|
476 |
*eof = true;
|
|
|
477 |
else if (a == '-' && b == '-') {
|
|
|
478 |
mimeSource->ungetChar();
|
|
|
479 |
mimeSource->ungetChar();
|
|
|
480 |
mimeSource->ungetChar();
|
|
|
481 |
mimeSource->ungetChar();
|
|
|
482 |
} else {
|
|
|
483 |
mimeSource->ungetChar();
|
|
|
484 |
mimeSource->ungetChar();
|
|
|
485 |
}
|
|
|
486 |
|
|
|
487 |
*boundarysize += 2;
|
|
|
488 |
} else {
|
|
|
489 |
mimeSource->ungetChar();
|
|
|
490 |
mimeSource->ungetChar();
|
|
|
491 |
}
|
|
|
492 |
}
|
|
|
493 |
}
|
|
|
494 |
|
|
|
495 |
// make sure bodylength doesn't overflow
|
|
|
496 |
*bodylength = mimeSource->getOffset();
|
|
|
497 |
if (*bodylength >= bodystartoffsetcrlf) {
|
|
|
498 |
*bodylength -= bodystartoffsetcrlf;
|
|
|
499 |
if (*bodylength >= (unsigned int) *boundarysize) {
|
|
|
500 |
*bodylength -= (unsigned int) *boundarysize;
|
|
|
501 |
} else {
|
|
|
502 |
*bodylength = 0;
|
|
|
503 |
}
|
|
|
504 |
} else {
|
|
|
505 |
*bodylength = 0;
|
|
|
506 |
}
|
|
|
507 |
}
|
|
|
508 |
|
|
|
509 |
static void parseSinglePart(const string &toboundary,
|
|
|
510 |
int *boundarysize,
|
|
|
511 |
unsigned int *nbodylines,
|
|
|
512 |
unsigned int *nlines,
|
|
|
513 |
bool *eof, bool *foundendofpart,
|
|
|
514 |
unsigned int *bodylength)
|
|
|
515 |
{
|
|
|
516 |
using namespace ::Binc;
|
|
|
517 |
unsigned int bodystartoffsetcrlf = mimeSource->getOffset();
|
|
|
518 |
|
|
|
519 |
// If toboundary is empty, then we read until the end of the
|
|
|
520 |
// file. Otherwise we will read until we encounter toboundary.
|
|
|
521 |
string _toboundary;
|
|
|
522 |
if (toboundary != "") {
|
|
|
523 |
_toboundary = "\r\n--";
|
|
|
524 |
_toboundary += toboundary;
|
|
|
525 |
}
|
|
|
526 |
|
|
|
527 |
// if (skipUntilBoundary(_toboundary, nlines, eof))
|
|
|
528 |
// *boundarysize = _toboundary.length();
|
|
|
529 |
|
|
|
530 |
char *boundaryqueue = 0;
|
|
|
531 |
int endpos = _toboundary.length();
|
|
|
532 |
if (toboundary != "") {
|
|
|
533 |
boundaryqueue = new char[endpos];
|
|
|
534 |
memset(boundaryqueue, 0, endpos);
|
|
|
535 |
}
|
|
|
536 |
int boundarypos = 0;
|
|
|
537 |
|
|
|
538 |
*boundarysize = 0;
|
|
|
539 |
|
|
|
540 |
const char *_toboundaryStr = _toboundary.c_str();
|
|
|
541 |
string line;
|
|
|
542 |
bool toboundaryIsEmpty = (toboundary == "");
|
|
|
543 |
char c;
|
|
|
544 |
while (mimeSource->getChar(&c)) {
|
|
|
545 |
if (c == '\n') { ++*nbodylines; ++*nlines; }
|
|
|
546 |
|
|
|
547 |
if (toboundaryIsEmpty)
|
|
|
548 |
continue;
|
|
|
549 |
|
|
|
550 |
// find boundary
|
|
|
551 |
boundaryqueue[boundarypos++ % endpos] = c;
|
|
|
552 |
|
|
|
553 |
if (compareStringToQueue(_toboundaryStr, boundaryqueue,
|
|
|
554 |
boundarypos, endpos)) {
|
|
|
555 |
*boundarysize = _toboundary.length();
|
|
|
556 |
break;
|
|
|
557 |
}
|
|
|
558 |
}
|
|
|
559 |
|
|
|
560 |
delete [] boundaryqueue;
|
|
|
561 |
|
|
|
562 |
if (toboundary != "") {
|
|
|
563 |
char a;
|
|
|
564 |
if (!mimeSource->getChar(&a))
|
|
|
565 |
*eof = true;
|
|
|
566 |
|
|
|
567 |
if (a == '\n')
|
|
|
568 |
++*nlines;
|
|
|
569 |
char b;
|
|
|
570 |
if (!mimeSource->getChar(&b))
|
|
|
571 |
*eof = true;
|
|
|
572 |
|
|
|
573 |
if (b == '\n')
|
|
|
574 |
++*nlines;
|
|
|
575 |
|
|
|
576 |
if (a == '-' && b == '-') {
|
|
|
577 |
*boundarysize += 2;
|
|
|
578 |
*foundendofpart = true;
|
|
|
579 |
if (!mimeSource->getChar(&a))
|
|
|
580 |
*eof = true;
|
|
|
581 |
|
|
|
582 |
if (a == '\n')
|
|
|
583 |
++*nlines;
|
|
|
584 |
|
|
|
585 |
if (!mimeSource->getChar(&b))
|
|
|
586 |
*eof = true;
|
|
|
587 |
|
|
|
588 |
if (b == '\n')
|
|
|
589 |
++*nlines;
|
|
|
590 |
}
|
|
|
591 |
|
|
|
592 |
if (a == '\r' && b == '\n') {
|
|
|
593 |
// This exception is to handle a special case where the
|
|
|
594 |
// delimiter of one part is not followed by CRLF, but
|
|
|
595 |
// immediately followed by a CRLF prefixed delimiter.
|
|
|
596 |
if (!mimeSource->getChar(&a) || !mimeSource->getChar(&b))
|
|
|
597 |
*eof = true;
|
|
|
598 |
else if (a == '-' && b == '-') {
|
|
|
599 |
mimeSource->ungetChar();
|
|
|
600 |
mimeSource->ungetChar();
|
|
|
601 |
mimeSource->ungetChar();
|
|
|
602 |
mimeSource->ungetChar();
|
|
|
603 |
} else {
|
|
|
604 |
mimeSource->ungetChar();
|
|
|
605 |
mimeSource->ungetChar();
|
|
|
606 |
}
|
|
|
607 |
|
|
|
608 |
*boundarysize += 2;
|
|
|
609 |
} else {
|
|
|
610 |
mimeSource->ungetChar();
|
|
|
611 |
mimeSource->ungetChar();
|
|
|
612 |
}
|
|
|
613 |
}
|
|
|
614 |
|
|
|
615 |
// make sure bodylength doesn't overflow
|
|
|
616 |
*bodylength = mimeSource->getOffset();
|
|
|
617 |
if (*bodylength >= bodystartoffsetcrlf) {
|
|
|
618 |
*bodylength -= bodystartoffsetcrlf;
|
|
|
619 |
if (*bodylength >= (unsigned int) *boundarysize) {
|
|
|
620 |
*bodylength -= (unsigned int) *boundarysize;
|
|
|
621 |
} else {
|
|
|
622 |
*bodylength = 0;
|
|
|
623 |
}
|
|
|
624 |
} else {
|
|
|
625 |
*bodylength = 0;
|
|
|
626 |
}
|
|
|
627 |
|
|
|
628 |
}
|
|
|
629 |
|
|
|
630 |
//------------------------------------------------------------------------
|
|
|
631 |
int Binc::MimePart::parseFull(const string &toboundary,
|
|
|
632 |
int &boundarysize) const
|
|
|
633 |
{
|
|
|
634 |
headerstartoffsetcrlf = mimeSource->getOffset();
|
|
|
635 |
|
|
|
636 |
// Parse the header of this mime part.
|
|
|
637 |
parseHeader(&h, &nlines);
|
|
|
638 |
|
|
|
639 |
// Headerlength includes the seperating CRLF. Body starts after the
|
|
|
640 |
// CRLF.
|
|
|
641 |
headerlength = mimeSource->getOffset() - headerstartoffsetcrlf;
|
|
|
642 |
bodystartoffsetcrlf = mimeSource->getOffset();
|
|
|
643 |
bodylength = 0;
|
|
|
644 |
|
|
|
645 |
// Determine the type of mime part by looking at fields in the
|
|
|
646 |
// header.
|
|
|
647 |
analyzeHeader(&h, &multipart, &messagerfc822, &subtype, &boundary);
|
|
|
648 |
|
|
|
649 |
bool eof = false;
|
|
|
650 |
bool foundendofpart = false;
|
|
|
651 |
|
|
|
652 |
if (messagerfc822) {
|
|
|
653 |
parseMessageRFC822(&members, &foundendofpart, &bodylength,
|
|
|
654 |
&nbodylines, toboundary);
|
|
|
655 |
|
|
|
656 |
} else if (multipart) {
|
|
|
657 |
parseMultipart(boundary, toboundary, &eof, &nlines, &boundarysize,
|
|
|
658 |
&foundendofpart, &bodylength,
|
|
|
659 |
&members);
|
|
|
660 |
} else {
|
|
|
661 |
parseSinglePart(toboundary, &boundarysize, &nbodylines, &nlines,
|
|
|
662 |
&eof, &foundendofpart, &bodylength);
|
|
|
663 |
}
|
|
|
664 |
|
|
|
665 |
return (eof || foundendofpart) ? 1 : 0;
|
|
|
666 |
}
|