1 /* -*- Mode: C; tab-width: 8; indent-tabs-mode: t; c-basic-offset: 8 -*- */
2 /* GMime
3 * Copyright (C) 2000-2014 Jeffrey Stedfast
4 *
5 * This library is free software; you can redistribute it and/or
6 * modify it under the terms of the GNU Lesser General Public License
7 * as published by the Free Software Foundation; either version 2.1
8 * of the License, or (at your option) any later version.
9 *
10 * This library is distributed in the hope that it will be useful,
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
13 * Lesser General Public License for more details.
14 *
15 * You should have received a copy of the GNU Lesser General Public
16 * License along with this library; if not, write to the Free
17 * Software Foundation, 51 Franklin Street, Fifth Floor, Boston, MA
18 * 02110-1301, USA.
19 */
20
21
22 #ifdef HAVE_CONFIG_H
23 #include <config.h>
24 #endif
25
26 #include <string.h>
27
28 #include "gmime-filter-best.h"
29
30
31 /**
32 * SECTION: gmime-filter-best
33 * @title: GMimeFilterBest
34 * @short_description: Determine the best charset/encoding to use for a stream
35 * @see_also: #GMimeFilter
36 *
37 * A #GMimeFilter which is meant to determine the best charset and/or
38 * transfer encoding suitable for the stream which is filtered through
39 * it.
40 **/
41
42
43 static void g_mime_filter_best_class_init (GMimeFilterBestClass *klass);
44 static void g_mime_filter_best_init (GMimeFilterBest *filter, GMimeFilterBestClass *klass);
45 static void g_mime_filter_best_finalize (GObject *object);
46
47 static GMimeFilter *filter_copy (GMimeFilter *filter);
48 static void filter_filter (GMimeFilter *filter, char *in, size_t len, size_t prespace,
49 char **out, size_t *outlen, size_t *outprespace);
50 static void filter_complete (GMimeFilter *filter, char *in, size_t len, size_t prespace,
51 char **out, size_t *outlen, size_t *outprespace);
52 static void filter_reset (GMimeFilter *filter);
53
54
55 static GMimeFilterClass *parent_class = NULL;
56
57
58 GType
g_mime_filter_best_get_type(void)59 g_mime_filter_best_get_type (void)
60 {
61 static GType type = 0;
62
63 if (!type) {
64 static const GTypeInfo info = {
65 sizeof (GMimeFilterBestClass),
66 NULL, /* base_class_init */
67 NULL, /* base_class_finalize */
68 (GClassInitFunc) g_mime_filter_best_class_init,
69 NULL, /* class_finalize */
70 NULL, /* class_data */
71 sizeof (GMimeFilterBest),
72 0, /* n_preallocs */
73 (GInstanceInitFunc) g_mime_filter_best_init,
74 };
75
76 type = g_type_register_static (GMIME_TYPE_FILTER, "GMimeFilterBest", &info, 0);
77 }
78
79 return type;
80 }
81
82
83 static void
g_mime_filter_best_class_init(GMimeFilterBestClass * klass)84 g_mime_filter_best_class_init (GMimeFilterBestClass *klass)
85 {
86 GObjectClass *object_class = G_OBJECT_CLASS (klass);
87 GMimeFilterClass *filter_class = GMIME_FILTER_CLASS (klass);
88
89 parent_class = g_type_class_ref (GMIME_TYPE_FILTER);
90
91 object_class->finalize = g_mime_filter_best_finalize;
92
93 filter_class->copy = filter_copy;
94 filter_class->filter = filter_filter;
95 filter_class->complete = filter_complete;
96 filter_class->reset = filter_reset;
97 }
98
99 static void
g_mime_filter_best_init(GMimeFilterBest * filter,GMimeFilterBestClass * klass)100 g_mime_filter_best_init (GMimeFilterBest *filter, GMimeFilterBestClass *klass)
101 {
102 filter->frombuf[5] = '\0';
103 }
104
105 static void
g_mime_filter_best_finalize(GObject * object)106 g_mime_filter_best_finalize (GObject *object)
107 {
108 G_OBJECT_CLASS (parent_class)->finalize (object);
109 }
110
111
112 static GMimeFilter *
filter_copy(GMimeFilter * filter)113 filter_copy (GMimeFilter *filter)
114 {
115 GMimeFilterBest *best = (GMimeFilterBest *) filter;
116
117 return g_mime_filter_best_new (best->flags);
118 }
119
120 static void
filter_filter(GMimeFilter * filter,char * inbuf,size_t inlen,size_t prespace,char ** outbuf,size_t * outlen,size_t * outprespace)121 filter_filter (GMimeFilter *filter, char *inbuf, size_t inlen, size_t prespace,
122 char **outbuf, size_t *outlen, size_t *outprespace)
123 {
124 GMimeFilterBest *best = (GMimeFilterBest *) filter;
125 register unsigned char *inptr, *inend;
126 register unsigned char c;
127 size_t left;
128
129 if (best->flags & GMIME_FILTER_BEST_CHARSET)
130 g_mime_charset_step (&best->charset, inbuf, inlen);
131
132 if (best->flags & GMIME_FILTER_BEST_ENCODING) {
133 best->total += inlen;
134
135 inptr = (unsigned char *) inbuf;
136 inend = inptr + inlen;
137
138 while (inptr < inend) {
139 c = 0;
140
141 if (best->midline) {
142 while (inptr < inend && (c = *inptr++) != '\n') {
143 if (c == 0)
144 best->count0++;
145 else if (c & 0x80)
146 best->count8++;
147
148 if (best->fromlen > 0 && best->fromlen < 5)
149 best->frombuf[best->fromlen++] = c & 0xff;
150
151 best->linelen++;
152 }
153
154 if (c == '\n') {
155 best->maxline = MAX (best->maxline, best->linelen);
156 best->startline = TRUE;
157 best->midline = FALSE;
158 best->linelen = 0;
159 }
160 }
161
162 /* check our from-save buffer for "From " */
163 if (best->fromlen == 5 && !strcmp ((char *) best->frombuf, "From "))
164 best->hadfrom = TRUE;
165
166 best->fromlen = 0;
167
168 left = inend - inptr;
169
170 /* if we have not yet found a from-line, check for one */
171 if (best->startline && !best->hadfrom && left > 0) {
172 if (left < 5) {
173 if (!strncmp ((char *) inptr, "From ", left)) {
174 memcpy (best->frombuf, inptr, left);
175 best->frombuf[left] = '\0';
176 best->fromlen = left;
177 break;
178 }
179 } else {
180 if (!strncmp ((char *) inptr, "From ", 5)) {
181 best->hadfrom = TRUE;
182 inptr += 5;
183 }
184 }
185 }
186
187 best->startline = FALSE;
188 best->midline = TRUE;
189 }
190 }
191
192 *outprespace = prespace;
193 *outlen = inlen;
194 *outbuf = inbuf;
195 }
196
197 static void
filter_complete(GMimeFilter * filter,char * in,size_t len,size_t prespace,char ** out,size_t * outlen,size_t * outprespace)198 filter_complete (GMimeFilter *filter, char *in, size_t len, size_t prespace,
199 char **out, size_t *outlen, size_t *outprespace)
200 {
201 GMimeFilterBest *best = (GMimeFilterBest *) filter;
202
203 filter_filter (filter, in, len, prespace, out, outlen, outprespace);
204
205 best->maxline = MAX (best->maxline, best->linelen);
206 }
207
208 static void
filter_reset(GMimeFilter * filter)209 filter_reset (GMimeFilter *filter)
210 {
211 GMimeFilterBest *best = (GMimeFilterBest *) filter;
212
213 g_mime_charset_init (&best->charset);
214 best->count0 = 0;
215 best->count8 = 0;
216 best->total = 0;
217 best->maxline = 0;
218 best->linelen = 0;
219 best->fromlen = 0;
220 best->hadfrom = FALSE;
221 best->startline = TRUE;
222 best->midline = FALSE;
223 }
224
225
226 /**
227 * g_mime_filter_best_new:
228 * @flags: filter flags
229 *
230 * Creates a new GMimeFilterBest filter. @flags are used to determine
231 * which information to keep statistics of. If the
232 * #GMIME_FILTER_BEST_CHARSET bit is set, the filter will be able to
233 * compute the best charset for encoding the stream of data
234 * filtered. If the #GMIME_FILTER_BEST_ENCODING bit is set, the filter
235 * will be able to compute the best Content-Transfer-Encoding for use
236 * with the stream being filtered.
237 *
238 * Note: In order for the g_mime_filter_best_charset() function to
239 * work, the stream being filtered MUST already be encoded in UTF-8.
240 *
241 * Returns: a new best filter with flags @flags.
242 **/
243 GMimeFilter *
g_mime_filter_best_new(GMimeFilterBestFlags flags)244 g_mime_filter_best_new (GMimeFilterBestFlags flags)
245 {
246 GMimeFilterBest *new;
247
248 new = g_object_newv (GMIME_TYPE_FILTER_BEST, 0, NULL);
249 new->flags = flags;
250 filter_reset ((GMimeFilter *) new);
251
252 return (GMimeFilter *) new;
253 }
254
255
256 /**
257 * g_mime_filter_best_charset:
258 * @best: best filter
259 *
260 * Calculates the best charset for encoding the stream filtered
261 * through the @best filter.
262 *
263 * Returns: a pointer to a string containing the name of the charset
264 * best suited for the text filtered through @best.
265 **/
266 const char *
g_mime_filter_best_charset(GMimeFilterBest * best)267 g_mime_filter_best_charset (GMimeFilterBest *best)
268 {
269 const char *charset;
270
271 g_return_val_if_fail (GMIME_IS_FILTER_BEST (best), NULL);
272
273 if (!(best->flags & GMIME_FILTER_BEST_CHARSET))
274 return NULL;
275
276 charset = g_mime_charset_best_name (&best->charset);
277
278 return charset ? charset : "us-ascii";
279 }
280
281
282 /**
283 * g_mime_filter_best_encoding:
284 * @best: a #GMimeFilterBest
285 * @constraint: a #GMimeEncodingConstraint
286 *
287 * Calculates the most efficient Content-Transfer-Encoding for the
288 * stream filtered through @best that fits within the encoding
289 * @constraint.
290 *
291 * Returns: the best encoding for the stream filtered by @best.
292 **/
293 GMimeContentEncoding
g_mime_filter_best_encoding(GMimeFilterBest * best,GMimeEncodingConstraint constraint)294 g_mime_filter_best_encoding (GMimeFilterBest *best, GMimeEncodingConstraint constraint)
295 {
296 GMimeContentEncoding encoding = GMIME_CONTENT_ENCODING_DEFAULT;
297
298 g_return_val_if_fail (GMIME_IS_FILTER_BEST (best), GMIME_CONTENT_ENCODING_DEFAULT);
299
300 if (!(best->flags & GMIME_FILTER_BEST_ENCODING))
301 return GMIME_CONTENT_ENCODING_DEFAULT;
302
303 switch (constraint) {
304 case GMIME_ENCODING_CONSTRAINT_7BIT:
305 if (best->count0 > 0) {
306 encoding = GMIME_CONTENT_ENCODING_BASE64;
307 } else if (best->count8 > 0) {
308 if (best->count8 >= (unsigned int) (best->total * (17.0 / 100.0)))
309 encoding = GMIME_CONTENT_ENCODING_BASE64;
310 else
311 encoding = GMIME_CONTENT_ENCODING_QUOTEDPRINTABLE;
312 } else if (best->maxline > 998) {
313 encoding = GMIME_CONTENT_ENCODING_QUOTEDPRINTABLE;
314 }
315 break;
316 case GMIME_ENCODING_CONSTRAINT_8BIT:
317 if (best->count0 > 0) {
318 encoding = GMIME_CONTENT_ENCODING_BASE64;
319 } else if (best->maxline > 998) {
320 encoding = GMIME_CONTENT_ENCODING_QUOTEDPRINTABLE;
321 }
322 break;
323 case GMIME_ENCODING_CONSTRAINT_BINARY:
324 if (best->count0 + best->count8 > 0)
325 encoding = GMIME_CONTENT_ENCODING_BINARY;
326 break;
327 }
328
329 if (encoding == GMIME_CONTENT_ENCODING_DEFAULT && best->hadfrom)
330 encoding = GMIME_CONTENT_ENCODING_QUOTEDPRINTABLE;
331
332 return encoding;
333 }
334