1 /*
2 * "streamable kanji code filter and converter"
3 * Copyright (c) 1998-2002 HappySize, Inc. All rights reserved.
4 *
5 * LICENSE NOTICES
6 *
7 * This file is part of "streamable kanji code filter and converter",
8 * which is distributed under the terms of GNU Lesser General Public
9 * License (version 2) as published by the Free Software Foundation.
10 *
11 * This software is distributed in the hope that it will be useful,
12 * but WITHOUT ANY WARRANTY; without even the implied warranty of
13 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14 * GNU Lesser General Public License for more details.
15 *
16 * You should have received a copy of the GNU Lesser General Public
17 * License along with "streamable kanji code filter and converter";
18 * if not, write to the Free Software Foundation, Inc., 59 Temple Place,
19 * Suite 330, Boston, MA 02111-1307 USA
20 *
21 * The author of this file:
22 *
23 */
24 /*
25 * The source code included in this files was separated from mbfilter.c
26 * by moriyoshi koizumi <moriyoshi@php.net> on 4 dec 2002.
27 *
28 */
29
30 /* Modified UTF-7 used for 'international mailbox names' in the IMAP protocol
31 * Also known as mUTF-7
32 * Defined in RFC 3501 5.1.3 (https://tools.ietf.org/html/rfc3501)
33 *
34 * Quoting from the RFC:
35 *
36 ***********************************************************************
37 * In modified UTF-7, printable US-ASCII characters, except for "&",
38 * represent themselves; that is, characters with octet values 0x20-0x25
39 * and 0x27-0x7e. The character "&" (0x26) is represented by the
40 * two-octet sequence "&-".
41 *
42 * All other characters (octet values 0x00-0x1f and 0x7f-0xff) are
43 * represented in modified BASE64, with a further modification from
44 * UTF-7 that "," is used instead of "/". Modified BASE64 MUST NOT be
45 * used to represent any printing US-ASCII character which can represent
46 * itself.
47 *
48 * "&" is used to shift to modified BASE64 and "-" to shift back to
49 * US-ASCII. There is no implicit shift from BASE64 to US-ASCII, and
50 * null shifts ("-&" while in BASE64; note that "&-" while in US-ASCII
51 * means "&") are not permitted. However, all names start in US-ASCII,
52 * and MUST end in US-ASCII; that is, a name that ends with a non-ASCII
53 * ISO-10646 character MUST end with a "-").
54 ***********************************************************************
55 *
56 * The purpose of all this is: 1) to keep all parts of IMAP messages 7-bit clean,
57 * 2) to avoid giving special treatment to +, /, \, and ~, since these are
58 * commonly used in mailbox names, and 3) to ensure there is only one
59 * representation of any mailbox name (vanilla UTF-7 does allow multiple
60 * representations of the same string, by Base64-encoding characters which
61 * could have been included as ASCII literals.)
62 *
63 * RFC 2152 also applies, since it defines vanilla UTF-7 (minus IMAP modifications)
64 * The following paragraph is notable:
65 *
66 ***********************************************************************
67 * Unicode is encoded using Modified Base64 by first converting Unicode
68 * 16-bit quantities to an octet stream (with the most significant octet first).
69 * Surrogate pairs (UTF-16) are converted by treating each half of the pair as
70 * a separate 16 bit quantity (i.e., no special treatment). Text with an odd
71 * number of octets is ill-formed. ISO 10646 characters outside the range
72 * addressable via surrogate pairs cannot be encoded.
73 ***********************************************************************
74 *
75 * So after reversing the modified Base64 encoding on an encoded section,
76 * the contents are interpreted as UTF-16BE. */
77
78 #include "mbfilter.h"
79 #include "mbfilter_utf7imap.h"
80
81 static int mbfl_filt_conv_wchar_utf7imap_flush(mbfl_convert_filter *filter);
82 static int mbfl_filt_conv_utf7imap_wchar_flush(mbfl_convert_filter *filter);
83
84 static const char *mbfl_encoding_utf7imap_aliases[] = {"mUTF-7", NULL};
85
86 const mbfl_encoding mbfl_encoding_utf7imap = {
87 mbfl_no_encoding_utf7imap,
88 "UTF7-IMAP",
89 NULL,
90 mbfl_encoding_utf7imap_aliases,
91 NULL,
92 0,
93 &vtbl_utf7imap_wchar,
94 &vtbl_wchar_utf7imap
95 };
96
97 const struct mbfl_convert_vtbl vtbl_utf7imap_wchar = {
98 mbfl_no_encoding_utf7imap,
99 mbfl_no_encoding_wchar,
100 mbfl_filt_conv_common_ctor,
101 NULL,
102 mbfl_filt_conv_utf7imap_wchar,
103 mbfl_filt_conv_utf7imap_wchar_flush,
104 NULL,
105 };
106
107 const struct mbfl_convert_vtbl vtbl_wchar_utf7imap = {
108 mbfl_no_encoding_wchar,
109 mbfl_no_encoding_utf7imap,
110 mbfl_filt_conv_common_ctor,
111 NULL,
112 mbfl_filt_conv_wchar_utf7imap,
113 mbfl_filt_conv_wchar_utf7imap_flush,
114 NULL,
115 };
116
117 #define CK(statement) do { if ((statement) < 0) return (-1); } while (0)
118
mbfl_filt_conv_utf7imap_wchar(int c,mbfl_convert_filter * filter)119 int mbfl_filt_conv_utf7imap_wchar(int c, mbfl_convert_filter *filter)
120 {
121 int s, n = -1;
122
123 if (filter->status != 0) { /* Modified Base64 */
124 if (c >= 'A' && c <= 'Z') {
125 n = c - 65;
126 } else if (c >= 'a' && c <= 'z') {
127 n = c - 71;
128 } else if (c >= '0' && c <= '9') {
129 n = c + 4;
130 } else if (c == '+') {
131 n = 62;
132 } else if (c == ',') {
133 n = 63;
134 }
135
136 if (n < 0 || n > 63) {
137 if (c == '-') {
138 if (filter->status == 1) { /* "&-" -> "&" */
139 CK((*filter->output_function)('&', filter->data));
140 } else if (filter->cache) {
141 /* Base64-encoded section ended abruptly, with partially encoded characters,
142 * or it could be that it ended on the first half of a surrogate pair */
143 CK((*filter->output_function)(MBFL_BAD_INPUT, filter->data));
144 }
145 } else { /* illegal character */
146 CK((*filter->output_function)(MBFL_BAD_INPUT, filter->data));
147 }
148 filter->cache = filter->status = 0;
149 return 0;
150 }
151 }
152
153 switch (filter->status) {
154 /* directly encoded characters */
155 case 0:
156 if (c == '&') { /* shift character */
157 filter->status++;
158 } else if (c >= 0x20 && c <= 0x7E) { /* ASCII */
159 CK((*filter->output_function)(c, filter->data));
160 } else { /* illegal character */
161 CK((*filter->output_function)(MBFL_BAD_INPUT, filter->data));
162 }
163 break;
164
165 /* decode Modified Base64 */
166 case 1:
167 case 2:
168 filter->cache |= n << 10;
169 filter->status = 3;
170 break;
171 case 3:
172 filter->cache |= n << 4;
173 filter->status = 4;
174 break;
175 case 4:
176 s = ((n >> 2) & 0xf) | (filter->cache & 0xffff);
177 n = (n & 0x3) << 14;
178 filter->status = 5;
179 if (s >= 0xd800 && s < 0xdc00) {
180 /* 1st part of surrogate pair */
181 s = (((s & 0x3ff) << 16) + 0x400000) | n;
182 filter->cache = s;
183 } else if (s >= 0xdc00 && s < 0xe000) {
184 /* 2nd part of surrogate pair */
185 if (filter->cache & 0xfff0000) {
186 s &= 0x3ff;
187 s |= (filter->cache & 0xfff0000) >> 6;
188 filter->cache = n;
189 CK((*filter->output_function)(s, filter->data));
190 } else { /* illegal character */
191 CK((*filter->output_function)(MBFL_BAD_INPUT, filter->data));
192 }
193 } else {
194 filter->cache = n;
195 /* Characters which can be expressed as literal, ASCII characters
196 * should not be Base64-encoded */
197 if (s < 0x20 || s > 0x7E || s == '&') {
198 CK((*filter->output_function)(s, filter->data));
199 } else {
200 CK((*filter->output_function)(MBFL_BAD_INPUT, filter->data));
201 }
202 }
203 break;
204
205 case 5:
206 filter->cache |= n << 8;
207 filter->status = 6;
208 break;
209 case 6:
210 filter->cache |= n << 2;
211 filter->status = 7;
212 break;
213 case 7:
214 s = ((n >> 4) & 0x3) | (filter->cache & 0xffff);
215 n = (n & 0xf) << 12;
216 filter->status = 8;
217 if (s >= 0xd800 && s < 0xdc00) {
218 s = (((s & 0x3ff) << 16) + 0x400000) | n;
219 filter->cache = s;
220 } else if (s >= 0xdc00 && s < 0xe000) {
221 if (filter->cache & 0xfff0000) {
222 s &= 0x3ff;
223 s |= (filter->cache & 0xfff0000) >> 6;
224 filter->cache = n;
225 CK((*filter->output_function)(s, filter->data));
226 } else { /* illegal character */
227 CK((*filter->output_function)(MBFL_BAD_INPUT, filter->data));
228 }
229 } else {
230 filter->cache = n;
231 /* Characters which can be expressed as literal, ASCII characters
232 * should not be Base64-encoded */
233 if (s < 0x20 || s > 0x7E || s == '&') {
234 CK((*filter->output_function)(s, filter->data));
235 } else {
236 CK((*filter->output_function)(MBFL_BAD_INPUT, filter->data));
237 }
238 }
239 break;
240
241 case 8:
242 filter->cache |= n << 6;
243 filter->status = 9;
244 break;
245 case 9:
246 s = n | (filter->cache & 0xffff);
247 filter->status = 2;
248 if (s >= 0xd800 && s < 0xdc00) {
249 s = (((s & 0x3ff) << 16) + 0x400000);
250 filter->cache = s;
251 } else if (s >= 0xdc00 && s < 0xe000) {
252 if (filter->cache & 0xfff0000) {
253 s &= 0x3ff;
254 s |= (filter->cache & 0xfff0000) >> 6;
255 filter->cache = 0;
256 CK((*filter->output_function)(s, filter->data));
257 } else { /* illegal character */
258 CK((*filter->output_function)(MBFL_BAD_INPUT, filter->data));
259 }
260 } else {
261 filter->cache = 0;
262 /* Characters which can be expressed as literal, ASCII characters
263 * should not be Base64-encoded */
264 if (s < 0x20 || s > 0x7E || s == '&') {
265 CK((*filter->output_function)(s, filter->data));
266 } else {
267 CK((*filter->output_function)(MBFL_BAD_INPUT, filter->data));
268 }
269 }
270 break;
271
272 default:
273 filter->status = 0;
274 break;
275 }
276
277 return 0;
278 }
279
mbfl_filt_conv_utf7imap_wchar_flush(mbfl_convert_filter * filter)280 static int mbfl_filt_conv_utf7imap_wchar_flush(mbfl_convert_filter *filter)
281 {
282 if (filter->status) {
283 /* It is illegal for a UTF-7 IMAP string to end in a Base-64 encoded
284 * section. It should always change back to ASCII before the end. */
285 (*filter->output_function)(MBFL_BAD_INPUT, filter->data);
286 }
287
288 if (filter->flush_function) {
289 (*filter->flush_function)(filter->data);
290 }
291
292 return 0;
293 }
294
295 static const unsigned char mbfl_utf7imap_base64_table[] =
296 {
297 /* 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'I', 'J', 'K', 'L', 'M', */
298 0x41,0x42,0x43,0x44,0x45,0x46,0x47,0x48,0x49,0x4a,0x4b,0x4c,0x4d,
299 /* 'N', 'O', 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W', 'X', 'Y', 'Z', */
300 0x4e,0x4f,0x50,0x51,0x52,0x53,0x54,0x55,0x56,0x57,0x58,0x59,0x5a,
301 /* 'a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 'j', 'k', 'l', 'm', */
302 0x61,0x62,0x63,0x64,0x65,0x66,0x67,0x68,0x69,0x6a,0x6b,0x6c,0x6d,
303 /* 'n', 'o', 'p', 'q', 'r', 's', 't', 'u', 'v', 'w', 'x', 'y', 'z', */
304 0x6e,0x6f,0x70,0x71,0x72,0x73,0x74,0x75,0x76,0x77,0x78,0x79,0x7a,
305 /* '0', '1', '2', '3', '4', '5', '6', '7', '8', '9', '+', ',', '\0' */
306 0x30,0x31,0x32,0x33,0x34,0x35,0x36,0x37,0x38,0x39,0x2b,0x2c,0x00
307 };
308
mbfl_filt_conv_wchar_utf7imap(int c,mbfl_convert_filter * filter)309 int mbfl_filt_conv_wchar_utf7imap(int c, mbfl_convert_filter *filter)
310 {
311 int n = 0, s;
312
313 if (c == '&') {
314 n = 1;
315 } else if ((c >= 0x20 && c <= 0x7e) || c == 0) {
316 n = 2;
317 } else if (c >= 0 && c < MBFL_WCSPLANE_UCS2MAX) {
318 ;
319 } else if (c >= MBFL_WCSPLANE_SUPMIN && c < MBFL_WCSPLANE_SUPMAX) {
320 s = ((c >> 10) - 0x40) | 0xd800;
321 CK((*filter->filter_function)(s, filter));
322 s = (c & 0x3ff) | 0xdc00;
323 CK((*filter->filter_function)(s, filter));
324 return 0;
325 } else {
326 CK(mbfl_filt_conv_illegal_output(c, filter));
327 return 0;
328 }
329
330 switch (filter->status) {
331 case 0:
332 if (n != 0) { /* directly encode characters */
333 CK((*filter->output_function)(c, filter->data));
334 if (n == 1) {
335 CK((*filter->output_function)(0x2d, filter->data)); /* '-' */
336 }
337 } else { /* Modified Base64 */
338 CK((*filter->output_function)(0x26, filter->data)); /* '&' */
339 filter->status = 1;
340 filter->cache = c;
341 }
342 break;
343
344 /* encode Modified Base64 */
345 case 1:
346 s = filter->cache;
347 CK((*filter->output_function)(mbfl_utf7imap_base64_table[(s >> 10) & 0x3f], filter->data));
348 CK((*filter->output_function)(mbfl_utf7imap_base64_table[(s >> 4) & 0x3f], filter->data));
349 if (n != 0) {
350 CK((*filter->output_function)(mbfl_utf7imap_base64_table[(s << 2) & 0x3c], filter->data));
351 CK((*filter->output_function)('-', filter->data));
352 CK((*filter->output_function)(c, filter->data));
353 if (n == 1) {
354 CK((*filter->output_function)('-', filter->data));
355 }
356 filter->status = 0;
357 } else {
358 filter->status = 2;
359 filter->cache = ((s & 0xf) << 16) | c;
360 }
361 break;
362
363 case 2:
364 s = filter->cache;
365 CK((*filter->output_function)(mbfl_utf7imap_base64_table[(s >> 14) & 0x3f], filter->data));
366 CK((*filter->output_function)(mbfl_utf7imap_base64_table[(s >> 8) & 0x3f], filter->data));
367 CK((*filter->output_function)(mbfl_utf7imap_base64_table[(s >> 2) & 0x3f], filter->data));
368 if (n != 0) {
369 CK((*filter->output_function)(mbfl_utf7imap_base64_table[(s << 4) & 0x30], filter->data));
370 CK((*filter->output_function)('-', filter->data));
371 CK((*filter->output_function)(c, filter->data));
372 if (n == 1) {
373 CK((*filter->output_function)('-', filter->data));
374 }
375 filter->status = 0;
376 } else {
377 filter->status = 3;
378 filter->cache = ((s & 0x3) << 16) | c;
379 }
380 break;
381
382 case 3:
383 s = filter->cache;
384 CK((*filter->output_function)(mbfl_utf7imap_base64_table[(s >> 12) & 0x3f], filter->data));
385 CK((*filter->output_function)(mbfl_utf7imap_base64_table[(s >> 6) & 0x3f], filter->data));
386 CK((*filter->output_function)(mbfl_utf7imap_base64_table[s & 0x3f], filter->data));
387 if (n != 0) {
388 CK((*filter->output_function)('-', filter->data));
389 CK((*filter->output_function)(c, filter->data));
390 if (n == 1) {
391 CK((*filter->output_function)('-', filter->data));
392 }
393 filter->status = 0;
394 } else {
395 filter->status = 1;
396 filter->cache = c;
397 }
398 break;
399
400 default:
401 filter->status = 0;
402 break;
403 }
404
405 return 0;
406 }
407
mbfl_filt_conv_wchar_utf7imap_flush(mbfl_convert_filter * filter)408 static int mbfl_filt_conv_wchar_utf7imap_flush(mbfl_convert_filter *filter)
409 {
410 int status = filter->status, cache = filter->cache;
411 filter->status = filter->cache = 0;
412
413 /* flush fragments */
414 switch (status) {
415 case 1:
416 CK((*filter->output_function)(mbfl_utf7imap_base64_table[(cache >> 10) & 0x3f], filter->data));
417 CK((*filter->output_function)(mbfl_utf7imap_base64_table[(cache >> 4) & 0x3f], filter->data));
418 CK((*filter->output_function)(mbfl_utf7imap_base64_table[(cache << 2) & 0x3c], filter->data));
419 CK((*filter->output_function)('-', filter->data));
420 break;
421
422 case 2:
423 CK((*filter->output_function)(mbfl_utf7imap_base64_table[(cache >> 14) & 0x3f], filter->data));
424 CK((*filter->output_function)(mbfl_utf7imap_base64_table[(cache >> 8) & 0x3f], filter->data));
425 CK((*filter->output_function)(mbfl_utf7imap_base64_table[(cache >> 2) & 0x3f], filter->data));
426 CK((*filter->output_function)(mbfl_utf7imap_base64_table[(cache << 4) & 0x30], filter->data));
427 CK((*filter->output_function)('-', filter->data));
428 break;
429
430 case 3:
431 CK((*filter->output_function)(mbfl_utf7imap_base64_table[(cache >> 12) & 0x3f], filter->data));
432 CK((*filter->output_function)(mbfl_utf7imap_base64_table[(cache >> 6) & 0x3f], filter->data));
433 CK((*filter->output_function)(mbfl_utf7imap_base64_table[cache & 0x3f], filter->data));
434 CK((*filter->output_function)('-', filter->data));
435 break;
436 }
437 return 0;
438 }
439