icuSources/test/intltest/punyref.cpp

   1 // © 2016 and later: Unicode, Inc. and others.
   2 // License & terms of use: http://www.unicode.org/copyright.html
   3 /*
   4  *******************************************************************************
   5  *
   6  *   Copyright (C) 2003-2005, International Business Machines
   7  *   Corporation and others.  All Rights Reserved.
   8  *
   9  *******************************************************************************
  10  *   file name:  punyref.cpp
  11  *   encoding:   UTF-8
  12  *   tab size:   8 (not used)
  13  *   indentation:4
  14  *
  15  *   created on: 2003feb1
  16  *   created by: Ram Viswanadha
  17  */
  18
  19 /*
  20 Disclaimer and license
  21
  22     Regarding this entire document or any portion of it (including
  23     the pseudocode and C code), the author makes no guarantees and
  24     is not responsible for any damage resulting from its use.  The
  25     author grants irrevocable permission to anyone to use, modify,
  26     and distribute it in any way that does not diminish the rights
  27     of anyone else to use, modify, and distribute it, provided that
  28     redistributed derivative works do not contain misleading author or
  29     version information.  Derivative works need not be licensed under
  30     similar terms.
  31
  32 punycode.c 0.4.0 (2001-Nov-17-Sat)
  33 http://www.cs.berkeley.edu/~amc/idn/
  34 Adam M. Costello
  35 http://www.nicemice.net/amc/
  36 */
  37
  38 /**********************************************************/
  39 /* Implementation (would normally go in its own .c file): */
  40
  41 #include <string.h>
  42
  43 #include "unicode/utypes.h"
  44
  45 #if !UCONFIG_NO_IDNA
  46
  47 #include "punyref.h"
  48
  49 /*** Bootstring parameters for Punycode ***/
  50
  51 enum { base = 36, tmin = 1, tmax = 26, skew = 38, damp = 700,
  52        initial_bias = 72, initial_n = 0x80, delimiter = 0x2D };
  53
  54 /* basic(cp) tests whether cp is a basic code point: */
  55 #define basic(cp) ((punycode_uint)(cp) < 0x80)
  56
  57 /* delim(cp) tests whether cp is a delimiter: */
  58 #define delim(cp) ((cp) == delimiter)
  59
  60 U_CDECL_BEGIN
  61 /* decode_digit(cp) returns the numeric value of a basic code */
  62 /* point (for use in representing integers) in the range 0 to */
  63 /* base-1, or base if cp is does not represent a value.       */
  64
  65 static punycode_uint decode_digit(punycode_uint cp)
  66 {
  67   return  cp - 48 < 10 ? cp - 22 :  cp - 65 < 26 ? cp - 65 :
  68           cp - 97 < 26 ? cp - 97 :  base;
  69 }
  70
  71 /* encode_digit(d,flag) returns the basic code point whose value      */
  72 /* (when used for representing integers) is d, which needs to be in   */
  73 /* the range 0 to base-1.  The lowercase form is used unless flag is  */
  74 /* nonzero, in which case the uppercase form is used.  The behavior   */
  75 /* is undefined if flag is nonzero and digit d has no uppercase form. */
  76
  77 static char encode_digit(punycode_uint d, int flag)
  78 {
  79   return (char) d + 22 + 75 * (d < 26) - ((flag != 0) << 5);
  80   /*  0..25 map to ASCII a..z or A..Z */
  81   /* 26..35 map to ASCII 0..9         */
  82 }
  83
  84 /* flagged(bcp) tests whether a basic code point is flagged */
  85 /* (uppercase).  The behavior is undefined if bcp is not a  */
  86 /* basic code point.                                        */
  87
  88 #define flagged(bcp) ((punycode_uint)(bcp) - 65 < 26)
  89
  90 /* encode_basic(bcp,flag) forces a basic code point to lowercase */
  91 /* if flag is zero, uppercase if flag is nonzero, and returns    */
  92 /* the resulting code point.  The code point is unchanged if it  */
  93 /* is caseless.  The behavior is undefined if bcp is not a basic */
  94 /* code point.                                                   */
  95
  96 static char encode_basic(punycode_uint bcp, int flag)
  97 {
  98   bcp -= (bcp - 97 < 26) << 5;
  99   return (char) bcp + ((!flag && (bcp - 65 < 26)) << 5);
 100 }
 101
 102 /*** Platform-specific constants ***/
 103
 104 /* maxint is the maximum value of a punycode_uint variable: */
 105 static const punycode_uint maxint = (punycode_uint) (-1);
 106 /* Because maxint is unsigned, -1 becomes the maximum value. */
 107
 108 /*** Bias adaptation function ***/
 109
 110 static punycode_uint adapt(
 111   punycode_uint delta, punycode_uint numpoints, int firsttime )
 112 {
 113   punycode_uint k;
 114
 115   delta = firsttime ? delta / damp : delta >> 1;
 116   /* delta >> 1 is a faster way of doing delta / 2 */
 117   delta += delta / numpoints;
 118
 119   for (k = 0;  delta > ((base - tmin) * tmax) / 2;  k += base) {
 120     delta /= base - tmin;
 121   }
 122
 123   return k + (base - tmin + 1) * delta / (delta + skew);
 124 }
 125
 126 /*** Main encode function ***/
 127
 128 enum punycode_status punycode_encode(
 129   punycode_uint input_length,
 130   const punycode_uint input[],
 131   const unsigned char case_flags[],
 132   punycode_uint *output_length,
 133   char output[] )
 134 {
 135   punycode_uint n, delta, h, b, out, max_out, bias, j, m, q, k, t;
 136
 137   /* Initialize the state: */
 138
 139   n = initial_n;
 140   delta = out = 0;
 141   max_out = *output_length;
 142   bias = initial_bias;
 143
 144   /* Handle the basic code points: */
 145
 146   for (j = 0;  j < input_length;  ++j) {
 147     if (basic(input[j])) {
 148       if (max_out - out < 2) return punycode_big_output;
 149       output[out++] = (char)
 150         (case_flags ?  encode_basic(input[j], case_flags[j]) : input[j]);
 151     }
 152     /* else if (input[j] < n) return punycode_bad_input; */
 153     /* (not needed for Punycode with unsigned code points) */
 154   }
 155
 156   h = b = out;
 157
 158   /* h is the number of code points that have been handled, b is the  */
 159   /* number of basic code points, and out is the number of characters */
 160   /* that have been output.                                           */
 161
 162   if (b > 0) output[out++] = delimiter;
 163
 164   /* Main encoding loop: */
 165
 166   while (h < input_length) {
 167     /* All non-basic code points < n have been     */
 168     /* handled already.  Find the next larger one: */
 169
 170     for (m = maxint, j = 0;  j < input_length;  ++j) {
 171       /* if (basic(input[j])) continue; */
 172       /* (not needed for Punycode) */
 173       if (input[j] >= n && input[j] < m) m = input[j];
 174     }
 175
 176     /* Increase delta enough to advance the decoder's    */
 177     /* <n,i> state to <m,0>, but guard against overflow: */
 178
 179     if (m - n > (maxint - delta) / (h + 1)) return punycode_overflow;
 180     delta += (m - n) * (h + 1);
 181     n = m;
 182
 183     for (j = 0;  j < input_length;  ++j) {
 184       /* Punycode does not need to check whether input[j] is basic: */
 185       if (input[j] < n /* || basic(input[j]) */ ) {
 186         if (++delta == 0) return punycode_overflow;
 187       }
 188
 189       if (input[j] == n) {
 190         /* Represent delta as a generalized variable-length integer: */
 191
 192         for (q = delta, k = base;  ;  k += base) {
 193           if (out >= max_out) return punycode_big_output;
 194           t = k <= bias /* + tmin */ ? tmin :     /* +tmin not needed */
 195               k >= bias + tmax ? tmax : k - bias;
 196           if (q < t) break;
 197           output[out++] = encode_digit(t + (q - t) % (base - t), 0);
 198           q = (q - t) / (base - t);
 199         }
 200
 201         output[out++] = encode_digit(q, case_flags && case_flags[j]);
 202         bias = adapt(delta, h + 1, h == b);
 203         delta = 0;
 204         ++h;
 205       }
 206     }
 207
 208     ++delta, ++n;
 209   }
 210
 211   *output_length = out;
 212   return punycode_success;
 213 }
 214
 215 /*** Main decode function ***/
 216
 217 enum punycode_status punycode_decode(
 218   punycode_uint input_length,
 219   const char input[],
 220   punycode_uint *output_length,
 221   punycode_uint output[],
 222   unsigned char case_flags[] )
 223 {
 224   punycode_uint n, out, i, max_out, bias,
 225                  b, j, in, oldi, w, k, digit, t;
 226
 227   /* Initialize the state: */
 228
 229   n = initial_n;
 230   out = i = 0;
 231   max_out = *output_length;
 232   bias = initial_bias;
 233
 234   /* Handle the basic code points:  Let b be the number of input code */
 235   /* points before the last delimiter, or 0 if there is none, then    */
 236   /* copy the first b code points to the output.                      */
 237
 238   for (b = j = 0;  j < input_length;  ++j) if (delim(input[j])) b = j;
 239   if (b > max_out) return punycode_big_output;
 240
 241   for (j = 0;  j < b;  ++j) {
 242     if (case_flags) case_flags[out] = flagged(input[j]);
 243     if (!basic(input[j])) return punycode_bad_input;
 244     output[out++] = input[j];
 245   }
 246
 247   /* Main decoding loop:  Start just after the last delimiter if any  */
 248   /* basic code points were copied; start at the beginning otherwise. */
 249
 250   for (in = b > 0 ? b + 1 : 0;  in < input_length;  ++out) {
 251
 252     /* in is the index of the next character to be consumed, and */
 253     /* out is the number of code points in the output array.     */
 254
 255     /* Decode a generalized variable-length integer into delta,  */
 256     /* which gets added to i.  The overflow checking is easier   */
 257     /* if we increase i as we go, then subtract off its starting */
 258     /* value at the end to obtain delta.                         */
 259
 260     for (oldi = i, w = 1, k = base;  ;  k += base) {
 261       if (in >= input_length) return punycode_bad_input;
 262       digit = decode_digit(input[in++]);
 263       if (digit >= base) return punycode_bad_input;
 264       if (digit > (maxint - i) / w) return punycode_overflow;
 265       i += digit * w;
 266       t = k <= bias /* + tmin */ ? tmin :     /* +tmin not needed */
 267           k >= bias + tmax ? tmax : k - bias;
 268       if (digit < t) break;
 269       if (w > maxint / (base - t)) return punycode_overflow;
 270       w *= (base - t);
 271     }
 272
 273     bias = adapt(i - oldi, out + 1, oldi == 0);
 274
 275     /* i was supposed to wrap around from out+1 to 0,   */
 276     /* incrementing n each time, so we'll fix that now: */
 277
 278     if (i / (out + 1) > maxint - n) return punycode_overflow;
 279     n += i / (out + 1);
 280     i %= (out + 1);
 281
 282     /* Insert n at position i of the output: */
 283
 284     /* not needed for Punycode: */
 285     /* if (decode_digit(n) <= base) return punycode_invalid_input; */
 286     if (out >= max_out) return punycode_big_output;
 287
 288     if (case_flags) {
 289       memmove(case_flags + i + 1, case_flags + i, out - i);
 290       /* Case of last character determines uppercase flag: */
 291       case_flags[i] = flagged(input[in - 1]);
 292     }
 293
 294     memmove(output + i + 1, output + i, (out - i) * sizeof *output);
 295     output[i++] = n;
 296   }
 297
 298   *output_length = out;
 299   return punycode_success;
 300 }
 301
 302 U_CDECL_END
 303
 304 #endif /* #if !UCONFIG_NO_IDNA */