No issues found
1 /*
2 * This program is free software; you can redistribute it and/or
3 * modify it under the terms of the GNU Lesser General Public
4 * License as published by the Free Software Foundation; either
5 * version 2 of the License, or (at your option) version 3.
6 *
7 * This program is distributed in the hope that it will be useful,
8 * but WITHOUT ANY WARRANTY; without even the implied warranty of
9 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
10 * Lesser General Public License for more details.
11 *
12 * You should have received a copy of the GNU Lesser General Public
13 * License along with the program; if not, see <http://www.gnu.org/licenses/>
14 *
15 *
16 * Authors:
17 * Dan Winship <danw@ximian.com>
18 *
19 * Copyright (C) 1999-2008 Novell, Inc. (www.novell.com)
20 *
21 */
22
23 #ifdef HAVE_CONFIG_H
24 #include <config.h>
25 #endif
26
27 #include <ctype.h>
28 #include <stdio.h>
29 #include <string.h>
30
31 #include "e-html-utils.h"
32
33 static gchar *
34 check_size (gchar **buffer,
35 gint *buffer_size,
36 gchar *out,
37 gint len)
38 {
39 if (out + len + 1> *buffer + *buffer_size) {
40 gint index = out - *buffer;
41
42 *buffer_size = MAX (index + len + 1, *buffer_size * 2);
43 *buffer = g_realloc (*buffer, *buffer_size);
44 out = *buffer + index;
45 }
46 return out;
47 }
48
49 /* auto-urlification hints: the goal is not to be strictly RFC-compliant,
50 * but rather to accurately distinguish urls/addresses from non-urls/
51 * addresses in real-world email.
52 *
53 * 1 = non-email-address chars: ()<>@,;:\"[]`'{}|
54 * 2 = trailing url garbage: ,.!?;:>)]}`'-_
55 * 4 = allowed dns chars
56 * 8 = non-url chars: "|
57 */
58 static gint special_chars[] = {
59 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, /* nul - 0x0f */
60 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, /* 0x10 - 0x1f */
61 9, 2, 9, 0, 0, 0, 0, 3, 1, 3, 0, 0, 3, 6, 6, 0, /* sp - / */
62 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 3, 3, 1, 0, 3, 2, /* 0 - ? */
63 1, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, /* @ - O */
64 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 1, 1, 3, 0, 2, /* P - _ */
65 3, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, /* ` - o */
66 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 1, 9, 3, 0, 3 /* p - del */
67 };
68
69 #define is_addr_char(c) (c < 128 && !(special_chars[c] & 1))
70 #define is_url_char(c) (c < 128 && !(special_chars[c] & 8))
71 #define is_trailing_garbage(c) (c > 127 || (special_chars[c] & 2))
72 #define is_domain_name_char(c) (c < 128 && (special_chars[c] & 4))
73
74 /* (http|https|ftp|nntp)://[^ "|/]+\.([^ "|]*[^ ,.!?;:>)\]}`'"|_-])+ */
75 /* www\.[A-Za-z0-9.-]+(/([^ "|]*[^ ,.!?;:>)\]}`'"|_-])+) */
76
77 static gchar *
78 url_extract (const guchar **text,
79 gboolean full_url)
80 {
81 const guchar *end = *text, *p;
82 gchar *out;
83
84 while (*end && is_url_char (*end))
85 end++;
86
87 /* Back up if we probably went too far. */
88 while (end > *text && is_trailing_garbage (*(end - 1)))
89 end--;
90
91 if (full_url) {
92 /* Make sure this really looks like a URL. */
93 p = memchr (*text, ':', end - *text);
94 if (!p || end - p < 4)
95 return NULL;
96 } else {
97 /* Make sure this really looks like a hostname. */
98 p = memchr (*text, '.', end - *text);
99 if (!p || p >= end - 2)
100 return NULL;
101 p = memchr (p + 2, '.', end - (p + 2));
102 if (!p || p >= end - 2)
103 return NULL;
104 }
105
106 out = g_strndup ((gchar *) * text, end - *text);
107 *text = end;
108 return out;
109 }
110
111 static gchar *
112 email_address_extract (const guchar **cur,
113 gchar **out,
114 const guchar *linestart)
115 {
116 const guchar *start, *end, *dot;
117 gchar *addr;
118
119 /* *cur points to the '@'. Look backward for a valid local-part */
120 for (start = *cur; start - 1 >= linestart && is_addr_char (*(start - 1)); start--)
121 ;
122 if (start == *cur)
123 return NULL;
124 if (start > linestart + 2 &&
125 start[-1] == ':' && start[0] == '/' && start[1] == '/')
126 return NULL;
127
128 /* Now look forward for a valid domain part */
129 for (end = *cur + 1, dot = NULL; is_domain_name_char (*end); end++) {
130 if (*end == '.' && !dot)
131 dot = end;
132 }
133 if (!dot)
134 return NULL;
135
136 /* Remove trailing garbage */
137 while (is_trailing_garbage (*(end - 1)))
138 end--;
139 if (dot > end)
140 return NULL;
141
142 addr = g_strndup ((gchar *) start, end - start);
143 *out -= *cur - start;
144 *cur = end;
145
146 return addr;
147 }
148
149 static gboolean
150 is_citation (const guchar *c,
151 gboolean saw_citation)
152 {
153 const guchar *p;
154
155 if (*c != '>')
156 return FALSE;
157
158 /* A line that starts with a ">" is a citation, unless it's
159 * just mbox From-mangling...
160 */
161 if (strncmp ((const gchar *) c, ">From ", 6) != 0)
162 return TRUE;
163
164 /* If the previous line was a citation, then say this
165 * one is too.
166 */
167 if (saw_citation)
168 return TRUE;
169
170 /* Same if the next line is */
171 p = (const guchar *) strchr ((const gchar *) c, '\n');
172 if (p && *++p == '>')
173 return TRUE;
174
175 /* Otherwise, it was just an isolated ">From" line. */
176 return FALSE;
177 }
178
179 /**
180 * e_text_to_html_full:
181 * @input: a NUL-terminated input buffer
182 * @flags: some combination of the E_TEXT_TO_HTML_* flags defined
183 * in e-html-utils.h
184 * @color: color for citation highlighting
185 *
186 * This takes a buffer of text as input and produces a buffer of
187 * "equivalent" HTML, subject to certain transformation rules.
188 *
189 * The set of possible flags is:
190 *
191 * - E_TEXT_TO_HTML_PRE: wrap the output HTML in <PRE> and
192 * </PRE> Should only be used if @input is the entire
193 * buffer to be converted. If e_text_to_html is being called with
194 * small pieces of data, you should wrap the entire result in
195 * <PRE> yourself.
196 *
197 * - E_TEXT_TO_HTML_CONVERT_NL: convert "\n" to "<BR>n" on
198 * output. (Should not be used with E_TEXT_TO_HTML_PRE, since
199 * that would result in double-newlines.)
200 *
201 * - E_TEXT_TO_HTML_CONVERT_SPACES: convert a block of N spaces
202 * into N-1 non-breaking spaces and one normal space. A space
203 * at the start of the buffer is always converted to a
204 * non-breaking space, regardless of the following character,
205 * which probably means you don't want to use this flag on
206 * pieces of data that aren't delimited by at least line breaks.
207 *
208 * If E_TEXT_TO_HTML_CONVERT_NL and E_TEXT_TO_HTML_CONVERT_SPACES
209 * are both defined, then TABs will also be converted to spaces.
210 *
211 * - E_TEXT_TO_HTML_CONVERT_URLS: wrap <a href="..."> </a>
212 * around strings that look like URLs.
213 *
214 * - E_TEXT_TO_HTML_CONVERT_ADDRESSES: wrap <a href="mailto:...">
215 * </a> around strings that look like mail addresses.
216 *
217 * - E_TEXT_TO_HTML_MARK_CITATION: wrap <font color="...">
218 * </font> around citations (lines beginning with "> ", etc).
219 *
220 * - E_TEXT_TO_HTML_ESCAPE_8BIT: flatten everything to US-ASCII
221 *
222 * - E_TEXT_TO_HTML_CITE: quote the text with "> " at the start of each
223 * line.
224 *
225 * Returns: a newly-allocated string containing HTML
226 **/
227 gchar *
228 e_text_to_html_full (const gchar *input,
229 guint flags,
230 guint32 color)
231 {
232 const guchar *cur, *next, *linestart;
233 gchar *buffer = NULL;
234 gchar *out = NULL;
235 gint buffer_size = 0, col;
236 gboolean colored = FALSE, saw_citation = FALSE;
237
238 /* Allocate a translation buffer. */
239 buffer_size = strlen (input) * 2 + 5;
240 buffer = g_malloc (buffer_size);
241
242 out = buffer;
243 if (flags & E_TEXT_TO_HTML_PRE)
244 out += sprintf (out, "<PRE>");
245
246 col = 0;
247
248 for (cur = linestart = (const guchar *) input; cur && *cur; cur = next) {
249 gunichar u;
250
251 if (flags & E_TEXT_TO_HTML_MARK_CITATION && col == 0) {
252 saw_citation = is_citation (cur, saw_citation);
253 if (saw_citation) {
254 if (!colored) {
255 gchar font[25];
256
257 g_snprintf (font, 25, "<FONT COLOR=\"#%06x\">", color);
258
259 out = check_size (&buffer, &buffer_size, out, 25);
260 out += sprintf (out, "%s", font);
261 colored = TRUE;
262 }
263 } else if (colored) {
264 const gchar *no_font = "</FONT>";
265
266 out = check_size (&buffer, &buffer_size, out, 9);
267 out += sprintf (out, "%s", no_font);
268 colored = FALSE;
269 }
270
271 /* Display mbox-mangled ">From" as "From" */
272 if (*cur == '>' && !saw_citation)
273 cur++;
274 } else if (flags & E_TEXT_TO_HTML_CITE && col == 0) {
275 out = check_size (&buffer, &buffer_size, out, 5);
276 out += sprintf (out, "> ");
277 }
278
279 u = g_utf8_get_char ((gchar *) cur);
280 if (g_unichar_isalpha (u) &&
281 (flags & E_TEXT_TO_HTML_CONVERT_URLS)) {
282 gchar *tmpurl = NULL, *refurl = NULL, *dispurl = NULL;
283
284 if (!g_ascii_strncasecmp ((gchar *) cur, "http://", 7) ||
285 !g_ascii_strncasecmp ((gchar *) cur, "https://", 8) ||
286 !g_ascii_strncasecmp ((gchar *) cur, "ftp://", 6) ||
287 !g_ascii_strncasecmp ((gchar *) cur, "nntp://", 7) ||
288 !g_ascii_strncasecmp ((gchar *) cur, "mailto:", 7) ||
289 !g_ascii_strncasecmp ((gchar *) cur, "news:", 5) ||
290 !g_ascii_strncasecmp ((gchar *) cur, "file:", 5) ||
291 !g_ascii_strncasecmp ((gchar *) cur, "callto:", 7) ||
292 !g_ascii_strncasecmp ((gchar *) cur, "h323:", 5) ||
293 !g_ascii_strncasecmp ((gchar *) cur, "sip:", 4) ||
294 !g_ascii_strncasecmp ((gchar *) cur, "webcal:", 7)) {
295 tmpurl = url_extract (&cur, TRUE);
296 if (tmpurl) {
297 refurl = e_text_to_html (tmpurl, 0);
298 dispurl = g_strdup (refurl);
299 }
300 } else if (!g_ascii_strncasecmp ((gchar *) cur, "www.", 4) &&
301 is_url_char (*(cur + 4))) {
302 tmpurl = url_extract (&cur, FALSE);
303 if (tmpurl) {
304 dispurl = e_text_to_html (tmpurl, 0);
305 refurl = g_strdup_printf (
306 "http://%s", dispurl);
307 }
308 }
309
310 if (tmpurl) {
311 out = check_size (
312 &buffer, &buffer_size, out,
313 strlen (refurl) +
314 strlen (dispurl) + 15);
315 out += sprintf (out,
316 "<a href=\"%s\">%s</a>",
317 refurl, dispurl);
318 col += strlen (tmpurl);
319 g_free (tmpurl);
320 g_free (refurl);
321 g_free (dispurl);
322 }
323
324 if (!*cur)
325 break;
326 u = g_utf8_get_char ((gchar *) cur);
327 }
328
329 if (u == '@' && (flags & E_TEXT_TO_HTML_CONVERT_ADDRESSES)) {
330 gchar *addr, *dispaddr, *outaddr;
331
332 addr = email_address_extract (&cur, &out, linestart);
333 if (addr) {
334 dispaddr = e_text_to_html (addr, 0);
335 outaddr = g_strdup_printf (
336 "<a href=\"mailto:%s\">%s</a>",
337 addr, dispaddr);
338 out = check_size (&buffer, &buffer_size, out, strlen (outaddr));
339 out += sprintf (out, "%s", outaddr);
340 col += strlen (addr);
341 g_free (addr);
342 g_free (dispaddr);
343 g_free (outaddr);
344
345 if (!*cur)
346 break;
347 u = g_utf8_get_char ((gchar *) cur);
348 }
349 }
350
351 if (!g_unichar_validate (u)) {
352 /* Sigh. Someone sent undeclared 8-bit data.
353 * Assume it's iso-8859-1.
354 */
355 u = *cur;
356 next = cur + 1;
357 } else
358 next = (const guchar *) g_utf8_next_char (cur);
359
360 out = check_size (&buffer, &buffer_size, out, 10);
361
362 switch (u) {
363 case '<':
364 strcpy (out, "<");
365 out += 4;
366 col++;
367 break;
368
369 case '>':
370 strcpy (out, ">");
371 out += 4;
372 col++;
373 break;
374
375 case '&':
376 strcpy (out, "&");
377 out += 5;
378 col++;
379 break;
380
381 case '"':
382 strcpy (out, """);
383 out += 6;
384 col++;
385 break;
386
387 case '\n':
388 if (flags & E_TEXT_TO_HTML_CONVERT_NL) {
389 strcpy (out, "<br>");
390 out += 4;
391 }
392 *out++ = *cur;
393 linestart = cur;
394 col = 0;
395 break;
396
397 case '\t':
398 if (flags & (E_TEXT_TO_HTML_CONVERT_SPACES |
399 E_TEXT_TO_HTML_CONVERT_NL)) {
400 do {
401 out = check_size (
402 &buffer, &buffer_size, out, 7);
403 strcpy (out, " ");
404 out += 6;
405 col++;
406 } while (col % 8);
407 break;
408 }
409 /* otherwise, FALL THROUGH */
410
411 case ' ':
412 if (flags & E_TEXT_TO_HTML_CONVERT_SPACES) {
413 if (cur == (const guchar *) input ||
414 *(cur + 1) == ' ' || *(cur + 1) == '\t' ||
415 *(cur - 1) == '\n') {
416 strcpy (out, " ");
417 out += 6;
418 col++;
419 break;
420 }
421 }
422 /* otherwise, FALL THROUGH */
423
424 default:
425 if ((u >= 0x20 && u < 0x80) ||
426 (u == '\r' || u == '\t')) {
427 /* Default case, just copy. */
428 *out++ = u;
429 } else {
430 if (flags & E_TEXT_TO_HTML_ESCAPE_8BIT)
431 *out++ = '?';
432 else
433 out += g_snprintf (out, 9, "&#%d;", u);
434 }
435 col++;
436 break;
437 }
438 }
439
440 out = check_size (&buffer, &buffer_size, out, 7);
441 if (flags & E_TEXT_TO_HTML_PRE)
442 strcpy (out, "</PRE>");
443 else
444 *out = '\0';
445
446 return buffer;
447 }
448
449 gchar *
450 e_text_to_html (const gchar *input,
451 guint flags)
452 {
453 return e_text_to_html_full (input, flags, 0);
454 }
455
456 #ifdef E_HTML_UTILS_TEST
457
458 struct {
459 gchar *text, *url;
460 } url_tests[] = {
461 { "bob@foo.com", "mailto:bob@foo.com" },
462 { "Ends with bob@foo.com", "mailto:bob@foo.com" },
463 { "bob@foo.com at start", "mailto:bob@foo.com" },
464 { "bob@foo.com.", "mailto:bob@foo.com" },
465 { "\"bob@foo.com\"", "mailto:bob@foo.com" },
466 { "<bob@foo.com>", "mailto:bob@foo.com" },
467 { "(bob@foo.com)", "mailto:bob@foo.com" },
468 { "bob@foo.com, 555-9999", "mailto:bob@foo.com" },
469 { "|bob@foo.com|555-9999|", "mailto:bob@foo.com" },
470 { "bob@ no match bob@", NULL },
471 { "@foo.com no match @foo.com", NULL },
472 { "\"bob\"@foo.com", NULL },
473 { "M@ke money fast!", NULL },
474 { "ASCII art @_@ @>->-", NULL },
475
476 { "http://www.foo.com", "http://www.foo.com" },
477 { "Ends with http://www.foo.com", "http://www.foo.com" },
478 { "http://www.foo.com at start", "http://www.foo.com" },
479 { "http://www.foo.com.", "http://www.foo.com" },
480 { "http://www.foo.com/.", "http://www.foo.com/" },
481 { "<http://www.foo.com>", "http://www.foo.com" },
482 { "(http://www.foo.com)", "http://www.foo.com" },
483 { "http://www.foo.com, 555-9999", "http://www.foo.com" },
484 { "|http://www.foo.com|555-9999|", "http://www.foo.com" },
485 { "foo http://www.foo.com/ bar", "http://www.foo.com/" },
486 { "foo http://www.foo.com/index.html bar",
487 "http://www.foo.com/index.html" },
488 { "foo http://www.foo.com/q?99 bar", "http://www.foo.com/q?99" },
489 { "foo http://www.foo.com/;foo=bar&baz=quux bar",
490 "http://www.foo.com/;foo=bar&baz=quux" },
491 { "foo http://www.foo.com/index.html#anchor bar",
492 "http://www.foo.com/index.html#anchor" },
493 { "http://www.foo.com/index.html; foo",
494 "http://www.foo.com/index.html" },
495 { "http://www.foo.com/index.html: foo",
496 "http://www.foo.com/index.html" },
497 { "http://www.foo.com/index.html-- foo",
498 "http://www.foo.com/index.html" },
499 { "http://www.foo.com/index.html?",
500 "http://www.foo.com/index.html" },
501 { "http://www.foo.com/index.html!",
502 "http://www.foo.com/index.html" },
503 { "\"http://www.foo.com/index.html\"",
504 "http://www.foo.com/index.html" },
505 { "'http://www.foo.com/index.html'",
506 "http://www.foo.com/index.html" },
507 { "http://bob@www.foo.com/bar/baz/",
508 "http://bob@www.foo.com/bar/baz/" },
509 { "http no match http", NULL },
510 { "http: no match http:", NULL },
511 { "http:// no match http://", NULL },
512 { "unrecognized://bob@foo.com/path", NULL },
513
514 { "src/www.c", NULL },
515 { "Ewwwwww.Gross.", NULL },
516
517 };
518 gint num_url_tests = G_N_ELEMENTS (url_tests);
519
520 gint
521 main (gint argc,
522 gchar **argv)
523 {
524 gint i, errors = 0;
525 gchar *html, *url, *p;
526
527 for (i = 0; i < num_url_tests; i++) {
528 html = e_text_to_html (
529 url_tests[i].text,
530 E_TEXT_TO_HTML_CONVERT_URLS |
531 E_TEXT_TO_HTML_CONVERT_ADDRESSES);
532
533 url = strstr (html, "href=\"");
534 if (url) {
535 url += 6;
536 p = strchr (url, '"');
537 if (p)
538 *p = '\0';
539
540 while ((p = strstr (url, "&")))
541 memmove (p + 1, p + 5, strlen (p + 5) + 1);
542 }
543
544 if ((url && (!url_tests[i].url || strcmp (url, url_tests[i].url) != 0)) ||
545 (!url && url_tests[i].url)) {
546 printf (
547 "FAILED on \"%s\" -> %s\n (got %s)\n\n",
548 url_tests[i].text,
549 url_tests[i].url ? url_tests[i].url : "(nothing)",
550 url ? url : "(nothing)");
551 errors++;
552 }
553
554 g_free (html);
555 }
556
557 printf ("\n%d errors\n", errors);
558 return errors;
559 }
560 #endif