LCOV - code coverage report
Current view: top level - lib - punycode.c (source / functions) Hit Total Coverage
Test: GNU Libidn Lines: 94 111 84.7 %
Date: 2015-08-01 15:30:48 Functions: 5 6 83.3 %
Legend: Lines: hit not hit

          Line data    Source code
       1             : /* punycode.c --- Implementation of punycode used to ASCII encode IDN's.
       2             :    Copyright (C) 2002-2015 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             :  * downloaded from http://www.nicemice.net/idn/punycode-spec.gz on
      33             :  * 2015-03-02 with SHA1 a966a8017f6be579d74a50a226accc7607c40133, a
      34             :  * copy of which is stored in the GNU Libidn version controlled
      35             :  * repository under doc/specification/punycode-spec.gz.
      36             :  *
      37             :  * The changes compared to Adam's file include: re-indentation, adding
      38             :  * the license boilerplate and this comment, #include of config.h and
      39             :  * punycode.h, adding GTK-DOC comments, changing the return code of
      40             :  * punycode_encode and punycode_decode from enum to int, renaming the
      41             :  * input_length_orig function input variable to input_length (and
      42             :  * renaming the internal input_length variable to input_len) in
      43             :  * punycode_encode.
      44             :  *
      45             :  * Adam's file contains the following:
      46             :  *
      47             :  * punycode-sample.c 2.0.0 (2004-Mar-21-Sun)
      48             :  * http://www.nicemice.net/idn/
      49             :  * Adam M. Costello
      50             :  * http://www.nicemice.net/amc/
      51             :  *
      52             :  * This is ANSI C code (C89) implementing Punycode 1.0.x.
      53             :  *
      54             :  * Disclaimer and license: Regarding this entire document or any
      55             :  * portion of it (including the pseudocode and C code), the author
      56             :  * makes no guarantees and is not responsible for any damage resulting
      57             :  * from its use.  The author grants irrevocable permission to anyone
      58             :  * to use, modify, and distribute it in any way that does not diminish
      59             :  * the rights of anyone else to use, modify, and distribute it,
      60             :  * provided that redistributed derivative works do not contain
      61             :  * misleading author or version information.  Derivative works need
      62             :  * not be licensed under similar terms.
      63             :  */
      64             : 
      65             : #include <config.h>
      66             : 
      67             : /**********************************************************/
      68             : /* Implementation (would normally go in its own .c file): */
      69             : 
      70             : #include <string.h>
      71             : 
      72             : #include "punycode.h"
      73             : 
      74             : /*** Bootstring parameters for Punycode ***/
      75             : 
      76             : enum
      77             : { base = 36, tmin = 1, tmax = 26, skew = 38, damp = 700,
      78             :   initial_bias = 72, initial_n = 0x80, delimiter = 0x2D
      79             : };
      80             : 
      81             : /* basic(cp) tests whether cp is a basic code point: */
      82             : #define basic(cp) ((punycode_uint)(cp) < 0x80)
      83             : 
      84             : /* delim(cp) tests whether cp is a delimiter: */
      85             : #define delim(cp) ((cp) == delimiter)
      86             : 
      87             : /* decode_digit(cp) returns the numeric value of a basic code */
      88             : /* point (for use in representing integers) in the range 0 to */
      89             : /* base-1, or base if cp does not represent a value.          */
      90             : 
      91             : static punycode_uint
      92        1019 : decode_digit (punycode_uint cp)
      93             : {
      94        1682 :   return cp - 48 < 10 ? cp - 22 : cp - 65 < 26 ? cp - 65 :
      95         663 :     cp - 97 < 26 ? cp - 97 : base;
      96             : }
      97             : 
      98             : /* encode_digit(d,flag) returns the basic code point whose value      */
      99             : /* (when used for representing integers) is d, which needs to be in   */
     100             : /* the range 0 to base-1.  The lowercase form is used unless flag is  */
     101             : /* nonzero, in which case the uppercase form is used.  The behavior   */
     102             : /* is undefined if flag is nonzero and digit d has no uppercase form. */
     103             : 
     104             : static char
     105        1857 : encode_digit (punycode_uint d, int flag)
     106             : {
     107        1857 :   return d + 22 + 75 * (d < 26) - ((flag != 0) << 5);
     108             :   /*  0..25 map to ASCII a..z or A..Z */
     109             :   /* 26..35 map to ASCII 0..9         */
     110             : }
     111             : 
     112             : /* flagged(bcp) tests whether a basic code point is flagged */
     113             : /* (uppercase).  The behavior is undefined if bcp is not a  */
     114             : /* basic code point.                                        */
     115             : 
     116             : #define flagged(bcp) ((punycode_uint)(bcp) - 65 < 26)
     117             : 
     118             : /* encode_basic(bcp,flag) forces a basic code point to lowercase */
     119             : /* if flag is zero, uppercase if flag is nonzero, and returns    */
     120             : /* the resulting code point.  The code point is unchanged if it  */
     121             : /* is caseless.  The behavior is undefined if bcp is not a basic */
     122             : /* code point.                                                   */
     123             : 
     124             : static char
     125           0 : encode_basic (punycode_uint bcp, int flag)
     126             : {
     127           0 :   bcp -= (bcp - 97 < 26) << 5;
     128           0 :   return bcp + ((!flag && (bcp - 65 < 26)) << 5);
     129             : }
     130             : 
     131             : /*** Platform-specific constants ***/
     132             : 
     133             : /* maxint is the maximum value of a punycode_uint variable: */
     134             : static const punycode_uint maxint = -1;
     135             : /* Because maxint is unsigned, -1 becomes the maximum value. */
     136             : 
     137             : /*** Bias adaptation function ***/
     138             : 
     139             : static punycode_uint
     140        1452 : adapt (punycode_uint delta, punycode_uint numpoints, int firsttime)
     141             : {
     142             :   punycode_uint k;
     143             : 
     144        1452 :   delta = firsttime ? delta / damp : delta >> 1;
     145             :   /* delta >> 1 is a faster way of doing delta / 2 */
     146        1452 :   delta += delta / numpoints;
     147             : 
     148        1745 :   for (k = 0; delta > ((base - tmin) * tmax) / 2; k += base)
     149             :     {
     150         293 :       delta /= base - tmin;
     151             :     }
     152             : 
     153        1452 :   return k + (base - tmin + 1) * delta / (delta + skew);
     154             : }
     155             : 
     156             : /*** Main encode function ***/
     157             : 
     158             : /**
     159             :  * punycode_encode:
     160             :  * @input_length: The number of code points in the @input array and
     161             :  *   the number of flags in the @case_flags array.
     162             :  * @input: An array of code points.  They are presumed to be Unicode
     163             :  *   code points, but that is not strictly REQUIRED.  The array
     164             :  *   contains code points, not code units.  UTF-16 uses code units
     165             :  *   D800 through DFFF to refer to code points 10000..10FFFF.  The
     166             :  *   code points D800..DFFF do not occur in any valid Unicode string.
     167             :  *   The code points that can occur in Unicode strings (0..D7FF and
     168             :  *   E000..10FFFF) are also called Unicode scalar values.
     169             :  * @case_flags: A %NULL pointer or an array of boolean values parallel
     170             :  *   to the @input array.  Nonzero (true, flagged) suggests that the
     171             :  *   corresponding Unicode character be forced to uppercase after
     172             :  *   being decoded (if possible), and zero (false, unflagged) suggests
     173             :  *   that it be forced to lowercase (if possible).  ASCII code points
     174             :  *   (0..7F) are encoded literally, except that ASCII letters are
     175             :  *   forced to uppercase or lowercase according to the corresponding
     176             :  *   case flags.  If @case_flags is a %NULL pointer then ASCII letters
     177             :  *   are left as they are, and other code points are treated as
     178             :  *   unflagged.
     179             :  * @output_length: The caller passes in the maximum number of ASCII
     180             :  *   code points that it can receive.  On successful return it will
     181             :  *   contain the number of ASCII code points actually output.
     182             :  * @output: An array of ASCII code points.  It is *not*
     183             :  *   null-terminated; it will contain zeros if and only if the @input
     184             :  *   contains zeros.  (Of course the caller can leave room for a
     185             :  *   terminator and add one if needed.)
     186             :  *
     187             :  * Converts a sequence of code points (presumed to be Unicode code
     188             :  * points) to Punycode.
     189             :  *
     190             :  * Return value: The return value can be any of the #Punycode_status
     191             :  *   values defined above except %PUNYCODE_BAD_INPUT.  If not
     192             :  *   %PUNYCODE_SUCCESS, then @output_size and @output might contain
     193             :  *   garbage.
     194             :  **/
     195             : int
     196         133 : punycode_encode (size_t input_length,
     197             :                  const punycode_uint input[],
     198             :                  const unsigned char case_flags[],
     199             :                  size_t * output_length, char output[])
     200             : {
     201             :   punycode_uint input_len, n, delta, h, b, bias, j, m, q, k, t;
     202             :   size_t out, max_out;
     203             : 
     204             :   /* The Punycode spec assumes that the input length is the same type */
     205             :   /* of integer as a code point, so we need to convert the size_t to  */
     206             :   /* a punycode_uint, which could overflow.                           */
     207             : 
     208         133 :   if (input_length > maxint)
     209           0 :     return punycode_overflow;
     210         133 :   input_len = (punycode_uint) input_length;
     211             : 
     212             :   /* Initialize the state: */
     213             : 
     214         133 :   n = initial_n;
     215         133 :   delta = 0;
     216         133 :   out = 0;
     217         133 :   max_out = *output_length;
     218         133 :   bias = initial_bias;
     219             : 
     220             :   /* Handle the basic code points: */
     221             : 
     222        1710 :   for (j = 0; j < input_len; ++j)
     223             :     {
     224        1577 :       if (basic (input[j]))
     225             :         {
     226         641 :           if (max_out - out < 2)
     227           0 :             return punycode_big_output;
     228        1282 :           output[out++] = case_flags ?
     229         641 :             encode_basic (input[j], case_flags[j]) : (char) input[j];
     230             :         }
     231             :       /* else if (input[j] < n) return punycode_bad_input; */
     232             :       /* (not needed for Punycode with unsigned code points) */
     233             :     }
     234             : 
     235         133 :   h = b = (punycode_uint) out;
     236             :   /* cannot overflow because out <= input_len <= maxint */
     237             : 
     238             :   /* h is the number of code points that have been handled, b is the  */
     239             :   /* number of basic code points, and out is the number of ASCII code */
     240             :   /* points that have been output.                                    */
     241             : 
     242         133 :   if (b > 0)
     243          64 :     output[out++] = delimiter;
     244             : 
     245             :   /* Main encoding loop: */
     246             : 
     247        1027 :   while (h < input_len)
     248             :     {
     249             :       /* All non-basic code points < n have been     */
     250             :       /* handled already.  Find the next larger one: */
     251             : 
     252       13420 :       for (m = maxint, j = 0; j < input_len; ++j)
     253             :         {
     254             :           /* if (basic(input[j])) continue; */
     255             :           /* (not needed for Punycode) */
     256       12657 :           if (input[j] >= n && input[j] < m)
     257        1447 :             m = input[j];
     258             :         }
     259             : 
     260             :       /* Increase delta enough to advance the decoder's    */
     261             :       /* <n,i> state to <m,0>, but guard against overflow: */
     262             : 
     263         763 :       if (m - n > (maxint - delta) / (h + 1))
     264           0 :         return punycode_overflow;
     265         763 :       delta += (m - n) * (h + 1);
     266         763 :       n = m;
     267             : 
     268       13414 :       for (j = 0; j < input_len; ++j)
     269             :         {
     270             :           /* Punycode does not need to check whether input[j] is basic: */
     271       12653 :           if (input[j] < n /* || basic(input[j]) */ )
     272             :             {
     273        6890 :               if (++delta == 0)
     274           0 :                 return punycode_overflow;
     275             :             }
     276             : 
     277       12653 :           if (input[j] == n)
     278             :             {
     279             :               /* Represent delta as a generalized variable-length integer: */
     280             : 
     281        1859 :               for (q = delta, k = base;; k += base)
     282             :                 {
     283        1859 :                   if (out >= max_out)
     284           2 :                     return punycode_big_output;
     285        3327 :                   t = k <= bias /* + tmin */ ? tmin :        /* +tmin not needed */
     286        1470 :                     k >= bias + tmax ? tmax : k - bias;
     287        1857 :                   if (q < t)
     288         928 :                     break;
     289         929 :                   output[out++] = encode_digit (t + (q - t) % (base - t), 0);
     290         929 :                   q = (q - t) / (base - t);
     291         929 :                 }
     292             : 
     293         928 :               output[out++] = encode_digit (q, case_flags && case_flags[j]);
     294         928 :               bias = adapt (delta, h + 1, h == b);
     295         928 :               delta = 0;
     296         928 :               ++h;
     297             :             }
     298             :         }
     299             : 
     300         761 :       ++delta, ++n;
     301             :     }
     302             : 
     303         131 :   *output_length = out;
     304         131 :   return punycode_success;
     305             : }
     306             : 
     307             : /*** Main decode function ***/
     308             : 
     309             : /**
     310             :  * punycode_decode:
     311             :  * @input_length: The number of ASCII code points in the @input array.
     312             :  * @input: An array of ASCII code points (0..7F).
     313             :  * @output_length: The caller passes in the maximum number of code
     314             :  *   points that it can receive into the @output array (which is also
     315             :  *   the maximum number of flags that it can receive into the
     316             :  *   @case_flags array, if @case_flags is not a %NULL pointer).  On
     317             :  *   successful return it will contain the number of code points
     318             :  *   actually output (which is also the number of flags actually
     319             :  *   output, if case_flags is not a null pointer).  The decoder will
     320             :  *   never need to output more code points than the number of ASCII
     321             :  *   code points in the input, because of the way the encoding is
     322             :  *   defined.  The number of code points output cannot exceed the
     323             :  *   maximum possible value of a punycode_uint, even if the supplied
     324             :  *   @output_length is greater than that.
     325             :  * @output: An array of code points like the input argument of
     326             :  *   punycode_encode() (see above).
     327             :  * @case_flags: A %NULL pointer (if the flags are not needed by the
     328             :  *   caller) or an array of boolean values parallel to the @output
     329             :  *   array.  Nonzero (true, flagged) suggests that the corresponding
     330             :  *   Unicode character be forced to uppercase by the caller (if
     331             :  *   possible), and zero (false, unflagged) suggests that it be forced
     332             :  *   to lowercase (if possible).  ASCII code points (0..7F) are output
     333             :  *   already in the proper case, but their flags will be set
     334             :  *   appropriately so that applying the flags would be harmless.
     335             :  *
     336             :  * Converts Punycode to a sequence of code points (presumed to be
     337             :  * Unicode code points).
     338             :  *
     339             :  * Return value: The return value can be any of the #Punycode_status
     340             :  *   values defined above.  If not %PUNYCODE_SUCCESS, then
     341             :  *   @output_length, @output, and @case_flags might contain garbage.
     342             :  *
     343             :  **/
     344             : int
     345          47 : punycode_decode (size_t input_length,
     346             :                  const char input[],
     347             :                  size_t * output_length,
     348             :                  punycode_uint output[], unsigned char case_flags[])
     349             : {
     350             :   punycode_uint n, out, i, max_out, bias, oldi, w, k, digit, t;
     351             :   size_t b, j, in;
     352             : 
     353             :   /* Initialize the state: */
     354             : 
     355          47 :   n = initial_n;
     356          47 :   out = i = 0;
     357          94 :   max_out = *output_length > maxint ? maxint
     358          47 :     : (punycode_uint) * output_length;
     359          47 :   bias = initial_bias;
     360             : 
     361             :   /* Handle the basic code points:  Let b be the number of input code */
     362             :   /* points before the last delimiter, or 0 if there is none, then    */
     363             :   /* copy the first b code points to the output.                      */
     364             : 
     365        1454 :   for (b = j = 0; j < input_length; ++j)
     366        1407 :     if (delim (input[j]))
     367          40 :       b = j;
     368          47 :   if (b > max_out)
     369           0 :     return punycode_big_output;
     370             : 
     371         326 :   for (j = 0; j < b; ++j)
     372             :     {
     373         279 :       if (case_flags)
     374           0 :         case_flags[out] = flagged (input[j]);
     375         279 :       if (!basic (input[j]))
     376           0 :         return punycode_bad_input;
     377         279 :       output[out++] = input[j];
     378             :     }
     379             : 
     380             :   /* Main decoding loop:  Start just after the last delimiter if any  */
     381             :   /* basic code points were copied; start at the beginning otherwise. */
     382             : 
     383         571 :   for (in = b > 0 ? b + 1 : 0; in < input_length; ++out)
     384             :     {
     385             : 
     386             :       /* in is the index of the next ASCII code point to be consumed, */
     387             :       /* and out is the number of code points in the output array.    */
     388             : 
     389             :       /* Decode a generalized variable-length integer into delta,  */
     390             :       /* which gets added to i.  The overflow checking is easier   */
     391             :       /* if we increase i as we go, then subtract off its starting */
     392             :       /* value at the end to obtain delta.                         */
     393             : 
     394        1019 :       for (oldi = i, w = 1, k = base;; k += base)
     395             :         {
     396        1019 :           if (in >= input_length)
     397           0 :             return punycode_bad_input;
     398        1019 :           digit = decode_digit (input[in++]);
     399        1019 :           if (digit >= base)
     400           1 :             return punycode_bad_input;
     401        1018 :           if (digit > (maxint - i) / w)
     402           0 :             return punycode_overflow;
     403        1018 :           i += digit * w;
     404        1860 :           t = k <= bias /* + tmin */ ? tmin :        /* +tmin not needed */
     405         842 :             k >= bias + tmax ? tmax : k - bias;
     406        1018 :           if (digit < t)
     407         524 :             break;
     408         494 :           if (w > maxint / (base - t))
     409           0 :             return punycode_overflow;
     410         494 :           w *= (base - t);
     411         494 :         }
     412             : 
     413         524 :       bias = adapt (i - oldi, out + 1, oldi == 0);
     414             : 
     415             :       /* i was supposed to wrap around from out+1 to 0,   */
     416             :       /* incrementing n each time, so we'll fix that now: */
     417             : 
     418         524 :       if (i / (out + 1) > maxint - n)
     419           0 :         return punycode_overflow;
     420         524 :       n += i / (out + 1);
     421         524 :       i %= (out + 1);
     422             : 
     423             :       /* Insert n at position i of the output: */
     424             : 
     425             :       /* not needed for Punycode: */
     426             :       /* if (basic(n)) return punycode_bad_input; */
     427         524 :       if (out >= max_out)
     428           0 :         return punycode_big_output;
     429             : 
     430         524 :       if (case_flags)
     431             :         {
     432           0 :           memmove (case_flags + i + 1, case_flags + i, out - i);
     433             :           /* Case of last ASCII code point determines case flag: */
     434           0 :           case_flags[i] = flagged (input[in - 1]);
     435             :         }
     436             : 
     437         524 :       memmove (output + i + 1, output + i, (out - i) * sizeof *output);
     438         524 :       output[i++] = n;
     439             :     }
     440             : 
     441          46 :   *output_length = (size_t) out;
     442             :   /* cannot overflow because out <= old value of *output_length */
     443          46 :   return punycode_success;
     444             : }
     445             : 
     446             : /**
     447             :  * punycode_uint
     448             :  *
     449             :  * Unicode code point data type, this is always a 32 bit unsigned
     450             :  * integer.
     451             :  */
     452             : 
     453             : /**
     454             :  * Punycode_status
     455             :  * @PUNYCODE_SUCCESS: Successful operation.  This value is guaranteed
     456             :  *   to always be zero, the remaining ones are only guaranteed to hold
     457             :  *   non-zero values, for logical comparison purposes.
     458             :  * @PUNYCODE_BAD_INPUT: Input is invalid.
     459             :  * @PUNYCODE_BIG_OUTPUT: Output would exceed the space provided.
     460             :  * @PUNYCODE_OVERFLOW: Input needs wider integers to process.
     461             :  *
     462             :  * Enumerated return codes of punycode_encode() and punycode_decode().
     463             :  * The value 0 is guaranteed to always correspond to success.
     464             :  */

Generated by: LCOV version 1.11