1 /*
2 punycode-sample.c 2.0.0 (2004-Mar-21-Sun)
3 http://www.nicemice.net/idn/
4 Adam M. Costello
5 http://www.nicemice.net/amc/
6
7 This is ANSI C code (C89) implementing Punycode 1.0.x.
8 */
9
10 /**********************************************************/
11 /* Implementation (would normally go in its own .c file): */
12
13 #include <string.h>
14
15 #include "punycode.h"
16
17 /*** Bootstring parameters for Punycode ***/
18
19 enum { base = 36, tmin = 1, tmax = 26, skew = 38, damp = 700,
20 initial_bias = 72, initial_n = 0x80, delimiter = 0x2D };
21
22 /* basic(cp) tests whether cp is a basic code point: */
23 #define basic(cp) ((punycode_uint)(cp) < 0x80)
24
25 /* delim(cp) tests whether cp is a delimiter: */
26 #define delim(cp) ((cp) == delimiter)
27
28 /* decode_digit(cp) returns the numeric value of a basic code */
29 /* point (for use in representing integers) in the range 0 to */
30 /* base-1, or base if cp does not represent a value. */
31
decode_digit(punycode_uint cp)32 static punycode_uint decode_digit(punycode_uint cp)
33 {
34 return cp - 48 < 10 ? cp - 22 : cp - 65 < 26 ? cp - 65 :
35 cp - 97 < 26 ? cp - 97 : base;
36 }
37
38 /* encode_digit(d,flag) returns the basic code point whose value */
39 /* (when used for representing integers) is d, which needs to be in */
40 /* the range 0 to base-1. The lowercase form is used unless flag is */
41 /* nonzero, in which case the uppercase form is used. The behavior */
42 /* is undefined if flag is nonzero and digit d has no uppercase form. */
43
encode_digit(punycode_uint d,int flag)44 static char encode_digit(punycode_uint d, int flag)
45 {
46 return d + 22 + 75 * (d < 26) - ((flag != 0) << 5);
47 /* 0..25 map to ASCII a..z or A..Z */
48 /* 26..35 map to ASCII 0..9 */
49 }
50
51 /* flagged(bcp) tests whether a basic code point is flagged */
52 /* (uppercase). The behavior is undefined if bcp is not a */
53 /* basic code point. */
54
55 #define flagged(bcp) ((punycode_uint)(bcp) - 65 < 26)
56
57 /* encode_basic(bcp,flag) forces a basic code point to lowercase */
58 /* if flag is zero, uppercase if flag is nonzero, and returns */
59 /* the resulting code point. The code point is unchanged if it */
60 /* is caseless. The behavior is undefined if bcp is not a basic */
61 /* code point. */
62
encode_basic(punycode_uint bcp,int flag)63 static char encode_basic(punycode_uint bcp, int flag)
64 {
65 bcp -= (bcp - 97 < 26) << 5;
66 return bcp + ((!flag && (bcp - 65 < 26)) << 5);
67 }
68
69 /*** Platform-specific constants ***/
70
71 /* maxint is the maximum value of a punycode_uint variable: */
72 static const punycode_uint maxint = -1;
73 /* Because maxint is unsigned, -1 becomes the maximum value. */
74
75 /*** Bias adaptation function ***/
76
adapt(punycode_uint delta,punycode_uint numpoints,int firsttime)77 static punycode_uint adapt(
78 punycode_uint delta, punycode_uint numpoints, int firsttime )
79 {
80 punycode_uint k;
81
82 delta = firsttime ? delta / damp : delta >> 1;
83 /* delta >> 1 is a faster way of doing delta / 2 */
84 delta += delta / numpoints;
85
86 for (k = 0; delta > ((base - tmin) * tmax) / 2; k += base) {
87 delta /= base - tmin;
88 }
89
90 return k + (base - tmin + 1) * delta / (delta + skew);
91 }
92
93 /*** Main encode function ***/
94
punycode_encode(size_t input_length_orig,const punycode_uint input[],const unsigned char case_flags[],size_t * output_length,char output[])95 enum punycode_status punycode_encode(
96 size_t input_length_orig,
97 const punycode_uint input[],
98 const unsigned char case_flags[],
99 size_t *output_length,
100 char output[] )
101 {
102 punycode_uint input_length, n, delta, h, b, bias, j, m, q, k, t;
103 size_t out, max_out;
104
105 /* The Punycode spec assumes that the input length is the same type */
106 /* of integer as a code point, so we need to convert the size_t to */
107 /* a punycode_uint, which could overflow. */
108
109 if (input_length_orig > maxint) return punycode_overflow;
110 input_length = (punycode_uint) input_length_orig;
111
112 /* Initialize the state: */
113
114 n = initial_n;
115 delta = 0;
116 out = 0;
117 max_out = *output_length;
118 bias = initial_bias;
119
120 /* Handle the basic code points: */
121
122 for (j = 0; j < input_length; ++j) {
123 if (basic(input[j])) {
124 if (max_out - out < 2) return punycode_big_output;
125 output[out++] = case_flags ?
126 encode_basic(input[j], case_flags[j]) : (char) input[j];
127 }
128 /* else if (input[j] < n) return punycode_bad_input; */
129 /* (not needed for Punycode with unsigned code points) */
130 }
131
132 h = b = (punycode_uint) out;
133 /* cannot overflow because out <= input_length <= maxint */
134
135 /* h is the number of code points that have been handled, b is the */
136 /* number of basic code points, and out is the number of ASCII code */
137 /* points that have been output. */
138
139 if (b > 0) output[out++] = delimiter;
140
141 /* Main encoding loop: */
142
143 while (h < input_length) {
144 /* All non-basic code points < n have been */
145 /* handled already. Find the next larger one: */
146
147 for (m = maxint, j = 0; j < input_length; ++j) {
148 /* if (basic(input[j])) continue; */
149 /* (not needed for Punycode) */
150 if (input[j] >= n && input[j] < m) m = input[j];
151 }
152
153 /* Increase delta enough to advance the decoder's */
154 /* <n,i> state to <m,0>, but guard against overflow: */
155
156 if (m - n > (maxint - delta) / (h + 1)) return punycode_overflow;
157 delta += (m - n) * (h + 1);
158 n = m;
159
160 for (j = 0; j < input_length; ++j) {
161 /* Punycode does not need to check whether input[j] is basic: */
162 if (input[j] < n /* || basic(input[j]) */ ) {
163 if (++delta == 0) return punycode_overflow;
164 }
165
166 if (input[j] == n) {
167 /* Represent delta as a generalized variable-length integer: */
168
169 for (q = delta, k = base; ; k += base) {
170 if (out >= max_out) return punycode_big_output;
171 t = k <= bias /* + tmin */ ? tmin : /* +tmin not needed */
172 k >= bias + tmax ? tmax : k - bias;
173 if (q < t) break;
174 output[out++] = encode_digit(t + (q - t) % (base - t), 0);
175 q = (q - t) / (base - t);
176 }
177
178 output[out++] = encode_digit(q, case_flags && case_flags[j]);
179 bias = adapt(delta, h + 1, h == b);
180 delta = 0;
181 ++h;
182 }
183 }
184
185 ++delta, ++n;
186 }
187
188 *output_length = out;
189 return punycode_success;
190 }
191
192 /*** Main decode function ***/
193
punycode_decode(size_t input_length,const char input[],size_t * output_length,punycode_uint output[],unsigned char case_flags[])194 enum punycode_status punycode_decode(
195 size_t input_length,
196 const char input[],
197 size_t *output_length,
198 punycode_uint output[],
199 unsigned char case_flags[] )
200 {
201 punycode_uint n, out, i, max_out, bias, oldi, w, k, digit, t;
202 size_t b, j, in;
203
204 /* Initialize the state: */
205
206 n = initial_n;
207 out = i = 0;
208 max_out = *output_length > maxint ? maxint
209 : (punycode_uint) *output_length;
210 bias = initial_bias;
211
212 /* Handle the basic code points: Let b be the number of input code */
213 /* points before the last delimiter, or 0 if there is none, then */
214 /* copy the first b code points to the output. */
215
216 for (b = j = 0; j < input_length; ++j) if (delim(input[j])) b = j;
217 if (b > max_out) return punycode_big_output;
218
219 for (j = 0; j < b; ++j) {
220 if (case_flags) case_flags[out] = flagged(input[j]);
221 if (!basic(input[j])) return punycode_bad_input;
222 output[out++] = input[j];
223 }
224
225 /* Main decoding loop: Start just after the last delimiter if any */
226 /* basic code points were copied; start at the beginning otherwise. */
227
228 for (in = b > 0 ? b + 1 : 0; in < input_length; ++out) {
229
230 /* in is the index of the next ASCII code point to be consumed, */
231 /* and out is the number of code points in the output array. */
232
233 /* Decode a generalized variable-length integer into delta, */
234 /* which gets added to i. The overflow checking is easier */
235 /* if we increase i as we go, then subtract off its starting */
236 /* value at the end to obtain delta. */
237
238 for (oldi = i, w = 1, k = base; ; k += base) {
239 if (in >= input_length) return punycode_bad_input;
240 digit = decode_digit(input[in++]);
241 if (digit >= base) return punycode_bad_input;
242 if (digit > (maxint - i) / w) return punycode_overflow;
243 i += digit * w;
244 t = k <= bias /* + tmin */ ? tmin : /* +tmin not needed */
245 k >= bias + tmax ? tmax : k - bias;
246 if (digit < t) break;
247 if (w > maxint / (base - t)) return punycode_overflow;
248 w *= (base - t);
249 }
250
251 bias = adapt(i - oldi, out + 1, oldi == 0);
252
253 /* i was supposed to wrap around from out+1 to 0, */
254 /* incrementing n each time, so we'll fix that now: */
255
256 if (i / (out + 1) > maxint - n) return punycode_overflow;
257 n += i / (out + 1);
258 i %= (out + 1);
259
260 /* Insert n at position i of the output: */
261
262 /* not needed for Punycode: */
263 /* if (basic(n)) return punycode_bad_input; */
264 if (out >= max_out) return punycode_big_output;
265
266 if (case_flags) {
267 memmove(case_flags + i + 1, case_flags + i, out - i);
268 /* Case of last ASCII code point determines case flag: */
269 case_flags[i] = flagged(input[in - 1]);
270 }
271
272 memmove(output + i + 1, output + i, (out - i) * sizeof *output);
273 output[i++] = n;
274 }
275
276 *output_length = (size_t) out;
277 /* cannot overflow because out <= old value of *output_length */
278 return punycode_success;
279 }
280
281