Branch data Line data Source code
1 : : /* punycode.c --- Implementation of punycode used to ASCII encode IDN's.
2 : : Copyright (C) 2002-2012 Simon Josefsson
3 : :
4 : : This file is part of GNU Libidn.
5 : :
6 : : GNU Libidn is free software: you can redistribute it and/or
7 : : modify it under the terms of either:
8 : :
9 : : * the GNU Lesser General Public License as published by the Free
10 : : Software Foundation; either version 3 of the License, or (at
11 : : your option) any later version.
12 : :
13 : : or
14 : :
15 : : * the GNU General Public License as published by the Free
16 : : Software Foundation; either version 2 of the License, or (at
17 : : your option) any later version.
18 : :
19 : : or both in parallel, as here.
20 : :
21 : : GNU Libidn is distributed in the hope that it will be useful,
22 : : but WITHOUT ANY WARRANTY; without even the implied warranty of
23 : : MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
24 : : General Public License for more details.
25 : :
26 : : You should have received copies of the GNU General Public License and
27 : : the GNU Lesser General Public License along with this program. If
28 : : not, see <http://www.gnu.org/licenses/>. */
29 : :
30 : : /*
31 : : * This file is derived from RFC 3492bis written by Adam M. Costello.
32 : : *
33 : : * Disclaimer and license: Regarding this entire document or any
34 : : * portion of it (including the pseudocode and C code), the author
35 : : * makes no guarantees and is not responsible for any damage resulting
36 : : * from its use. The author grants irrevocable permission to anyone
37 : : * to use, modify, and distribute it in any way that does not diminish
38 : : * the rights of anyone else to use, modify, and distribute it,
39 : : * provided that redistributed derivative works do not contain
40 : : * misleading author or version information. Derivative works need
41 : : * not be licensed under similar terms.
42 : : *
43 : : * Copyright (C) The Internet Society (2003). All Rights Reserved.
44 : : *
45 : : * This document and translations of it may be copied and furnished to
46 : : * others, and derivative works that comment on or otherwise explain it
47 : : * or assist in its implementation may be prepared, copied, published
48 : : * and distributed, in whole or in part, without restriction of any
49 : : * kind, provided that the above copyright notice and this paragraph are
50 : : * included on all such copies and derivative works. However, this
51 : : * document itself may not be modified in any way, such as by removing
52 : : * the copyright notice or references to the Internet Society or other
53 : : * Internet organizations, except as needed for the purpose of
54 : : * developing Internet standards in which case the procedures for
55 : : * copyrights defined in the Internet Standards process must be
56 : : * followed, or as required to translate it into languages other than
57 : : * English.
58 : : *
59 : : * The limited permissions granted above are perpetual and will not be
60 : : * revoked by the Internet Society or its successors or assigns.
61 : : *
62 : : * This document and the information contained herein is provided on an
63 : : * "AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING
64 : : * TASK FORCE DISCLAIMS ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING
65 : : * BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION
66 : : * HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF
67 : : * MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
68 : : */
69 : :
70 : : #include <config.h>
71 : : #include <string.h>
72 : :
73 : : #include "punycode.h"
74 : :
75 : : /*** Bootstring parameters for Punycode ***/
76 : :
77 : : enum
78 : : { base = 36, tmin = 1, tmax = 26, skew = 38, damp = 700,
79 : : initial_bias = 72, initial_n = 0x80, delimiter = 0x2D
80 : : };
81 : :
82 : : /* basic(cp) tests whether cp is a basic code point: */
83 : : #define basic(cp) ((punycode_uint)(cp) < 0x80)
84 : :
85 : : /* delim(cp) tests whether cp is a delimiter: */
86 : : #define delim(cp) ((cp) == delimiter)
87 : :
88 : : /* decode_digit(cp) returns the numeric value of a basic code */
89 : : /* point (for use in representing integers) in the range 0 to */
90 : : /* base-1, or base if cp does not represent a value. */
91 : :
92 : : static punycode_uint
93 : 1019 : decode_digit (punycode_uint cp)
94 : : {
95 [ + + ][ + + ]: 1682 : return cp - 48 < 10 ? cp - 22 : cp - 65 < 26 ? cp - 65 :
96 [ + + ]: 663 : cp - 97 < 26 ? cp - 97 : base;
97 : : }
98 : :
99 : : /* encode_digit(d,flag) returns the basic code point whose value */
100 : : /* (when used for representing integers) is d, which needs to be in */
101 : : /* the range 0 to base-1. The lowercase form is used unless flag is */
102 : : /* nonzero, in which case the uppercase form is used. The behavior */
103 : : /* is undefined if flag is nonzero and digit d has no uppercase form. */
104 : :
105 : : static char
106 : 1857 : encode_digit (punycode_uint d, int flag)
107 : : {
108 [ + + ][ - + ]: 1857 : return d + 22 + 75 * (d < 26) - ((flag != 0) << 5);
109 : : /* 0..25 map to ASCII a..z or A..Z */
110 : : /* 26..35 map to ASCII 0..9 */
111 : : }
112 : :
113 : : /* flagged(bcp) tests whether a basic code point is flagged */
114 : : /* (uppercase). The behavior is undefined if bcp is not a */
115 : : /* basic code point. */
116 : :
117 : : #define flagged(bcp) ((punycode_uint)(bcp) - 65 < 26)
118 : :
119 : : /* encode_basic(bcp,flag) forces a basic code point to lowercase */
120 : : /* if flag is zero, uppercase if flag is nonzero, and returns */
121 : : /* the resulting code point. The code point is unchanged if it */
122 : : /* is caseless. The behavior is undefined if bcp is not a basic */
123 : : /* code point. */
124 : :
125 : : static char
126 : 0 : encode_basic (punycode_uint bcp, int flag)
127 : : {
128 [ # # ]: 0 : bcp -= (bcp - 97 < 26) << 5;
129 [ # # ][ # # ]: 0 : return bcp + ((!flag && (bcp - 65 < 26)) << 5);
130 : : }
131 : :
132 : : /*** Platform-specific constants ***/
133 : :
134 : : /* maxint is the maximum value of a punycode_uint variable: */
135 : : static const punycode_uint maxint = -1;
136 : : /* Because maxint is unsigned, -1 becomes the maximum value. */
137 : :
138 : : /*** Bias adaptation function ***/
139 : :
140 : : static punycode_uint
141 : 1452 : adapt (punycode_uint delta, punycode_uint numpoints, int firsttime)
142 : : {
143 : : punycode_uint k;
144 : :
145 [ + + ]: 1452 : delta = firsttime ? delta / damp : delta >> 1;
146 : : /* delta >> 1 is a faster way of doing delta / 2 */
147 : 1452 : delta += delta / numpoints;
148 : :
149 [ + + ]: 1745 : for (k = 0; delta > ((base - tmin) * tmax) / 2; k += base)
150 : : {
151 : 293 : delta /= base - tmin;
152 : : }
153 : :
154 : 1452 : return k + (base - tmin + 1) * delta / (delta + skew);
155 : : }
156 : :
157 : : /*** Main encode function ***/
158 : :
159 : : /**
160 : : * punycode_encode:
161 : : * @input_length: The number of code points in the @input array and
162 : : * the number of flags in the @case_flags array.
163 : : * @input: An array of code points. They are presumed to be Unicode
164 : : * code points, but that is not strictly REQUIRED. The array
165 : : * contains code points, not code units. UTF-16 uses code units
166 : : * D800 through DFFF to refer to code points 10000..10FFFF. The
167 : : * code points D800..DFFF do not occur in any valid Unicode string.
168 : : * The code points that can occur in Unicode strings (0..D7FF and
169 : : * E000..10FFFF) are also called Unicode scalar values.
170 : : * @case_flags: A %NULL pointer or an array of boolean values parallel
171 : : * to the @input array. Nonzero (true, flagged) suggests that the
172 : : * corresponding Unicode character be forced to uppercase after
173 : : * being decoded (if possible), and zero (false, unflagged) suggests
174 : : * that it be forced to lowercase (if possible). ASCII code points
175 : : * (0..7F) are encoded literally, except that ASCII letters are
176 : : * forced to uppercase or lowercase according to the corresponding
177 : : * case flags. If @case_flags is a %NULL pointer then ASCII letters
178 : : * are left as they are, and other code points are treated as
179 : : * unflagged.
180 : : * @output_length: The caller passes in the maximum number of ASCII
181 : : * code points that it can receive. On successful return it will
182 : : * contain the number of ASCII code points actually output.
183 : : * @output: An array of ASCII code points. It is *not*
184 : : * null-terminated; it will contain zeros if and only if the @input
185 : : * contains zeros. (Of course the caller can leave room for a
186 : : * terminator and add one if needed.)
187 : : *
188 : : * Converts a sequence of code points (presumed to be Unicode code
189 : : * points) to Punycode.
190 : : *
191 : : * Return value: The return value can be any of the #Punycode_status
192 : : * values defined above except %PUNYCODE_BAD_INPUT. If not
193 : : * %PUNYCODE_SUCCESS, then @output_size and @output might contain
194 : : * garbage.
195 : : **/
196 : : int
197 : 133 : punycode_encode (size_t input_length,
198 : : const punycode_uint input[],
199 : : const unsigned char case_flags[],
200 : : size_t * output_length, char output[])
201 : : {
202 : : punycode_uint input_len, n, delta, h, b, bias, j, m, q, k, t;
203 : : size_t out, max_out;
204 : :
205 : : /* The Punycode spec assumes that the input length is the same type */
206 : : /* of integer as a code point, so we need to convert the size_t to */
207 : : /* a punycode_uint, which could overflow. */
208 : :
209 [ - + ]: 133 : if (input_length > maxint)
210 : 0 : return punycode_overflow;
211 : 133 : input_len = (punycode_uint) input_length;
212 : :
213 : : /* Initialize the state: */
214 : :
215 : 133 : n = initial_n;
216 : 133 : delta = 0;
217 : 133 : out = 0;
218 : 133 : max_out = *output_length;
219 : 133 : bias = initial_bias;
220 : :
221 : : /* Handle the basic code points: */
222 : :
223 [ + + ]: 1710 : for (j = 0; j < input_len; ++j)
224 : : {
225 [ + + ]: 1577 : if (basic (input[j]))
226 : : {
227 [ - + ]: 641 : if (max_out - out < 2)
228 : 0 : return punycode_big_output;
229 [ - + ]: 641 : output[out++] = case_flags ?
230 : 641 : encode_basic (input[j], case_flags[j]) : (char) input[j];
231 : : }
232 : : /* else if (input[j] < n) return punycode_bad_input; */
233 : : /* (not needed for Punycode with unsigned code points) */
234 : : }
235 : :
236 : 133 : h = b = (punycode_uint) out;
237 : : /* cannot overflow because out <= input_len <= maxint */
238 : :
239 : : /* h is the number of code points that have been handled, b is the */
240 : : /* number of basic code points, and out is the number of ASCII code */
241 : : /* points that have been output. */
242 : :
243 [ + + ]: 133 : if (b > 0)
244 : 64 : output[out++] = delimiter;
245 : :
246 : : /* Main encoding loop: */
247 : :
248 [ + + ]: 894 : while (h < input_len)
249 : : {
250 : : /* All non-basic code points < n have been */
251 : : /* handled already. Find the next larger one: */
252 : :
253 [ + + ]: 13420 : for (m = maxint, j = 0; j < input_len; ++j)
254 : : {
255 : : /* if (basic(input[j])) continue; */
256 : : /* (not needed for Punycode) */
257 [ + + ][ + + ]: 12657 : if (input[j] >= n && input[j] < m)
258 : 1447 : m = input[j];
259 : : }
260 : :
261 : : /* Increase delta enough to advance the decoder's */
262 : : /* <n,i> state to <m,0>, but guard against overflow: */
263 : :
264 [ - + ]: 763 : if (m - n > (maxint - delta) / (h + 1))
265 : 0 : return punycode_overflow;
266 : 763 : delta += (m - n) * (h + 1);
267 : 763 : n = m;
268 : :
269 [ + + ]: 13414 : for (j = 0; j < input_len; ++j)
270 : : {
271 : : /* Punycode does not need to check whether input[j] is basic: */
272 [ + + ]: 12653 : if (input[j] < n /* || basic(input[j]) */ )
273 : : {
274 [ - + ]: 6890 : if (++delta == 0)
275 : 0 : return punycode_overflow;
276 : : }
277 : :
278 [ + + ]: 12653 : if (input[j] == n)
279 : : {
280 : : /* Represent delta as a generalized variable-length integer: */
281 : :
282 : 930 : for (q = delta, k = base;; k += base)
283 : : {
284 [ + + ]: 1859 : if (out >= max_out)
285 : 2 : return punycode_big_output;
286 [ + + ]: 3327 : t = k <= bias /* + tmin */ ? tmin : /* +tmin not needed */
287 [ + + ]: 1470 : k >= bias + tmax ? tmax : k - bias;
288 [ + + ]: 1857 : if (q < t)
289 : 928 : break;
290 : 929 : output[out++] = encode_digit (t + (q - t) % (base - t), 0);
291 : 929 : q = (q - t) / (base - t);
292 : 929 : }
293 : :
294 [ - + ][ # # ]: 928 : output[out++] = encode_digit (q, case_flags && case_flags[j]);
295 : 928 : bias = adapt (delta, h + 1, h == b);
296 : 928 : delta = 0;
297 : 928 : ++h;
298 : : }
299 : : }
300 : :
301 : 761 : ++delta, ++n;
302 : : }
303 : :
304 : 131 : *output_length = out;
305 : 133 : return punycode_success;
306 : : }
307 : :
308 : : /*** Main decode function ***/
309 : :
310 : : /**
311 : : * punycode_decode:
312 : : * @input_length: The number of ASCII code points in the @input array.
313 : : * @input: An array of ASCII code points (0..7F).
314 : : * @output_length: The caller passes in the maximum number of code
315 : : * points that it can receive into the @output array (which is also
316 : : * the maximum number of flags that it can receive into the
317 : : * @case_flags array, if @case_flags is not a %NULL pointer). On
318 : : * successful return it will contain the number of code points
319 : : * actually output (which is also the number of flags actually
320 : : * output, if case_flags is not a null pointer). The decoder will
321 : : * never need to output more code points than the number of ASCII
322 : : * code points in the input, because of the way the encoding is
323 : : * defined. The number of code points output cannot exceed the
324 : : * maximum possible value of a punycode_uint, even if the supplied
325 : : * @output_length is greater than that.
326 : : * @output: An array of code points like the input argument of
327 : : * punycode_encode() (see above).
328 : : * @case_flags: A %NULL pointer (if the flags are not needed by the
329 : : * caller) or an array of boolean values parallel to the @output
330 : : * array. Nonzero (true, flagged) suggests that the corresponding
331 : : * Unicode character be forced to uppercase by the caller (if
332 : : * possible), and zero (false, unflagged) suggests that it be forced
333 : : * to lowercase (if possible). ASCII code points (0..7F) are output
334 : : * already in the proper case, but their flags will be set
335 : : * appropriately so that applying the flags would be harmless.
336 : : *
337 : : * Converts Punycode to a sequence of code points (presumed to be
338 : : * Unicode code points).
339 : : *
340 : : * Return value: The return value can be any of the #Punycode_status
341 : : * values defined above. If not %PUNYCODE_SUCCESS, then
342 : : * @output_length, @output, and @case_flags might contain garbage.
343 : : *
344 : : **/
345 : : int
346 : 47 : punycode_decode (size_t input_length,
347 : : const char input[],
348 : : size_t * output_length,
349 : : punycode_uint output[], unsigned char case_flags[])
350 : : {
351 : : punycode_uint n, out, i, max_out, bias, oldi, w, k, digit, t;
352 : : size_t b, j, in;
353 : :
354 : : /* Initialize the state: */
355 : :
356 : 47 : n = initial_n;
357 : 47 : out = i = 0;
358 : 94 : max_out = *output_length > maxint ? maxint
359 [ + - ]: 47 : : (punycode_uint) * output_length;
360 : 47 : bias = initial_bias;
361 : :
362 : : /* Handle the basic code points: Let b be the number of input code */
363 : : /* points before the last delimiter, or 0 if there is none, then */
364 : : /* copy the first b code points to the output. */
365 : :
366 [ + + ]: 1454 : for (b = j = 0; j < input_length; ++j)
367 [ + + ]: 1407 : if (delim (input[j]))
368 : 40 : b = j;
369 [ - + ]: 47 : if (b > max_out)
370 : 0 : return punycode_big_output;
371 : :
372 [ + + ]: 326 : for (j = 0; j < b; ++j)
373 : : {
374 [ - + ]: 279 : if (case_flags)
375 : 0 : case_flags[out] = flagged (input[j]);
376 [ - + ]: 279 : if (!basic (input[j]))
377 : 0 : return punycode_bad_input;
378 : 279 : output[out++] = input[j];
379 : : }
380 : :
381 : : /* Main decoding loop: Start just after the last delimiter if any */
382 : : /* basic code points were copied; start at the beginning otherwise. */
383 : :
384 [ + + ][ + + ]: 571 : for (in = b > 0 ? b + 1 : 0; in < input_length; ++out)
385 : : {
386 : :
387 : : /* in is the index of the next ASCII code point to be consumed, */
388 : : /* and out is the number of code points in the output array. */
389 : :
390 : : /* Decode a generalized variable-length integer into delta, */
391 : : /* which gets added to i. The overflow checking is easier */
392 : : /* if we increase i as we go, then subtract off its starting */
393 : : /* value at the end to obtain delta. */
394 : :
395 : 525 : for (oldi = i, w = 1, k = base;; k += base)
396 : : {
397 [ - + ]: 1019 : if (in >= input_length)
398 : 0 : return punycode_bad_input;
399 : 1019 : digit = decode_digit (input[in++]);
400 [ + + ]: 1019 : if (digit >= base)
401 : 1 : return punycode_bad_input;
402 [ - + ]: 1018 : if (digit > (maxint - i) / w)
403 : 0 : return punycode_overflow;
404 : 1018 : i += digit * w;
405 [ + + ]: 1860 : t = k <= bias /* + tmin */ ? tmin : /* +tmin not needed */
406 [ + + ]: 842 : k >= bias + tmax ? tmax : k - bias;
407 [ + + ]: 1018 : if (digit < t)
408 : 524 : break;
409 [ - + ]: 494 : if (w > maxint / (base - t))
410 : 0 : return punycode_overflow;
411 : 494 : w *= (base - t);
412 : 494 : }
413 : :
414 : 524 : bias = adapt (i - oldi, out + 1, oldi == 0);
415 : :
416 : : /* i was supposed to wrap around from out+1 to 0, */
417 : : /* incrementing n each time, so we'll fix that now: */
418 : :
419 [ - + ]: 524 : if (i / (out + 1) > maxint - n)
420 : 0 : return punycode_overflow;
421 : 524 : n += i / (out + 1);
422 : 524 : i %= (out + 1);
423 : :
424 : : /* Insert n at position i of the output: */
425 : :
426 : : /* not needed for Punycode: */
427 : : /* if (basic(n)) return punycode_invalid_input; */
428 [ - + ]: 524 : if (out >= max_out)
429 : 0 : return punycode_big_output;
430 : :
431 [ - + ]: 524 : if (case_flags)
432 : : {
433 : 0 : memmove (case_flags + i + 1, case_flags + i, out - i);
434 : : /* Case of last ASCII code point determines case flag: */
435 : 0 : case_flags[i] = flagged (input[in - 1]);
436 : : }
437 : :
438 : 524 : memmove (output + i + 1, output + i, (out - i) * sizeof *output);
439 : 524 : output[i++] = n;
440 : : }
441 : :
442 : 46 : *output_length = (size_t) out;
443 : : /* cannot overflow because out <= old value of *output_length */
444 : 47 : return punycode_success;
445 : : }
446 : :
447 : : /**
448 : : * punycode_uint
449 : : *
450 : : * Unicode code point data type, this is always a 32 bit unsigned
451 : : * integer.
452 : : */
453 : :
454 : : /**
455 : : * Punycode_status
456 : : * @PUNYCODE_SUCCESS: Successful operation. This value is guaranteed
457 : : * to always be zero, the remaining ones are only guaranteed to hold
458 : : * non-zero values, for logical comparison purposes.
459 : : * @PUNYCODE_BAD_INPUT: Input is invalid.
460 : : * @PUNYCODE_BIG_OUTPUT: Output would exceed the space provided.
461 : : * @PUNYCODE_OVERFLOW: Input needs wider integers to process.
462 : : *
463 : : * Enumerated return codes of punycode_encode() and punycode_decode().
464 : : * The value 0 is guaranteed to always correspond to success.
465 : : */
|
