2 * fmt_rfc2047.c -- decode RFC-2047 header format
4 * This code is Copyright (c) 2002, by the authors of nmh. See the
5 * COPYRIGHT file in the root directory of the nmh distribution for
6 * complete copyright information.
16 static signed char hexindex[] = {
17 -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,
18 -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,
19 -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,
20 0, 1, 2, 3, 4, 5, 6, 7, 8, 9,-1,-1,-1,-1,-1,-1,
21 -1,10,11,12,13,14,15,-1,-1,-1,-1,-1,-1,-1,-1,-1,
22 -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,
23 -1,10,11,12,13,14,15,-1,-1,-1,-1,-1,-1,-1,-1,-1,
24 -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1
27 static signed char index_64[128] = {
28 -1,-1,-1,-1, -1,-1,-1,-1, -1,-1,-1,-1, -1,-1,-1,-1,
29 -1,-1,-1,-1, -1,-1,-1,-1, -1,-1,-1,-1, -1,-1,-1,-1,
30 -1,-1,-1,-1, -1,-1,-1,-1, -1,-1,-1,62, -1,-1,-1,63,
31 52,53,54,55, 56,57,58,59, 60,61,-1,-1, -1,-1,-1,-1,
32 -1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9,10, 11,12,13,14,
33 15,16,17,18, 19,20,21,22, 23,24,25,-1, -1,-1,-1,-1,
34 -1,26,27,28, 29,30,31,32, 33,34,35,36, 37,38,39,40,
35 41,42,43,44, 45,46,47,48, 49,50,51,-1, -1,-1,-1,-1
38 #define char64(c) (((unsigned char) (c) > 127) ? -1 : index_64[(unsigned char) (c)])
41 unqp (unsigned char byte1, unsigned char byte2)
43 if (hexindex[byte1] == -1 || hexindex[byte2] == -1)
45 return (hexindex[byte1] << 4 | hexindex[byte2]);
48 /* Check if character is linear whitespace */
49 #define is_lws(c) ((c) == ' ' || (c) == '\t' || (c) == '\n')
53 * Decode the string as a RFC-2047 header field
56 /* Add character to the destination buffer, and bomb out if it fills up */
57 #define ADDCHR(C) do { *q++ = (C); dstlen--; if (!dstlen) goto buffull; } while (0)
60 decode_rfc2047 (char *str, char *dst, size_t dstlen)
63 char *startofmime, *endofmime;
64 int c, quoted_printable;
65 int encoding_found = 0; /* did we decode anything? */
66 int between_encodings = 0; /* are we between two encodings? */
67 int equals_pending = 0; /* is there a '=' pending? */
68 int whitespace = 0; /* how much whitespace between encodings? */
70 int use_iconv = 0; /* are we converting encoding with iconv? */
73 char *saveq, *convbuf = NULL;
81 * Do a quick and dirty check for the '=' character.
82 * This should quickly eliminate many cases.
84 if (!strchr (str, '='))
87 for (p = str, q = dst; *p; p++) {
97 * If we had an '=' character pending from
98 * last iteration, then add it first.
100 if (equals_pending) {
103 between_encodings = 0; /* we have added non-whitespace text */
107 /* count linear whitespace while between encodings */
108 if (between_encodings && is_lws(*p))
111 between_encodings = 0; /* we have added non-whitespace text */
116 equals_pending = 1; /* we have a '=' pending */
118 /* Check for initial =? */
119 if (*p == '=' && p[1] && p[1] == '?' && p[2]) {
122 /* Scan ahead for the next '?' character */
123 for (pp = startofmime; *pp && *pp != '?'; pp++)
129 /* Check if character set can be handled natively */
130 if (!check_charset(startofmime, pp - startofmime)) {
132 /* .. it can't. We'll use iconv then. */
134 cd = iconv_open(get_charset(), startofmime);
135 fromutf8 = !mh_strcasecmp(startofmime, "UTF-8");
137 if (cd == (iconv_t)-1) continue;
144 startofmime = pp + 1;
146 /* Check for valid encoding type */
147 if (*startofmime != 'B' && *startofmime != 'b' &&
148 *startofmime != 'Q' && *startofmime != 'q')
151 /* Is encoding quoted printable or base64? */
152 quoted_printable = (*startofmime == 'Q' || *startofmime == 'q');
155 /* Check for next '?' character */
156 if (*startofmime != '?')
161 * Scan ahead for the ending ?=
163 * While doing this, we will also check if encoded
164 * word has any embedded linear whitespace.
167 for (pp = startofmime; *pp && *(pp+1); pp++) {
170 } else if (*pp == '?' && pp[1] == '=') {
175 if (is_lws(*pp) || endofmime == NULL)
179 * We've found an encoded word, so we can drop
180 * the '=' that was pending
185 * If we are between two encoded words separated only by
186 * linear whitespace, then we ignore the whitespace.
187 * We will roll back the buffer the number of whitespace
188 * characters we've seen since last encoded word.
190 if (between_encodings) {
192 dstlen += whitespace;
197 * empty encoded text. This ensures that we don't
198 * malloc 0 bytes but skip on to the end
200 if (endofmime == startofmime && use_iconv) {
208 q = convbuf = (char *) mh_xmalloc(endofmime - startofmime);
210 /* ADDCHR2 is for adding characters when q is or might be convbuf:
211 * in this case on buffer-full we want to run iconv before returning.
212 * I apologise for the dreadful name.
214 # define ADDCHR2(C) do { *q++ = (C); dstlen--; if (!dstlen) goto iconvbuffull; } while (0)
216 # define ADDCHR2(C) ADDCHR(C)
219 /* Now decode the text */
220 if (quoted_printable) {
221 for (pp = startofmime; pp < endofmime; pp++) {
223 c = unqp (pp[1], pp[2]);
229 } else if (*pp == '_') {
238 c1 = c2 = c3 = c4 = -1;
241 while (pp < endofmime) {
243 while ((pp < endofmime) &&
244 ((c1 = char64(*pp)) == -1)) {
247 if (pp < endofmime) {
250 while ((pp < endofmime) &&
251 ((c2 = char64(*pp)) == -1)) {
254 if (pp < endofmime && c1 != -1 && c2 != -1) {
255 ADDCHR2((c1 << 2) | (c2 >> 4));
259 while ((pp < endofmime) &&
260 ((c3 = char64(*pp)) == -1)) {
263 if (pp < endofmime && c2 != -1 && c3 != -1) {
264 ADDCHR2(((c2 & 0xF) << 4) | (c3 >> 2));
268 while ((pp < endofmime) &&
269 ((c4 = char64(*pp)) == -1)) {
272 if (pp < endofmime && c3 != -1 && c4 != -1) {
273 ADDCHR2(((c3 & 0x3) << 6) | (c4));
281 /* NB that the string at convbuf is not necessarily
282 * NUL terminated here:
283 * q points to the first byte after the valid part.
285 /* Convert to native character set */
287 size_t inbytes = q - convbuf;
288 ICONV_CONST char *start = convbuf;
291 if (iconv(cd, &start, &inbytes, &saveq, &savedstlen) ==
295 /* character couldn't be converted. we output a `?'
296 * and try to carry on which won't work if
297 * either encoding was stateful */
298 iconv (cd, 0, 0, &saveq, &savedstlen);
305 /* skip to next input character */
307 for (start++;(start < q) && ((*start & 192) == 128);start++)
316 /* Stop now if (1) we hit the end of the buffer trying to do
317 * MIME decoding and have just iconv-converted a partial string
318 * or (2) our iconv-conversion hit the end of the buffer.
320 if (!dstlen || !savedstlen)
328 * Now that we are done decoding this particular
329 * encoded word, advance string to trailing '='.
333 encoding_found = 1; /* we found (at least 1) encoded word */
334 between_encodings = 1; /* we have just decoded something */
335 whitespace = 0; /* re-initialize amount of whitespace */
339 if (use_iconv) iconv_close(cd);
342 /* If an equals was pending at end of string, add it now. */
347 return encoding_found;
350 /* q is currently just off the end of the buffer,
351 * so rewind to NUL terminate */
354 return encoding_found;