Annotation of embedaddon/libiconv/lib/utf7.h, revision 1.1.1.1
1.1 misho 1: /*
2: * Copyright (C) 1999-2001, 2008 Free Software Foundation, Inc.
3: * This file is part of the GNU LIBICONV Library.
4: *
5: * The GNU LIBICONV Library is free software; you can redistribute it
6: * and/or modify it under the terms of the GNU Library General Public
7: * License as published by the Free Software Foundation; either version 2
8: * of the License, or (at your option) any later version.
9: *
10: * The GNU LIBICONV Library is distributed in the hope that it will be
11: * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of
12: * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
13: * Library General Public License for more details.
14: *
15: * You should have received a copy of the GNU Library General Public
16: * License along with the GNU LIBICONV Library; see the file COPYING.LIB.
17: * If not, write to the Free Software Foundation, Inc., 51 Franklin Street,
18: * Fifth Floor, Boston, MA 02110-1301, USA.
19: */
20:
21: /*
22: * UTF-7
23: */
24:
25: /* Specification: RFC 2152 (and old RFC 1641, RFC 1642) */
26: /* The original Base64 encoding is defined in RFC 2045. */
27:
28: /* Set of direct characters:
29: * A-Z a-z 0-9 ' ( ) , - . / : ? space tab lf cr
30: */
31: static const unsigned char direct_tab[128/8] = {
32: 0x00, 0x26, 0x00, 0x00, 0x81, 0xf3, 0xff, 0x87,
33: 0xfe, 0xff, 0xff, 0x07, 0xfe, 0xff, 0xff, 0x07,
34: };
35: #define isdirect(ch) ((ch) < 128 && ((direct_tab[(ch)>>3] >> (ch & 7)) & 1))
36:
37: /* Set of direct and optional direct characters:
38: * A-Z a-z 0-9 ' ( ) , - . / : ? space tab lf cr
39: * ! " # $ % & * ; < = > @ [ ] ^ _ ` { | }
40: */
41: static const unsigned char xdirect_tab[128/8] = {
42: 0x00, 0x26, 0x00, 0x00, 0xff, 0xf7, 0xff, 0xff,
43: 0xff, 0xff, 0xff, 0xef, 0xff, 0xff, 0xff, 0x3f,
44: };
45: #define isxdirect(ch) ((ch) < 128 && ((xdirect_tab[(ch)>>3] >> (ch & 7)) & 1))
46:
47: /* Set of base64 characters, extended:
48: * A-Z a-z 0-9 + / -
49: */
50: static const unsigned char xbase64_tab[128/8] = {
51: 0x00, 0x00, 0x00, 0x00, 0x00, 0xa8, 0xff, 0x03,
52: 0xfe, 0xff, 0xff, 0x07, 0xfe, 0xff, 0xff, 0x07,
53: };
54: #define isxbase64(ch) ((ch) < 128 && ((xbase64_tab[(ch)>>3] >> (ch & 7)) & 1))
55:
56: /*
57: * The state is structured as follows:
58: * bit 1..0: shift
59: * bit 7..2: data
60: * Precise meaning:
61: * shift data
62: * 0 0 not inside base64 encoding
63: * 1 0 inside base64, no pending bits
64: * 2 XXXX00 inside base64, 4 bits remain from 2nd byte
65: * 3 XX0000 inside base64, 2 bits remain from 3rd byte
66: */
67:
68: static int
69: utf7_mbtowc (conv_t conv, ucs4_t *pwc, const unsigned char *s, int n)
70: {
71: state_t state = conv->istate;
72: int count = 0; /* number of input bytes already read */
73: if (state & 3)
74: goto active;
75: else
76: goto inactive;
77:
78: inactive:
79: {
80: /* Here (state & 3) == 0 */
81: if (n < count+1)
82: goto none;
83: {
84: unsigned char c = *s;
85: if (isxdirect(c)) {
86: *pwc = (ucs4_t) c;
87: conv->istate = state;
88: return count+1;
89: }
90: if (c == '+') {
91: if (n < count+2)
92: goto none;
93: if (s[1] == '-') {
94: *pwc = (ucs4_t) '+';
95: conv->istate = state;
96: return count+2;
97: }
98: s++; count++;
99: state = 1;
100: goto active;
101: }
102: goto ilseq;
103: }
104: }
105:
106: active:
107: {
108: /* base64 encoding active */
109: unsigned int wc = 0;
110: state_t base64state = state;
111: unsigned int kmax = 2; /* number of payload bytes to read */
112: unsigned int k = 0; /* number of payload bytes already read */
113: unsigned int base64count = 0; /* number of base64 bytes already read */
114: for (;;) {
115: unsigned char c = *s;
116: unsigned int i;
117: if (c >= 'A' && c <= 'Z')
118: i = c-'A';
119: else if (c >= 'a' && c <= 'z')
120: i = c-'a'+26;
121: else if (c >= '0' && c <= '9')
122: i = c-'0'+52;
123: else if (c == '+')
124: i = 62;
125: else if (c == '/')
126: i = 63;
127: else {
128: /* c terminates base64 encoding */
129: if (base64state & -4)
130: goto ilseq; /* data must be 0, otherwise illegal */
131: if (base64count)
132: goto ilseq; /* partial UTF-16 characters are invalid */
133: if (c == '-') {
134: s++; count++;
135: }
136: state = 0;
137: goto inactive;
138: }
139: s++; base64count++;
140: /* read 6 bits: 0 <= i < 64 */
141: switch (base64state & 3) {
142: case 1: /* inside base64, no pending bits */
143: base64state = (i << 2) | 0; break;
144: case 0: /* inside base64, 6 bits remain from 1st byte */
145: wc = (wc << 8) | (base64state & -4) | (i >> 4); k++;
146: base64state = ((i & 15) << 4) | 2; break;
147: case 2: /* inside base64, 4 bits remain from 2nd byte */
148: wc = (wc << 8) | (base64state & -4) | (i >> 2); k++;
149: base64state = ((i & 3) << 6) | 3; break;
150: case 3: /* inside base64, 2 bits remain from 3rd byte */
151: wc = (wc << 8) | (base64state & -4) | i; k++;
152: base64state = 1; break;
153: }
154: if (k == kmax) {
155: /* UTF-16: When we see a High Surrogate, we must also decode
156: the following Low Surrogate. */
157: if (kmax == 2 && (wc >= 0xd800 && wc < 0xdc00))
158: kmax = 4;
159: else
160: break;
161: }
162: if (n < count+base64count+1)
163: goto none;
164: }
165: /* Here k = kmax > 0, hence base64count > 0. */
166: if ((base64state & 3) == 0) abort();
167: if (kmax == 4) {
168: ucs4_t wc1 = wc >> 16;
169: ucs4_t wc2 = wc & 0xffff;
170: if (!(wc1 >= 0xd800 && wc1 < 0xdc00)) abort();
171: if (!(wc2 >= 0xdc00 && wc2 < 0xe000)) goto ilseq;
172: *pwc = 0x10000 + ((wc1 - 0xd800) << 10) + (wc2 - 0xdc00);
173: } else {
174: *pwc = wc;
175: }
176: conv->istate = base64state;
177: return count+base64count;
178: }
179:
180: none:
181: conv->istate = state;
182: return RET_TOOFEW(count);
183:
184: ilseq:
185: conv->istate = state;
186: return RET_SHIFT_ILSEQ(count);
187: }
188:
189: /*
190: * The state is structured as follows:
191: * bit 1..0: shift
192: * bit 7..2: data
193: * Precise meaning:
194: * shift data
195: * 0 0 not inside base64 encoding
196: * 1 0 inside base64, no pending bits
197: * 2 XX00 inside base64, 2 bits known for 2nd byte
198: * 3 XXXX inside base64, 4 bits known for 3rd byte
199: */
200:
201: /* Define this to 1 if you want the so-called "optional direct" characters
202: ! " # $ % & * ; < = > @ [ ] ^ _ ` { | }
203: to be encoded. Define to 0 if you want them to be passed straight through,
204: like the so-called "direct" characters.
205: We set this to 1 because it's safer.
206: */
207: #define UTF7_ENCODE_OPTIONAL_CHARS 1
208:
209: static int
210: utf7_wctomb (conv_t conv, unsigned char *r, ucs4_t iwc, int n)
211: {
212: state_t state = conv->ostate;
213: unsigned int wc = iwc;
214: int count = 0;
215: if (state & 3)
216: goto active;
217:
218: /*inactive:*/
219: {
220: if (UTF7_ENCODE_OPTIONAL_CHARS ? isdirect(wc) : isxdirect(wc)) {
221: r[0] = (unsigned char) wc;
222: /*conv->ostate = state;*/
223: return 1;
224: } else {
225: *r++ = '+';
226: if (wc == '+') {
227: if (n < 2)
228: return RET_TOOSMALL;
229: *r = '-';
230: /*conv->ostate = state;*/
231: return 2;
232: }
233: count = 1;
234: state = 1;
235: goto active;
236: }
237: }
238:
239: active:
240: {
241: /* base64 encoding active */
242: if (UTF7_ENCODE_OPTIONAL_CHARS ? isdirect(wc) : isxdirect(wc)) {
243: /* deactivate base64 encoding */
244: count += ((state & 3) >= 2 ? 1 : 0) + (isxbase64(wc) ? 1 : 0) + 1;
245: if (n < count)
246: return RET_TOOSMALL;
247: if ((state & 3) >= 2) {
248: unsigned int i = state & -4;
249: unsigned char c;
250: if (i < 26)
251: c = i+'A';
252: else if (i < 52)
253: c = i-26+'a';
254: else if (i < 62)
255: c = i-52+'0';
256: else if (i == 62)
257: c = '+';
258: else if (i == 63)
259: c = '/';
260: else
261: abort();
262: *r++ = c;
263: }
264: if (isxbase64(wc))
265: *r++ = '-';
266: state = 0;
267: *r++ = (unsigned char) wc;
268: conv->ostate = state;
269: return count;
270: } else {
271: unsigned int k; /* number of payload bytes to write */
272: if (wc < 0x10000) {
273: k = 2;
274: count += ((state & 3) >= 2 ? 3 : 2);
275: } else if (wc < 0x110000) {
276: unsigned int wc1 = 0xd800 + ((wc - 0x10000) >> 10);
277: unsigned int wc2 = 0xdc00 + ((wc - 0x10000) & 0x3ff);
278: wc = (wc1 << 16) | wc2;
279: k = 4;
280: count += ((state & 3) >= 3 ? 6 : 5);
281: } else
282: return RET_ILUNI;
283: if (n < count)
284: return RET_TOOSMALL;
285: for (;;) {
286: unsigned int i;
287: unsigned char c;
288: switch (state & 3) {
289: case 0: /* inside base64, 6 bits known for 4th byte */
290: c = (state & -4) >> 2; state = 1; break;
291: case 1: /* inside base64, no pending bits */
292: i = (wc >> (8 * --k)) & 0xff;
293: c = i >> 2; state = ((i & 3) << 4) | 2; break;
294: case 2: /* inside base64, 2 bits known for 2nd byte */
295: i = (wc >> (8 * --k)) & 0xff;
296: c = (state & -4) | (i >> 4); state = ((i & 15) << 2) | 3; break;
297: case 3: /* inside base64, 4 bits known for 3rd byte */
298: i = (wc >> (8 * --k)) & 0xff;
299: c = (state & -4) | (i >> 6); state = ((i & 63) << 2) | 0; break;
300: default: abort(); /* stupid gcc */
301: }
302: if (c < 26)
303: c = c+'A';
304: else if (c < 52)
305: c = c-26+'a';
306: else if (c < 62)
307: c = c-52+'0';
308: else if (c == 62)
309: c = '+';
310: else if (c == 63)
311: c = '/';
312: else
313: abort();
314: *r++ = c;
315: if ((state & 3) && (k == 0))
316: break;
317: }
318: conv->ostate = state;
319: return count;
320: }
321: }
322: }
323:
324: static int
325: utf7_reset (conv_t conv, unsigned char *r, int n)
326: {
327: state_t state = conv->ostate;
328: if (state & 3) {
329: /* deactivate base64 encoding */
330: unsigned int count = ((state & 3) >= 2 ? 1 : 0) + 1;
331: if (n < count)
332: return RET_TOOSMALL;
333: if ((state & 3) >= 2) {
334: unsigned int i = state & -4;
335: unsigned char c;
336: if (i < 26)
337: c = i+'A';
338: else if (i < 52)
339: c = i-26+'a';
340: else if (i < 62)
341: c = i-52+'0';
342: else if (i == 62)
343: c = '+';
344: else if (i == 63)
345: c = '/';
346: else
347: abort();
348: *r++ = c;
349: }
350: *r++ = '-';
351: /* conv->ostate = 0; will be done by the caller */
352: return count;
353: } else
354: return 0;
355: }
FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>