Annotation of parser3/src/lib/punycode/pa_punycode.c, revision 1.4
1.1 moko 1: /* punycode.c --- Implementation of punycode used to ASCII encode IDN's.
2: Copyright (C) 2002-2013 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 "pa_punycode.h"
71:
72: /*** Bootstring parameters for Punycode ***/
73:
74: enum
75: { base = 36, tmin = 1, tmax = 26, skew = 38, damp = 700,
76: initial_bias = 72, initial_n = 0x80, delimiter = 0x2D
77: };
78:
79: /* basic(cp) tests whether cp is a basic code point: */
80: #define basic(cp) ((punycode_uint)(cp) < 0x80)
81:
82: /* delim(cp) tests whether cp is a delimiter: */
83: #define delim(cp) ((cp) == delimiter)
84:
85: /* decode_digit(cp) returns the numeric value of a basic code */
86: /* point (for use in representing integers) in the range 0 to */
87: /* base-1, or base if cp does not represent a value. */
88:
1.2 moko 89: static punycode_uint decode_digit (punycode_uint cp)
1.1 moko 90: {
91: return cp - 48 < 10 ? cp - 22 : cp - 65 < 26 ? cp - 65 :
92: cp - 97 < 26 ? cp - 97 : base;
93: }
94:
95: /* encode_digit(d,flag) returns the basic code point whose value */
96: /* (when used for representing integers) is d, which needs to be in */
97: /* the range 0 to base-1. The lowercase form is used unless flag is */
98: /* nonzero, in which case the uppercase form is used. The behavior */
99: /* is undefined if flag is nonzero and digit d has no uppercase form. */
100:
1.2 moko 101: static char encode_digit (punycode_uint d, int flag)
1.1 moko 102: {
1.2 moko 103: return (char)(d + 22 + 75 * (d < 26) - ((flag != 0) << 5));
1.1 moko 104: /* 0..25 map to ASCII a..z or A..Z */
105: /* 26..35 map to ASCII 0..9 */
106: }
107:
108: /* flagged(bcp) tests whether a basic code point is flagged */
109: /* (uppercase). The behavior is undefined if bcp is not a */
110: /* basic code point. */
111:
1.4 ! moko 112: #define flagged(bcp) (unsigned char)((punycode_uint)(bcp) - 65 < 26)
1.1 moko 113:
114: /* encode_basic(bcp,flag) forces a basic code point to lowercase */
115: /* if flag is zero, uppercase if flag is nonzero, and returns */
116: /* the resulting code point. The code point is unchanged if it */
117: /* is caseless. The behavior is undefined if bcp is not a basic */
118: /* code point. */
119:
1.2 moko 120: static char encode_basic (punycode_uint bcp, int flag)
1.1 moko 121: {
122: bcp -= (bcp - 97 < 26) << 5;
1.2 moko 123: return (char)(bcp + ((!flag && (bcp - 65 < 26)) << 5));
1.1 moko 124: }
125:
126: /*** Platform-specific constants ***/
127:
128: /* maxint is the maximum value of a punycode_uint variable: */
1.2 moko 129: static const punycode_uint maxint = 0xFFFFFFFF;
1.1 moko 130: /* Because maxint is unsigned, -1 becomes the maximum value. */
131:
132: /*** Bias adaptation function ***/
133:
1.2 moko 134: static punycode_uint adapt (punycode_uint delta, punycode_uint numpoints, int firsttime)
1.1 moko 135: {
136: punycode_uint k;
137:
138: delta = firsttime ? delta / damp : delta >> 1;
139: /* delta >> 1 is a faster way of doing delta / 2 */
140: delta += delta / numpoints;
141:
142: for (k = 0; delta > ((base - tmin) * tmax) / 2; k += base)
143: {
144: delta /= base - tmin;
145: }
146:
147: return k + (base - tmin + 1) * delta / (delta + skew);
148: }
149:
150: /*** Main encode function ***/
151:
152: /**
153: * punycode_encode:
154: * @input_length: The number of code points in the @input array and
155: * the number of flags in the @case_flags array.
156: * @input: An array of code points. They are presumed to be Unicode
157: * code points, but that is not strictly REQUIRED. The array
158: * contains code points, not code units. UTF-16 uses code units
159: * D800 through DFFF to refer to code points 10000..10FFFF. The
160: * code points D800..DFFF do not occur in any valid Unicode string.
161: * The code points that can occur in Unicode strings (0..D7FF and
162: * E000..10FFFF) are also called Unicode scalar values.
163: * @case_flags: A %NULL pointer or an array of boolean values parallel
164: * to the @input array. Nonzero (true, flagged) suggests that the
165: * corresponding Unicode character be forced to uppercase after
166: * being decoded (if possible), and zero (false, unflagged) suggests
167: * that it be forced to lowercase (if possible). ASCII code points
168: * (0..7F) are encoded literally, except that ASCII letters are
169: * forced to uppercase or lowercase according to the corresponding
170: * case flags. If @case_flags is a %NULL pointer then ASCII letters
171: * are left as they are, and other code points are treated as
172: * unflagged.
173: * @output_length: The caller passes in the maximum number of ASCII
174: * code points that it can receive. On successful return it will
175: * contain the number of ASCII code points actually output.
176: * @output: An array of ASCII code points. It is *not*
177: * null-terminated; it will contain zeros if and only if the @input
178: * contains zeros. (Of course the caller can leave room for a
179: * terminator and add one if needed.)
180: *
181: * Converts a sequence of code points (presumed to be Unicode code
182: * points) to Punycode.
183: *
184: * Return value: The return value can be any of the #Punycode_status
185: * values defined above except %PUNYCODE_BAD_INPUT. If not
186: * %PUNYCODE_SUCCESS, then @output_size and @output might contain
187: * garbage.
188: **/
1.2 moko 189: int punycode_encode (size_t input_length,
1.1 moko 190: const punycode_uint input[],
191: const unsigned char case_flags[],
192: size_t * output_length, char output[])
193: {
194: punycode_uint input_len, n, delta, h, b, bias, j, m, q, k, t;
195: size_t out, max_out;
196:
197: /* The Punycode spec assumes that the input length is the same type */
198: /* of integer as a code point, so we need to convert the size_t to */
199: /* a punycode_uint, which could overflow. */
200:
201: if (input_length > maxint)
202: return punycode_overflow;
203: input_len = (punycode_uint) input_length;
204:
205: /* Initialize the state: */
206:
207: n = initial_n;
208: delta = 0;
209: out = 0;
210: max_out = *output_length;
211: bias = initial_bias;
212:
213: /* Handle the basic code points: */
214:
215: for (j = 0; j < input_len; ++j)
216: {
217: if (basic (input[j]))
218: {
219: if (max_out - out < 2)
220: return punycode_big_output;
221: output[out++] = case_flags ?
222: encode_basic (input[j], case_flags[j]) : (char) input[j];
223: }
224: /* else if (input[j] < n) return punycode_bad_input; */
225: /* (not needed for Punycode with unsigned code points) */
226: }
227:
228: h = b = (punycode_uint) out;
229: /* cannot overflow because out <= input_len <= maxint */
230:
231: /* h is the number of code points that have been handled, b is the */
232: /* number of basic code points, and out is the number of ASCII code */
233: /* points that have been output. */
234:
235: if (b > 0)
236: output[out++] = delimiter;
237:
238: /* Main encoding loop: */
239:
240: while (h < input_len)
241: {
242: /* All non-basic code points < n have been */
243: /* handled already. Find the next larger one: */
244:
245: for (m = maxint, j = 0; j < input_len; ++j)
246: {
247: /* if (basic(input[j])) continue; */
248: /* (not needed for Punycode) */
249: if (input[j] >= n && input[j] < m)
250: m = input[j];
251: }
252:
253: /* Increase delta enough to advance the decoder's */
254: /* <n,i> state to <m,0>, but guard against overflow: */
255:
256: if (m - n > (maxint - delta) / (h + 1))
257: return punycode_overflow;
258: delta += (m - n) * (h + 1);
259: n = m;
260:
261: for (j = 0; j < input_len; ++j)
262: {
263: /* Punycode does not need to check whether input[j] is basic: */
264: if (input[j] < n /* || basic(input[j]) */ )
265: {
266: if (++delta == 0)
267: return punycode_overflow;
268: }
269:
270: if (input[j] == n)
271: {
272: /* Represent delta as a generalized variable-length integer: */
273:
274: for (q = delta, k = base;; k += base)
275: {
276: if (out >= max_out)
277: return punycode_big_output;
278: t = k <= bias /* + tmin */ ? tmin : /* +tmin not needed */
279: k >= bias + tmax ? tmax : k - bias;
280: if (q < t)
281: break;
282: output[out++] = encode_digit (t + (q - t) % (base - t), 0);
283: q = (q - t) / (base - t);
284: }
285:
286: output[out++] = encode_digit (q, case_flags && case_flags[j]);
287: bias = adapt (delta, h + 1, h == b);
288: delta = 0;
289: ++h;
290: }
291: }
292:
293: ++delta, ++n;
294: }
295:
296: *output_length = out;
297: return punycode_success;
298: }
299:
300: /*** Main decode function ***/
301:
302: /**
303: * punycode_decode:
304: * @input_length: The number of ASCII code points in the @input array.
305: * @input: An array of ASCII code points (0..7F).
306: * @output_length: The caller passes in the maximum number of code
307: * points that it can receive into the @output array (which is also
308: * the maximum number of flags that it can receive into the
309: * @case_flags array, if @case_flags is not a %NULL pointer). On
310: * successful return it will contain the number of code points
311: * actually output (which is also the number of flags actually
312: * output, if case_flags is not a null pointer). The decoder will
313: * never need to output more code points than the number of ASCII
314: * code points in the input, because of the way the encoding is
315: * defined. The number of code points output cannot exceed the
316: * maximum possible value of a punycode_uint, even if the supplied
317: * @output_length is greater than that.
318: * @output: An array of code points like the input argument of
319: * punycode_encode() (see above).
320: * @case_flags: A %NULL pointer (if the flags are not needed by the
321: * caller) or an array of boolean values parallel to the @output
322: * array. Nonzero (true, flagged) suggests that the corresponding
323: * Unicode character be forced to uppercase by the caller (if
324: * possible), and zero (false, unflagged) suggests that it be forced
325: * to lowercase (if possible). ASCII code points (0..7F) are output
326: * already in the proper case, but their flags will be set
327: * appropriately so that applying the flags would be harmless.
328: *
329: * Converts Punycode to a sequence of code points (presumed to be
330: * Unicode code points).
331: *
332: * Return value: The return value can be any of the #Punycode_status
333: * values defined above. If not %PUNYCODE_SUCCESS, then
334: * @output_length, @output, and @case_flags might contain garbage.
335: *
336: **/
1.2 moko 337: int punycode_decode (size_t input_length,
1.1 moko 338: const char input[],
339: size_t * output_length,
340: punycode_uint output[], unsigned char case_flags[])
341: {
342: punycode_uint n, out, i, max_out, bias, oldi, w, k, digit, t;
343: size_t b, j, in;
344:
345: /* Initialize the state: */
346:
347: n = initial_n;
348: out = i = 0;
349: max_out = *output_length > maxint ? maxint
350: : (punycode_uint) * output_length;
351: bias = initial_bias;
352:
353: /* Handle the basic code points: Let b be the number of input code */
354: /* points before the last delimiter, or 0 if there is none, then */
355: /* copy the first b code points to the output. */
356:
357: for (b = j = 0; j < input_length; ++j)
358: if (delim (input[j]))
359: b = j;
360: if (b > max_out)
361: return punycode_big_output;
362:
363: for (j = 0; j < b; ++j)
364: {
365: if (case_flags)
366: case_flags[out] = flagged (input[j]);
367: if (!basic (input[j]))
368: return punycode_bad_input;
369: output[out++] = input[j];
370: }
371:
372: /* Main decoding loop: Start just after the last delimiter if any */
373: /* basic code points were copied; start at the beginning otherwise. */
374:
375: for (in = b > 0 ? b + 1 : 0; in < input_length; ++out)
376: {
377:
378: /* in is the index of the next ASCII code point to be consumed, */
379: /* and out is the number of code points in the output array. */
380:
381: /* Decode a generalized variable-length integer into delta, */
382: /* which gets added to i. The overflow checking is easier */
383: /* if we increase i as we go, then subtract off its starting */
384: /* value at the end to obtain delta. */
385:
386: for (oldi = i, w = 1, k = base;; k += base)
387: {
388: if (in >= input_length)
389: return punycode_bad_input;
390: digit = decode_digit (input[in++]);
391: if (digit >= base)
392: return punycode_bad_input;
393: if (digit > (maxint - i) / w)
394: return punycode_overflow;
395: i += digit * w;
396: t = k <= bias /* + tmin */ ? tmin : /* +tmin not needed */
397: k >= bias + tmax ? tmax : k - bias;
398: if (digit < t)
399: break;
400: if (w > maxint / (base - t))
401: return punycode_overflow;
402: w *= (base - t);
403: }
404:
405: bias = adapt (i - oldi, out + 1, oldi == 0);
406:
407: /* i was supposed to wrap around from out+1 to 0, */
408: /* incrementing n each time, so we'll fix that now: */
409:
410: if (i / (out + 1) > maxint - n)
411: return punycode_overflow;
412: n += i / (out + 1);
413: i %= (out + 1);
414:
415: /* Insert n at position i of the output: */
416:
417: /* not needed for Punycode: */
418: /* if (basic(n)) return punycode_invalid_input; */
419: if (out >= max_out)
420: return punycode_big_output;
421:
422: if (case_flags)
423: {
424: memmove (case_flags + i + 1, case_flags + i, out - i);
425: /* Case of last ASCII code point determines case flag: */
1.4 ! moko 426: case_flags[i] = flagged (input[in - 1]);
1.1 moko 427: }
428:
429: memmove (output + i + 1, output + i, (out - i) * sizeof *output);
430: output[i++] = n;
431: }
432:
433: *output_length = (size_t) out;
434: /* cannot overflow because out <= old value of *output_length */
435: return punycode_success;
436: }
437:
438: /**
439: * punycode_uint
440: *
441: * Unicode code point data type, this is always a 32 bit unsigned
442: * integer.
443: */
444:
445: /**
446: * Punycode_status
447: * @PUNYCODE_SUCCESS: Successful operation. This value is guaranteed
448: * to always be zero, the remaining ones are only guaranteed to hold
449: * non-zero values, for logical comparison purposes.
450: * @PUNYCODE_BAD_INPUT: Input is invalid.
451: * @PUNYCODE_BIG_OUTPUT: Output would exceed the space provided.
452: * @PUNYCODE_OVERFLOW: Input needs wider integers to process.
453: *
454: * Enumerated return codes of punycode_encode() and punycode_decode().
455: * The value 0 is guaranteed to always correspond to success.
456: */
E-mail:
![[BACK]](/cvsweb/icons/back.gif){kind=icon}
Return to pa_punycode.c CVS log ![[TXT]](/cvsweb/icons/text.gif){kind=icon}
![[DIR]](/cvsweb/icons/dir.gif){kind=icon}
Up to [parser3project] / parser3 / src / lib / punycode