The Tor Browser / file comparison
comparison: extensions/spellcheck/hunspell/src/hunzip.cpp
extensions/spellcheck/hunspell/src/hunzip.cpp
- branch
- TOR_BUG_3246
- changeset 6
- 8bccb770b82d
equal
deleted
inserted
replaced
| -1:000000000000 | 0:9e9be5c2d9f7 |
|---|---|
| 1 /******* BEGIN LICENSE BLOCK ******* | |
| 2 * Version: MPL 1.1/GPL 2.0/LGPL 2.1 | |
| 3 * | |
| 4 * The contents of this file are subject to the Mozilla Public License Version | |
| 5 * 1.1 (the "License"); you may not use this file except in compliance with | |
| 6 * the License. You may obtain a copy of the License at | |
| 7 * http://www.mozilla.org/MPL/ | |
| 8 * | |
| 9 * Software distributed under the License is distributed on an "AS IS" basis, | |
| 10 * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the License | |
| 11 * for the specific language governing rights and limitations under the | |
| 12 * License. | |
| 13 * | |
| 14 * The Initial Developers of the Original Code are Kevin Hendricks (MySpell) | |
| 15 * and László Németh (Hunspell). Portions created by the Initial Developers | |
| 16 * are Copyright (C) 2002-2005 the Initial Developers. All Rights Reserved. | |
| 17 * | |
| 18 * Contributor(s): László Németh (nemethl@gyorsposta.hu) | |
| 19 * Caolan McNamara (caolanm@redhat.com) | |
| 20 * | |
| 21 * Alternatively, the contents of this file may be used under the terms of | |
| 22 * either the GNU General Public License Version 2 or later (the "GPL"), or | |
| 23 * the GNU Lesser General Public License Version 2.1 or later (the "LGPL"), | |
| 24 * in which case the provisions of the GPL or the LGPL are applicable instead | |
| 25 * of those above. If you wish to allow use of your version of this file only | |
| 26 * under the terms of either the GPL or the LGPL, and not to allow others to | |
| 27 * use your version of this file under the terms of the MPL, indicate your | |
| 28 * decision by deleting the provisions above and replace them with the notice | |
| 29 * and other provisions required by the GPL or the LGPL. If you do not delete | |
| 30 * the provisions above, a recipient may use your version of this file under | |
| 31 * the terms of any one of the MPL, the GPL or the LGPL. | |
| 32 * | |
| 33 ******* END LICENSE BLOCK *******/ | |
| 34 | |
| 35 #include <stdlib.h> | |
| 36 #include <string.h> | |
| 37 #include <stdio.h> | |
| 38 | |
| 39 #include "hunzip.hxx" | |
| 40 | |
| 41 #define CODELEN 65536 | |
| 42 #define BASEBITREC 5000 | |
| 43 | |
| 44 #define UNCOMPRESSED '\002' | |
| 45 #define MAGIC "hz0" | |
| 46 #define MAGIC_ENCRYPT "hz1" | |
| 47 #define MAGICLEN (sizeof(MAGIC) - 1) | |
| 48 | |
| 49 int Hunzip::fail(const char * err, const char * par) { | |
| 50 fprintf(stderr, err, par); | |
| 51 return -1; | |
| 52 } | |
| 53 | |
| 54 Hunzip::Hunzip(const char * file, const char * key) { | |
| 55 bufsiz = 0; | |
| 56 lastbit = 0; | |
| 57 inc = 0; | |
| 58 outc = 0; | |
| 59 dec = NULL; | |
| 60 fin = NULL; | |
| 61 filename = (char *) malloc(strlen(file) + 1); | |
| 62 if (filename) strcpy(filename, file); | |
| 63 if (getcode(key) == -1) bufsiz = -1; | |
| 64 else bufsiz = getbuf(); | |
| 65 } | |
| 66 | |
| 67 int Hunzip::getcode(const char * key) { | |
| 68 unsigned char c[2]; | |
| 69 int i, j, n, p; | |
| 70 int allocatedbit = BASEBITREC; | |
| 71 const char * enc = key; | |
| 72 | |
| 73 if (!filename) return -1; | |
| 74 | |
| 75 fin = fopen(filename, "rb"); | |
| 76 if (!fin) return -1; | |
| 77 | |
| 78 // read magic number | |
| 79 if ((fread(in, 1, 3, fin) < MAGICLEN) | |
| 80 || !(strncmp(MAGIC, in, MAGICLEN) == 0 || | |
| 81 strncmp(MAGIC_ENCRYPT, in, MAGICLEN) == 0)) { | |
| 82 return fail(MSG_FORMAT, filename); | |
| 83 } | |
| 84 | |
| 85 // check encryption | |
| 86 if (strncmp(MAGIC_ENCRYPT, in, MAGICLEN) == 0) { | |
| 87 unsigned char cs; | |
| 88 if (!key) return fail(MSG_KEY, filename); | |
| 89 if (fread(&c, 1, 1, fin) < 1) return fail(MSG_FORMAT, filename); | |
| 90 for (cs = 0; *enc; enc++) cs ^= *enc; | |
| 91 if (cs != c[0]) return fail(MSG_KEY, filename); | |
| 92 enc = key; | |
| 93 } else key = NULL; | |
| 94 | |
| 95 // read record count | |
| 96 if (fread(&c, 1, 2, fin) < 2) return fail(MSG_FORMAT, filename); | |
| 97 | |
| 98 if (key) { | |
| 99 c[0] ^= *enc; | |
| 100 if (*(++enc) == '\0') enc = key; | |
| 101 c[1] ^= *enc; | |
| 102 } | |
| 103 | |
| 104 n = ((int) c[0] << 8) + c[1]; | |
| 105 dec = (struct bit *) malloc(BASEBITREC * sizeof(struct bit)); | |
| 106 if (!dec) return fail(MSG_MEMORY, filename); | |
| 107 dec[0].v[0] = 0; | |
| 108 dec[0].v[1] = 0; | |
| 109 | |
| 110 // read codes | |
| 111 for (i = 0; i < n; i++) { | |
| 112 unsigned char l; | |
| 113 if (fread(c, 1, 2, fin) < 2) return fail(MSG_FORMAT, filename); | |
| 114 if (key) { | |
| 115 if (*(++enc) == '\0') enc = key; | |
| 116 c[0] ^= *enc; | |
| 117 if (*(++enc) == '\0') enc = key; | |
| 118 c[1] ^= *enc; | |
| 119 } | |
| 120 if (fread(&l, 1, 1, fin) < 1) return fail(MSG_FORMAT, filename); | |
| 121 if (key) { | |
| 122 if (*(++enc) == '\0') enc = key; | |
| 123 l ^= *enc; | |
| 124 } | |
| 125 if (fread(in, 1, l/8+1, fin) < (size_t) l/8+1) return fail(MSG_FORMAT, filename); | |
| 126 if (key) for (j = 0; j <= l/8; j++) { | |
| 127 if (*(++enc) == '\0') enc = key; | |
| 128 in[j] ^= *enc; | |
| 129 } | |
| 130 p = 0; | |
| 131 for (j = 0; j < l; j++) { | |
| 132 int b = (in[j/8] & (1 << (7 - (j % 8)))) ? 1 : 0; | |
| 133 int oldp = p; | |
| 134 p = dec[p].v[b]; | |
| 135 if (p == 0) { | |
| 136 lastbit++; | |
| 137 if (lastbit == allocatedbit) { | |
| 138 allocatedbit += BASEBITREC; | |
| 139 dec = (struct bit *) realloc(dec, allocatedbit * sizeof(struct bit)); | |
| 140 } | |
| 141 dec[lastbit].v[0] = 0; | |
| 142 dec[lastbit].v[1] = 0; | |
| 143 dec[oldp].v[b] = lastbit; | |
| 144 p = lastbit; | |
| 145 } | |
| 146 } | |
| 147 dec[p].c[0] = c[0]; | |
| 148 dec[p].c[1] = c[1]; | |
| 149 } | |
| 150 return 0; | |
| 151 } | |
| 152 | |
| 153 Hunzip::~Hunzip() | |
| 154 { | |
| 155 if (dec) free(dec); | |
| 156 if (fin) fclose(fin); | |
| 157 if (filename) free(filename); | |
| 158 } | |
| 159 | |
| 160 int Hunzip::getbuf() { | |
| 161 int p = 0; | |
| 162 int o = 0; | |
| 163 do { | |
| 164 if (inc == 0) inbits = fread(in, 1, BUFSIZE, fin) * 8; | |
| 165 for (; inc < inbits; inc++) { | |
| 166 int b = (in[inc / 8] & (1 << (7 - (inc % 8)))) ? 1 : 0; | |
| 167 int oldp = p; | |
| 168 p = dec[p].v[b]; | |
| 169 if (p == 0) { | |
| 170 if (oldp == lastbit) { | |
| 171 fclose(fin); | |
| 172 fin = NULL; | |
| 173 // add last odd byte | |
| 174 if (dec[lastbit].c[0]) out[o++] = dec[lastbit].c[1]; | |
| 175 return o; | |
| 176 } | |
| 177 out[o++] = dec[oldp].c[0]; | |
| 178 out[o++] = dec[oldp].c[1]; | |
| 179 if (o == BUFSIZE) return o; | |
| 180 p = dec[p].v[b]; | |
| 181 } | |
| 182 } | |
| 183 inc = 0; | |
| 184 } while (inbits == BUFSIZE * 8); | |
| 185 return fail(MSG_FORMAT, filename); | |
| 186 } | |
| 187 | |
| 188 const char * Hunzip::getline() { | |
| 189 char linebuf[BUFSIZE]; | |
| 190 int l = 0, eol = 0, left = 0, right = 0; | |
| 191 if (bufsiz == -1) return NULL; | |
| 192 while (l < bufsiz && !eol) { | |
| 193 linebuf[l++] = out[outc]; | |
| 194 switch (out[outc]) { | |
| 195 case '\t': break; | |
| 196 case 31: { // escape | |
| 197 if (++outc == bufsiz) { | |
| 198 bufsiz = getbuf(); | |
| 199 outc = 0; | |
| 200 } | |
| 201 linebuf[l - 1] = out[outc]; | |
| 202 break; | |
| 203 } | |
| 204 case ' ': break; | |
| 205 default: if (((unsigned char) out[outc]) < 47) { | |
| 206 if (out[outc] > 32) { | |
| 207 right = out[outc] - 31; | |
| 208 if (++outc == bufsiz) { | |
| 209 bufsiz = getbuf(); | |
| 210 outc = 0; | |
| 211 } | |
| 212 } | |
| 213 if (out[outc] == 30) left = 9; else left = out[outc]; | |
| 214 linebuf[l-1] = '\n'; | |
| 215 eol = 1; | |
| 216 } | |
| 217 } | |
| 218 if (++outc == bufsiz) { | |
| 219 outc = 0; | |
| 220 bufsiz = fin ? getbuf(): -1; | |
| 221 } | |
| 222 } | |
| 223 if (right) strcpy(linebuf + l - 1, line + strlen(line) - right - 1); | |
| 224 else linebuf[l] = '\0'; | |
| 225 strcpy(line + left, linebuf); | |
| 226 return line; | |
| 227 } |
