"Fossies" - the Fresh Open Source Software Archive 
Member "tin-2.4.2/src/cook.c" (8 Dec 2017, 28292 Bytes) of package /linux/misc/tin-2.4.2.tar.xz:
As a special service "Fossies" has tried to format the requested source page into HTML format using (guessed) C and C++ source code syntax highlighting (style:
standard) with prefixed line numbers and
code folding option.
Alternatively you can here
view or
download the uninterpreted source code file.
For more information about "cook.c" see the
Fossies "Dox" file reference documentation and the latest
Fossies "Diffs" side-by-side code changes report:
2.4.1_vs_2.4.2.
1 /*
2 * Project : tin - a Usenet reader
3 * Module : cook.c
4 * Author : J. Faultless
5 * Created : 2000-03-08
6 * Updated : 2016-02-28
7 * Notes : Split from page.c
8 *
9 * Copyright (c) 2000-2018 Jason Faultless <jason@altarstone.com>
10 * All rights reserved.
11 *
12 * Redistribution and use in source and binary forms, with or without
13 * modification, are permitted provided that the following conditions
14 * are met:
15 * 1. Redistributions of source code must retain the above copyright
16 * notice, this list of conditions and the following disclaimer.
17 * 2. Redistributions in binary form must reproduce the above copyright
18 * notice, this list of conditions and the following disclaimer in the
19 * documentation and/or other materials provided with the distribution.
20 * 3. The name of the author may not be used to endorse or promote
21 * products derived from this software without specific prior written
22 * permission.
23 *
24 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS
25 * OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
26 * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
27 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY
28 * DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
29 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE
30 * GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
31 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
32 * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
33 * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
34 * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
35 */
36
37 #ifndef TIN_H
38 # include "tin.h"
39 #endif /* !TIN_H */
40 #ifndef TCURSES_H
41 # include "tcurses.h"
42 #endif /* !TCURSES_H */
43
44
45 /*
46 * We malloc() this many t_lineinfo's at a time
47 */
48 #define CHUNK 50
49
50 #define STRIP_ALTERNATIVE(x) \
51 (curr_group->attribute->alternative_handling && \
52 (x)->hdr.ext->type == TYPE_MULTIPART && \
53 strcasecmp("alternative", (x)->hdr.ext->subtype) == 0)
54
55 #define MATCH_REGEX(x,y,z) (pcre_exec(x.re, x.extra, y, z, 0, 0, NULL, 0) >= 0)
56
57
58 static t_bool charset_unsupported(const char *charset);
59 static t_bool header_wanted(const char *line);
60 static t_part *new_uue(t_part **part, char *name);
61 static void process_text_body_part(t_bool wrap_lines, FILE *in, t_part *part, int hide_uue);
62 static void put_cooked(size_t buf_len, t_bool wrap_lines, int flags, const char *fmt, ...);
63 #if defined(MULTIBYTE_ABLE) && !defined(NO_LOCALE)
64 static t_bool wexpand_ctrl_chars(wchar_t **wline, size_t *length, size_t lcook_width);
65 #endif /* MULTIBYTE_ABLE && !NO_LOCALE */
66 #ifdef DEBUG_ART
67 static void dump_cooked(void);
68 #endif /* DEBUG_ART */
69
70
71 /*
72 * These are used globally within this module for access to the context
73 * currently being built. They must not leak outside.
74 */
75 static t_openartinfo *art;
76
77
78 /*
79 * Handle backspace, expand tabs, expand control chars to a literal ^[A-Z]
80 * Allows \n through
81 * Return TRUE if line contains a ^L (form-feed)
82 */
83 t_bool
84 expand_ctrl_chars(
85 char **line,
86 size_t *length,
87 size_t lcook_width)
88 {
89 t_bool ctrl_L = FALSE;
90 #if defined(MULTIBYTE_ABLE) && !defined(NO_LOCALE)
91 wchar_t *wline = char2wchar_t(*line);
92 size_t wlen;
93
94 /*
95 * remove the assert() before release
96 * it should help us find problems with wide-char strings
97 * in the development branch
98 */
99 assert(wline != NULL);
100 wlen = wcslen(wline);
101 ctrl_L = wexpand_ctrl_chars(&wline, &wlen, lcook_width);
102 free(*line);
103 *line = wchar_t2char(wline);
104 free(wline);
105 assert(line != NULL);
106 *length = strlen(*line);
107 #else
108 int curr_len = LEN;
109 unsigned int i = 0, j, ln = 0;
110 char *buf = my_malloc(curr_len);
111 unsigned char *c;
112
113 c = (unsigned char *) *line;
114 while (*c) {
115 if (i > curr_len - (lcook_width + 1)) {
116 curr_len <<= 1;
117 buf = my_realloc(buf, curr_len);
118 }
119 if (*c == '\n')
120 ln = i + 1;
121 if (*c == '\t') { /* expand tabs */
122 j = i + lcook_width - ((i - ln) % lcook_width);
123 for (; i < j; i++)
124 buf[i] = ' ';
125 } else if (((*c) & 0xFF) < ' ' && *c != '\n' && (!IS_LOCAL_CHARSET("Big5") || *c != 27)) { /* literal ctrl chars */
126 buf[i++] = '^';
127 buf[i++] = ((*c) & 0xFF) + '@';
128 if (*c == '\f') /* ^L detected */
129 ctrl_L = TRUE;
130 } else {
131 if (!my_isprint(*c) && *c != '\n')
132 buf[i++] = '?';
133 else
134 buf[i++] = *c;
135 }
136 c++;
137 }
138 buf[i] = '\0';
139 *length = i + 1;
140 *line = my_realloc(*line, *length);
141 strcpy(*line, buf);
142 free(buf);
143 #endif /* MULTIBYTE_ABLE && !NO_LOCALE */
144 return ctrl_L;
145 }
146
147
148 #if defined(MULTIBYTE_ABLE) && !defined(NO_LOCALE)
149 static t_bool
150 wexpand_ctrl_chars(
151 wchar_t **wline,
152 size_t *length,
153 size_t lcook_width)
154 {
155 size_t cur_len = LEN, i = 0, j, ln = 0;
156 wchar_t *wbuf = my_malloc(cur_len * sizeof(wchar_t));
157 wchar_t *wc;
158 t_bool ctrl_L = FALSE;
159
160 wc = *wline;
161 while (*wc) {
162 if (i > cur_len - (lcook_width + 1)) {
163 cur_len <<= 1;
164 wbuf = my_realloc(wbuf, cur_len * sizeof(wchar_t));
165 }
166 if (*wc == '\n')
167 ln = i + 1;
168 if (*wc == '\t') { /* expand_tabs */
169 j = i + lcook_width - ((i - ln) % lcook_width);
170 for (; i < j; i++)
171 wbuf[i] = ' ';
172 } else if (*wc < ' ' && *wc != '\n' && (!IS_LOCAL_CHARSET("Big5") || *wc != 27)) { /* literal ctrl chars */
173 wbuf[i++] = '^';
174 wbuf[i++] = *wc + '@';
175 if (*wc == '\f') /* ^L detected */
176 ctrl_L = TRUE;
177 } else {
178 if (!iswprint((wint_t) *wc) && *wc != '\n')
179 wbuf[i++] = '?';
180 else
181 wbuf[i++] = *wc;
182 }
183 wc++;
184 }
185 wbuf[i] = '\0';
186 *length = i + 1;
187 *wline = my_realloc(*wline, *length * sizeof(wchar_t));
188 wcscpy(*wline, wbuf);
189 free(wbuf);
190 return ctrl_L;
191 }
192 #endif /* MULTIBYTE_ABLE && !NO_LOCALE */
193
194
195 /*
196 * Output text to the cooked stream. Wrap lines as necessary.
197 * Update the line count and the array of line offsets
198 * Extend the lineoffset array as needed in CHUNK amounts.
199 * flags are 'hints' to the pager about line content.
200 * buf_len is the size put_cooked should use for its buffer.
201 */
202 static void
203 put_cooked(
204 size_t buf_len,
205 t_bool wrap_lines,
206 int flags,
207 const char *fmt,
208 ...)
209 {
210 char *p, *bufp, *buf;
211 int wrap_column;
212 int space;
213 static int saved_flags = 0;
214 va_list ap;
215 #if defined(MULTIBYTE_ABLE) && !defined(NO_LOCALE)
216 int bytes;
217 wint_t *wp;
218 #endif /* MULTIBYTE_ABLE && !NO_LOCALE */
219
220 buf = my_malloc(buf_len + 1);
221
222 va_start(ap, fmt);
223 vsnprintf(buf, buf_len + 1, fmt, ap);
224
225 if (tinrc.wrap_column < 0)
226 wrap_column = ((tinrc.wrap_column > -cCOLS) ? cCOLS + tinrc.wrap_column : cCOLS);
227 else
228 #if 1
229 wrap_column = ((tinrc.wrap_column > 0) ? tinrc.wrap_column : cCOLS);
230 #else /* never cut off long lines */
231 wrap_column = (((tinrc.wrap_column > 0) && (tinrc.wrap_column < cCOLS)) ? tinrc.wrap_column : cCOLS);
232 #endif /* 1 */
233
234 p = bufp = buf;
235
236 #if defined(MULTIBYTE_ABLE) && !defined(NO_LOCALE)
237 wp = my_malloc((MB_CUR_MAX + 1) * sizeof(wint_t));
238 #endif /* MULTIBYTE_ABLE && !NO_LOCALE */
239
240 while (*p) {
241 if (wrap_lines) {
242 space = wrap_column;
243 while (space > 0 && *p && *p != '\n') {
244 #if defined(MULTIBYTE_ABLE) && !defined(NO_LOCALE)
245 if ((bytes = mbtowc((wchar_t *) wp, p, MB_CUR_MAX)) > 0) {
246 if ((space -= wcwidth(*wp)) < 0)
247 break;
248 p += bytes;
249 } else
250 p++;
251 #else
252 p++;
253 space--;
254 #endif /* MULTIBYTE_ABLE && !NO_LOCALE */
255 }
256 } else {
257 while (*p && *p != '\n')
258 p++;
259 }
260 fwrite(bufp, 1, p - bufp, art->cooked);
261 fputs("\n", art->cooked);
262 if (*p == '\n')
263 p++;
264 bufp = p;
265
266 if (art->cooked_lines == 0) {
267 art->cookl = my_malloc(sizeof(t_lineinfo) * CHUNK);
268 art->cookl[0].offset = 0;
269 }
270
271 /*
272 * Pick up flags from a previous partial write
273 */
274 art->cookl[art->cooked_lines].flags = flags | saved_flags;
275 saved_flags = 0;
276 art->cooked_lines++;
277
278 /*
279 * Grow the array of lines if needed - we resize it properly at the end
280 */
281 if (art->cooked_lines % CHUNK == 0)
282 art->cookl = my_realloc(art->cookl, sizeof(t_lineinfo) * CHUNK * ((art->cooked_lines / CHUNK) + 1));
283
284 art->cookl[art->cooked_lines].offset = ftell(art->cooked);
285 }
286
287 #if defined(MULTIBYTE_ABLE) && !defined(NO_LOCALE)
288 free(wp);
289 #endif /* MULTIBYTE_ABLE && !NO_LOCALE */
290
291 /*
292 * If there is anything left over, then it must be a non \n terminated
293 * partial line from base64 decoding etc.. Dump it now and the rest of
294 * the line (with the \n) will fill in the t_lineinfo
295 * We must save the flags now as the rest of the line may not have the same properties
296 * We need to keep the length for accounting purposes
297 */
298 if (*bufp != '\0') {
299 fputs(bufp, art->cooked);
300 saved_flags = flags;
301 }
302
303 va_end(ap);
304 free(buf);
305 }
306
307
308 /*
309 * Add a new uuencode attachment description to the current part
310 */
311 static t_part *
312 new_uue(
313 t_part **part,
314 char *name)
315 {
316 t_part *ptr = new_part((*part)->uue);
317
318 if (!(*part)->uue) /* new_part() is simple and doesn't attach list heads */
319 (*part)->uue = ptr;
320
321 free_list(ptr->params);
322 /*
323 * Load the name into the parameter list
324 */
325 ptr->params = new_params();
326 ptr->params->name = my_strdup("name");
327 ptr->params->value = my_strdup(str_trim(name));
328
329 ptr->encoding = ENCODING_UUE; /* treat as x-uuencode */
330
331 ptr->offset = ftell(art->cooked);
332 ptr->depth = (*part)->depth; /* uue is at the same depth as the envelope */
333
334 /*
335 * If an extension is present, try and add a Content-Type
336 */
337 if ((name = strrchr(name, '.')) != NULL)
338 lookup_mimetype(name + 1, ptr);
339
340 return ptr;
341 }
342
343
344 /*
345 * Get the suggested filename for an attachment. RFC says Content-Disposition
346 * 'filename' supersedes Content-Type 'name'. We must also remove path
347 * information.
348 */
349 const char *
350 get_filename(
351 t_param *ptr)
352 {
353 const char *name;
354 char *p;
355
356 if (!(name = get_param(ptr, "filename"))) {
357 if (!(name = get_param(ptr, "name")))
358 return NULL;
359 }
360
361 if ((p = strrchr(name, DIRSEP)))
362 return p + 1;
363
364 return name;
365 }
366
367
368 #define PUT_UUE(part, qualifier_text) \
369 put_cooked(LEN, wrap_lines, C_UUE, _(txt_uue), \
370 part->depth ? (part->depth - 1) * 4 : 0, "", \
371 content_types[part->type], part->subtype, \
372 qualifier_text, part->line_count, get_filename(part->params))
373
374 #define PUT_ATTACH(part, depth, name, charset) \
375 put_cooked(LEN, wrap_lines, C_ATTACH, _(txt_attach), \
376 depth, "", \
377 content_types[part->type], part->subtype, \
378 content_encodings[part->encoding], \
379 charset ? _(txt_attach_charset) : "", BlankIfNull(charset), \
380 part->line_count, \
381 name ? _(txt_name) : "", BlankIfNull(name)); \
382 \
383 if (part->description) \
384 put_cooked(LEN, wrap_lines, C_ATTACH, \
385 _(txt_attach_description), \
386 depth, "", \
387 part->description); \
388 if (part->next != NULL || IS_PLAINTEXT(part)) \
389 put_cooked(1, wrap_lines, C_ATTACH, "\n")
390
391 /*
392 * Decodes text bodies, remove sig's, detects uuencoded sections
393 */
394 static void
395 process_text_body_part(
396 t_bool wrap_lines,
397 FILE *in,
398 t_part *part,
399 int hide_uue)
400 {
401 char *rest = NULL;
402 char *line = NULL, *buf, *tmpline;
403 const char *ncharset;
404 size_t max_line_len = 0;
405 int flags, len, lines_left, len_blank;
406 int offsets[6];
407 int size_offsets = ARRAY_SIZE(offsets);
408 unsigned int lines_skipped = 0;
409 t_bool in_sig = FALSE; /* Set when in sig portion */
410 t_bool in_uue = FALSE; /* Set when in uuencoded section */
411 t_bool in_verbatim = FALSE; /* Set when in verbatim section */
412 t_bool verbatim_begin = FALSE; /* Set when verbatim_begin_regex matches */
413 t_bool is_uubody; /* Set when current line looks like a uuencoded line */
414 t_bool first_line_blank = TRUE; /* Unset when first non-blank line is reached */
415 t_bool put_blank_lines = FALSE; /* Set when previously skipped lines needs to put */
416 t_part *curruue = NULL;
417
418 if (part->uue) { /* These are redone each time we recook/resize etc.. */
419 free_parts(part->uue);
420 part->uue = NULL;
421 }
422
423 fseek(in, part->offset, SEEK_SET);
424
425 if (part->encoding == ENCODING_BASE64)
426 (void) mmdecode(NULL, 'b', 0, NULL); /* flush */
427
428 lines_left = part->line_count;
429 while ((lines_left > 0) || rest) {
430 switch (part->encoding) {
431 case ENCODING_BASE64:
432 lines_left -= read_decoded_base64_line(in, &line, &max_line_len, lines_left, &rest);
433 break;
434
435 case ENCODING_QP:
436 lines_left -= read_decoded_qp_line(in, &line, &max_line_len, lines_left);
437 break;
438
439 default:
440 if ((buf = tin_fgets(in, FALSE)) == NULL) {
441 FreeAndNull(line);
442 break;
443 }
444
445 /*
446 * tin_fgets() uses the returned space also internally
447 * so it's not advisable to use it for our own purposes
448 * especially if we must resize it.
449 * So copy buf to line (and resize line if necessary).
450 */
451 if (max_line_len < strlen(buf) + 2) {
452 max_line_len = strlen(buf) + 2;
453 line = my_realloc(line, max_line_len);
454 }
455 strcpy(line, buf);
456
457 /*
458 * FIXME: Some code in cook.c expects a '\n' at the end
459 * of the line. As tin_fgets() strips trailing '\n', re-add it.
460 * This should problably be fixed in that other code.
461 */
462 strcat(line, "\n");
463
464 lines_left--;
465 break;
466 }
467 if (!(line && strlen(line))) {
468 FreeIfNeeded(rest);
469 break; /* premature end of file, file error etc. */
470 }
471
472 /* convert network to local charset, tex2iso, iso2asc etc. */
473 ncharset = get_param(part->params, "charset");
474 process_charsets(&line, &max_line_len, ncharset ? ncharset : "US-ASCII", tinrc.mm_local_charset, curr_group->attribute->tex2iso_conv && art->tex2iso);
475
476 #if defined(MULTIBYTE_ABLE) && !defined(NO_LOCALE)
477 if (IS_LOCAL_CHARSET("UTF-8"))
478 utf8_valid(line);
479 #endif /* MULTIBYTE_ABLE && !NO_LOCALE */
480
481 len = (int) strlen(line);
482
483 /*
484 * trim article body and sig (not verbatim blocks):
485 * - skip leading blank lines
486 * - replace multiple blank lines with one empty line
487 * - skip tailing blank lines, keep one if an
488 * attachment follows
489 */
490 if (curr_group->attribute->trim_article_body && !in_uue && !in_verbatim && !verbatim_begin) {
491 len_blank = 1;
492 tmpline = line;
493 /* check if line contains only whitespace */
494 while ((' ' == *tmpline) || ('\t' == *tmpline)) {
495 len_blank++;
496 tmpline++;
497 }
498 if (len_blank == len) { /* line is blank */
499 if (lines_left == 0 && (curr_group->attribute->trim_article_body & SKIP_TRAILING)) {
500 if (!(part->next == NULL || (STRIP_ALTERNATIVE(art) && !IS_PLAINTEXT(part->next))))
501 put_cooked(1, TRUE, in_sig ? C_SIG : C_BODY, "\n");
502 continue;
503 }
504 if (first_line_blank) {
505 if (curr_group->attribute->trim_article_body & SKIP_LEADING)
506 continue;
507 } else if ((curr_group->attribute->trim_article_body & (COMPACT_MULTIPLE | SKIP_TRAILING)) && (!in_sig || curr_group->attribute->show_signatures)) {
508 lines_skipped++;
509 if (lines_left == 0 && !(curr_group->attribute->trim_article_body & SKIP_TRAILING)) {
510 for (; lines_skipped > 0; lines_skipped--)
511 put_cooked(1, TRUE, in_sig ? C_SIG : C_BODY, "\n");
512 }
513 continue;
514 }
515 } else { /* line is not blank */
516 if (first_line_blank)
517 first_line_blank = FALSE;
518 if (lines_skipped && (!in_sig || curr_group->attribute->show_signatures)) {
519 if (strcmp(line, SIGDASHES) != 0 || curr_group->attribute->show_signatures) {
520 if (curr_group->attribute->trim_article_body & COMPACT_MULTIPLE)
521 put_cooked(1, TRUE, in_sig ? C_SIG : C_BODY, "\n");
522 else
523 put_blank_lines = TRUE;
524 } else if (!(curr_group->attribute->trim_article_body & SKIP_TRAILING))
525 put_blank_lines = TRUE;
526 if (put_blank_lines) {
527 for (; lines_skipped > 0; lines_skipped--)
528 put_cooked(1, TRUE, in_sig ? C_SIG : C_BODY, "\n");
529 }
530 put_blank_lines = FALSE;
531 lines_skipped = 0;
532 }
533 }
534 } /* if (tinrc.trim_article_body...) */
535
536 /* look for verbatim marks, set in_verbatim only for lines in between */
537 if (curr_group->attribute->verbatim_handling) {
538 if (verbatim_begin) {
539 in_verbatim = TRUE;
540 verbatim_begin = FALSE;
541 } else if (!in_sig && !in_uue && !in_verbatim && MATCH_REGEX(verbatim_begin_regex, line, len))
542 verbatim_begin = TRUE;
543 if (in_verbatim && MATCH_REGEX(verbatim_end_regex, line, len))
544 in_verbatim = FALSE;
545 }
546
547 if (!in_verbatim) {
548 /*
549 * Detect and skip signatures if necessary
550 */
551 if (!in_sig) {
552 if (strcmp(line, SIGDASHES) == 0) {
553 in_sig = TRUE;
554 if (in_uue) {
555 in_uue = FALSE;
556 if (hide_uue)
557 PUT_UUE(curruue, _(txt_incomplete));
558 }
559 }
560 }
561
562 if (in_sig && !(curr_group->attribute->show_signatures))
563 continue; /* No further processing needed */
564
565 /*
566 * Detect and process uuencoded sections
567 * Look for the start or the end of a uuencoded section
568 *
569 * TODO: look for a tailing size line after end (non standard
570 * extension)?
571 */
572 if (pcre_exec(uubegin_regex.re, uubegin_regex.extra, line, len, 0, 0, offsets, size_offsets) != PCRE_ERROR_NOMATCH) {
573 in_uue = TRUE;
574 curruue = new_uue(&part, line + offsets[1]);
575 if (hide_uue)
576 continue; /* Don't cook the 'begin' line */
577 } else if (strncmp(line, "end\n", 4) == 0) {
578 if (in_uue) {
579 in_uue = FALSE;
580 if (hide_uue) {
581 PUT_UUE(curruue, "");
582 continue; /* Don't cook the 'end' line */
583 }
584 }
585 }
586
587 /*
588 * See if this line looks like a uuencoded 'body' line
589 */
590 is_uubody = FALSE;
591
592 if (MATCH_REGEX(uubody_regex, line, len)) {
593 int sum = (((*line) - ' ') & 077) * 4 / 3; /* uuencode octet checksum */
594
595 /* sum = 0 in a uubody only on the last line, a single ` */
596 if (sum == 0 && len == 1 + 1) /* +1 for the \n */
597 is_uubody = TRUE;
598 else if (len == sum + 1 + 1)
599 is_uubody = TRUE;
600 #ifdef DEBUG_ART
601 if (debug & DEBUG_MISC)
602 fprintf(stderr, "%s sum=%d len=%d (%s)\n", bool_unparse(is_uubody), sum, len, line);
603 #endif /* DEBUG_ART */
604 }
605
606 if (in_uue) {
607 if (is_uubody)
608 curruue->line_count++;
609 else {
610 if (line[0] == '\n') { /* Blank line in a uubody - definitely a failure */
611 /* fprintf(stderr, "not a uue line while reading a uue body?\n"); */
612 in_uue = FALSE;
613 if (hide_uue)
614 /* don't continue here, so we see the line that 'broke' in_uue */
615 PUT_UUE(curruue, _(txt_incomplete));
616 }
617 }
618 } else {
619 /*
620 * UUE_ALL = 'Try harder' - we never saw a begin line, but useful
621 * when uue sections are split across > 1 article
622 */
623 if (is_uubody && hide_uue == UUE_ALL) {
624 char name[] = N_("(unknown)");
625
626 curruue = new_uue(&part, name);
627 curruue->line_count++;
628 in_uue = TRUE;
629 continue;
630 }
631 }
632
633 /*
634 * Skip output if we're hiding uue or the sig
635 */
636 if (in_uue && hide_uue)
637 continue; /* No further processing needed */
638 }
639
640 flags = in_verbatim ? C_VERBATIM : in_sig ? C_SIG : C_BODY;
641
642 /*
643 * Don't do any further handling of uue || verbatim lines
644 */
645 if (in_uue) {
646 put_cooked(max_line_len, wrap_lines, flags, "%s", line);
647 continue;
648 } else if (in_verbatim) {
649 expand_ctrl_chars(&line, &max_line_len, 8);
650 put_cooked(max_line_len, wrap_lines, flags, "%s", line);
651 continue;
652 }
653
654 #ifdef HAVE_COLOR
655 /* keep order in sync with color.c:draw_pager_line() */
656 if (quote_regex3.re) {
657 if (MATCH_REGEX(quote_regex3, line, len))
658 flags |= C_QUOTE3;
659 else if (quote_regex2.re) {
660 if (MATCH_REGEX(quote_regex2, line, len))
661 flags |= C_QUOTE2;
662 else if (curr_group->attribute->extquote_handling && extquote_regex.re) {
663 if (MATCH_REGEX(extquote_regex, line, len))
664 flags |= C_EXTQUOTE;
665 else if (quote_regex.re) {
666 if (MATCH_REGEX(quote_regex, line, len))
667 flags |= C_QUOTE1;
668 }
669 } else if (quote_regex.re) {
670 if (MATCH_REGEX(quote_regex, line, len))
671 flags |= C_QUOTE1;
672 }
673 }
674 }
675 #endif /* HAVE_COLOR */
676
677 if (MATCH_REGEX(url_regex, line, len))
678 flags |= C_URL;
679 if (MATCH_REGEX(mail_regex, line, len))
680 flags |= C_MAIL;
681 if (MATCH_REGEX(news_regex, line, len))
682 flags |= C_NEWS;
683
684 if (expand_ctrl_chars(&line, &max_line_len, tabwidth))
685 flags |= C_CTRLL; /* Line contains form-feed */
686
687 buf = line;
688
689 /*
690 * Skip over the first space in case of Format=Flowed (space-stuffing)
691 */
692 if (part->format == FORMAT_FLOWED) {
693 if (line[0] == ' ')
694 ++buf;
695 }
696
697 put_cooked(max_line_len, wrap_lines && (!IS_LOCAL_CHARSET("Big5")), flags, "%s", buf);
698 } /* while */
699
700 /*
701 * Were we reading uue and ran off the end ?
702 */
703 if (in_uue && hide_uue)
704 PUT_UUE(curruue, _(txt_incomplete));
705
706 free(line);
707 }
708
709
710 /*
711 * Return TRUE if this header should be printed as per
712 * news_headers_to_[not_]display
713 */
714 static t_bool
715 header_wanted(
716 const char *line)
717 {
718 int i;
719 t_bool ret = FALSE;
720
721 if (curr_group->attribute->headers_to_display->num && (curr_group->attribute->headers_to_display->header[0][0] == '*'))
722 ret = TRUE; /* wild do */
723 else {
724 for (i = 0; i < curr_group->attribute->headers_to_display->num; i++) {
725 if (!strncasecmp(line, curr_group->attribute->headers_to_display->header[i], strlen(curr_group->attribute->headers_to_display->header[i]))) {
726 ret = TRUE;
727 break;
728 }
729 }
730 }
731
732 if (curr_group->attribute->headers_to_not_display->num && (curr_group->attribute->headers_to_not_display->header[0][0] == '*'))
733 ret = FALSE; /* wild don't: doesn't make sense! */
734 else {
735 for (i = 0; i < curr_group->attribute->headers_to_not_display->num; i++) {
736 if (!strncasecmp(line, curr_group->attribute->headers_to_not_display->header[i], strlen(curr_group->attribute->headers_to_not_display->header[i]))) {
737 ret = FALSE;
738 break;
739 }
740 }
741 }
742
743 return ret;
744 }
745
746
747 /* #define DEBUG_ART 1 */
748 #ifdef DEBUG_ART
749 static void
750 dump_cooked(
751 void)
752 {
753 char *line;
754 int i;
755
756 for (i = 0; i < art->cooked_lines; i++) {
757 fseek(art->cooked, art->cookl[i].offset, SEEK_SET);
758 line = tin_fgets(art->cooked, FALSE);
759 fprintf(stderr, "[%3d] %4ld %3x [%s]\n", i, art->cookl[i].offset, art->cookl[i].flags, line);
760 }
761 fprintf(stderr, "%d lines cooked\n", art->cooked_lines);
762 }
763 #endif /* DEBUG_ART */
764
765
766 /*
767 * Check for charsets which may contain NULL bytes and thus break string
768 * functions. Possibly incomplete.
769 *
770 * TODO: fix the other code to handle those charsets properly.
771 */
772 static t_bool
773 charset_unsupported(
774 const char *charset)
775 {
776 static const char *charsets[] = {
777 "csUnicode", /* alias for ISO-10646-UCS-2 */
778 "csUCS4", /* alias for ISO-10646-UCS-4 */
779 "ISO-10646-UCS-2",
780 "ISO-10646-UCS-4",
781 "UTF-16", /* covers also BE/LE */
782 "UTF-32", /* covers also BE/LE */
783 NULL };
784 const char **charsetptr = charsets;
785 t_bool ret = FALSE;
786
787 if (!charset)
788 return ret;
789
790 do {
791 if (!strncasecmp(charset, *charsetptr, strlen(*charsetptr)))
792 ret = TRUE;
793 } while (!ret && *(++charsetptr) != NULL);
794
795 return ret;
796 }
797
798
799 /*
800 * 'cooks' an article, ie, prepare what will actually appear on the screen
801 * It is not easy to do this in the same pass as the initial read since
802 * boundary conditions for multipart articles make it harder to do on the
803 * fly decoding.
804 * We could have cooked the headers whilst they were being read but we're
805 * trying to keep this simple.
806 *
807 * Expects:
808 * Fresh article context to write into
809 * parse_uue is set only when the art is opened to create t_parts for
810 * uue sections found, when resizing this is not needed
811 * hide_uue determines the folding of uue sections
812 * Handles:
813 * multipart articles
814 * stripping of non text sections if skip_alternative
815 * Q and B decoding of text sections
816 * handling of uuencoded sections
817 * stripping of sigs if !show_signatures
818 * Returns:
819 * TRUE on success
820 *
821 * TODO:
822 * give an error-message on at least disk-full
823 */
824 t_bool
825 cook_article(
826 t_bool wrap_lines,
827 t_openartinfo *artinfo,
828 int hide_uue,
829 t_bool show_all_headers)
830 {
831 const char *charset;
832 const char *name;
833 char *line;
834 struct t_header *hdr = &artinfo->hdr;
835 t_bool header_put = FALSE;
836 static const char *struct_header[] = {
837 "Approved: ", "From: ", "Originator: ",
838 "Reply-To: ", "Sender: ", "X-Cancelled-By: ", "X-Comment-To: ",
839 "X-Submissions-To: ", "To: ", "Cc: ", "Bcc: ", "X-Originator: ", 0 };
840
841 art = artinfo; /* Global saves lots of passing artinfo around */
842
843 if (!(art->cooked = tmpfile()))
844 return FALSE;
845
846 art->cooked_lines = 0;
847
848 rewind(artinfo->raw);
849
850 /*
851 * Put down just the headers we want
852 */
853 while ((line = tin_fgets(artinfo->raw, TRUE)) != NULL) {
854 if (line[0] == '\0') { /* End of headers? */
855 if (STRIP_ALTERNATIVE(artinfo)) {
856 if (header_wanted(_(txt_info_x_conversion_note))) {
857 header_put = TRUE;
858 put_cooked(LEN, wrap_lines, C_HEADER, _(txt_info_x_conversion_note));
859 }
860 }
861 if (header_put)
862 put_cooked(1, TRUE, 0, "\n"); /* put a newline after headers */
863 break;
864 }
865
866 if (show_all_headers || header_wanted(line)) { /* Put cooked data */
867 const char **strptr = struct_header;
868 char *l = NULL, *ptr, *foo, *bar;
869 size_t i = LEN;
870 t_bool found = FALSE;
871
872 /* structured headers */
873 do {
874 if (!strncasecmp(line, *strptr, strlen(*strptr))) {
875 foo = my_strdup(*strptr);
876 if ((ptr = strchr(foo, ':'))) {
877 *ptr = '\0';
878 unfold_header(line);
879 if ((ptr = parse_header(line, foo, TRUE, TRUE, FALSE))) {
880 #if 0
881 /*
882 * TODO:
883 * idna_decode() currently expects just a FQDN
884 * or a mailaddress (with all comments stripped).
885 *
886 * we need to look for something like
887 * (?i)((?:\S+\.)?xn--[a-z0-9\.\-]{3,}\S+)\b
888 * and just decode $1
889 * maybe also in process_text_body_part()
890 */
891 bar = idna_decode(ptr);
892 #else
893 bar = my_strdup(ptr);
894 #endif /* 0 */
895 l = my_calloc(1, strlen(bar) + strlen(*strptr) + 1);
896 strncpy(l, line, strlen(*strptr));
897 strcat(l, bar);
898 free(bar);
899 }
900 }
901 free(foo);
902 found = TRUE;
903 }
904 } while (!found && *(++strptr) != 0);
905
906 /* unstructured but must not be decoded */
907 if (l == NULL && (!strncasecmp(line, "References: ", 12) || !strncasecmp(line, "Message-ID: ", 12) || !strncasecmp(line, "Date: ", 6) || !strncasecmp(line, "Newsgroups: ", 12) || !strncasecmp(line, "Distribution: ", 14) || !strncasecmp(line, "Followup-To: ", 13) || !strncasecmp(line, "X-Face: ", 8) || !strncasecmp(line, "Cancel-Lock: ", 13) || !strncasecmp(line, "Cancel-Key: ", 12) || !strncasecmp(line, "Supersedes: ", 12)))
908 l = my_strdup(line);
909
910 if (l == NULL)
911 l = my_strdup(rfc1522_decode(line));
912
913 #if defined(MULTIBYTE_ABLE) && !defined(NO_LOCALE)
914 if (IS_LOCAL_CHARSET("UTF-8"))
915 utf8_valid(l);
916 #endif /* MULTIBYTE_ABLE && !NO_LOCALE */
917 header_put = TRUE;
918 expand_ctrl_chars(&l, &i, tabwidth);
919 put_cooked(i, wrap_lines, C_HEADER, "%s", l);
920 free(l);
921 }
922 }
923
924 if (tin_errno != 0)
925 return FALSE;
926
927 /*
928 * Process the attachments in turn, print a neato header, and process/decode
929 * the body if of text type
930 */
931 if (hdr->mime && hdr->ext->type == TYPE_MULTIPART) {
932 t_part *ptr;
933
934 for (ptr = hdr->ext->next; ptr != NULL; ptr = ptr->next) {
935 /*
936 * Ignore non text/plain sections with alternative handling
937 */
938 if (STRIP_ALTERNATIVE(artinfo) && !IS_PLAINTEXT(ptr))
939 continue;
940
941 name = get_filename(ptr->params);
942 if (!strcmp(content_types[ptr->type], "text"))
943 charset = get_param(ptr->params, "charset");
944 else
945 charset = NULL;
946 PUT_ATTACH(ptr, (ptr->depth - 1) * 4, name, charset);
947
948 /* Try to view anything of type text, may need to review this */
949 if (IS_PLAINTEXT(ptr)) {
950 if (charset_unsupported(charset)) {
951 put_cooked(LEN, wrap_lines, C_ATTACH, _(txt_attach_unsup_charset), (ptr->depth - 1) * 4, "", charset);
952 if (ptr->next)
953 put_cooked(1, wrap_lines, C_ATTACH, "\n");
954 } else
955 process_text_body_part(wrap_lines, artinfo->raw, ptr, hide_uue);
956 }
957 }
958 } else {
959 if (!strcmp(content_types[hdr->ext->type], "text"))
960 charset = get_param(hdr->ext->params, "charset");
961 else
962 charset = NULL;
963 /*
964 * A regular single-body article
965 */
966 if (IS_PLAINTEXT(hdr->ext)) {
967 if (charset_unsupported(charset))
968 put_cooked(LEN, wrap_lines, C_ATTACH, _(txt_attach_unsup_charset), 0, "", charset);
969 else
970 process_text_body_part(wrap_lines, artinfo->raw, hdr->ext, hide_uue);
971 } else {
972 /*
973 * Non-textual main body
974 */
975 name = get_filename(hdr->ext->params);
976 PUT_ATTACH(hdr->ext, 0, name, charset);
977 }
978 }
979
980 #ifdef DEBUG_ART
981 dump_cooked();
982 #endif /* DEBUG_ART */
983
984 if (art->cooked_lines > 0)
985 art->cookl = my_realloc(art->cookl, sizeof(t_lineinfo) * art->cooked_lines);
986
987 rewind(art->cooked);
988 return (tin_errno != 0) ? FALSE : TRUE;
989 }