michael@0: /******* BEGIN LICENSE BLOCK *******
michael@0:  * Version: MPL 1.1/GPL 2.0/LGPL 2.1
michael@0:  * 
michael@0:  * The contents of this file are subject to the Mozilla Public License Version
michael@0:  * 1.1 (the "License"); you may not use this file except in compliance with
michael@0:  * the License. You may obtain a copy of the License at
michael@0:  * http://www.mozilla.org/MPL/
michael@0:  * 
michael@0:  * Software distributed under the License is distributed on an "AS IS" basis,
michael@0:  * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the License
michael@0:  * for the specific language governing rights and limitations under the
michael@0:  * License.
michael@0:  * 
michael@0:  * The Initial Developers of the Original Code are Kevin Hendricks (MySpell)
michael@0:  * and László Németh (Hunspell). Portions created by the Initial Developers
michael@0:  * are Copyright (C) 2002-2005 the Initial Developers. All Rights Reserved.
michael@0:  * 
michael@0:  * Contributor(s): László Németh (nemethl@gyorsposta.hu)
michael@0:  *                 Caolan McNamara (caolanm@redhat.com)
michael@0:  * 
michael@0:  * Alternatively, the contents of this file may be used under the terms of
michael@0:  * either the GNU General Public License Version 2 or later (the "GPL"), or
michael@0:  * the GNU Lesser General Public License Version 2.1 or later (the "LGPL"),
michael@0:  * in which case the provisions of the GPL or the LGPL are applicable instead
michael@0:  * of those above. If you wish to allow use of your version of this file only
michael@0:  * under the terms of either the GPL or the LGPL, and not to allow others to
michael@0:  * use your version of this file under the terms of the MPL, indicate your
michael@0:  * decision by deleting the provisions above and replace them with the notice
michael@0:  * and other provisions required by the GPL or the LGPL. If you do not delete
michael@0:  * the provisions above, a recipient may use your version of this file under
michael@0:  * the terms of any one of the MPL, the GPL or the LGPL.
michael@0:  *
michael@0:  ******* END LICENSE BLOCK *******/
michael@0: 
michael@0: #include <stdlib.h> 
michael@0: #include <string.h>
michael@0: #include <stdio.h> 
michael@0: 
michael@0: #include "hunzip.hxx"
michael@0: 
michael@0: #define CODELEN  65536
michael@0: #define BASEBITREC 5000
michael@0: 
michael@0: #define UNCOMPRESSED '\002'
michael@0: #define MAGIC "hz0"
michael@0: #define MAGIC_ENCRYPT "hz1"
michael@0: #define MAGICLEN (sizeof(MAGIC) - 1)
michael@0: 
michael@0: int Hunzip::fail(const char * err, const char * par) {
michael@0:     fprintf(stderr, err, par);
michael@0:     return -1;
michael@0: }
michael@0: 
michael@0: Hunzip::Hunzip(const char * file, const char * key) {
michael@0:     bufsiz = 0;
michael@0:     lastbit = 0;
michael@0:     inc = 0;
michael@0:     outc = 0;
michael@0:     dec = NULL;
michael@0:     fin = NULL;
michael@0:     filename = (char *) malloc(strlen(file) + 1);
michael@0:     if (filename) strcpy(filename, file);
michael@0:     if (getcode(key) == -1) bufsiz = -1;
michael@0:     else bufsiz = getbuf();
michael@0: }
michael@0: 
michael@0: int Hunzip::getcode(const char * key) {
michael@0:     unsigned char c[2];
michael@0:     int i, j, n, p;
michael@0:     int allocatedbit = BASEBITREC;
michael@0:     const char * enc = key;
michael@0: 
michael@0:     if (!filename) return -1;
michael@0: 
michael@0:     fin = fopen(filename, "rb");
michael@0:     if (!fin) return -1;
michael@0: 
michael@0:     // read magic number
michael@0:     if ((fread(in, 1, 3, fin) < MAGICLEN)
michael@0:         || !(strncmp(MAGIC, in, MAGICLEN) == 0 ||
michael@0:                 strncmp(MAGIC_ENCRYPT, in, MAGICLEN) == 0)) {
michael@0:             return fail(MSG_FORMAT, filename);
michael@0:     }
michael@0: 
michael@0:     // check encryption
michael@0:     if (strncmp(MAGIC_ENCRYPT, in, MAGICLEN) == 0) {
michael@0:         unsigned char cs;
michael@0:         if (!key) return fail(MSG_KEY, filename);
michael@0:         if (fread(&c, 1, 1, fin) < 1) return fail(MSG_FORMAT, filename);
michael@0:         for (cs = 0; *enc; enc++) cs ^= *enc;
michael@0:         if (cs != c[0]) return fail(MSG_KEY, filename);
michael@0:         enc = key;
michael@0:     } else key = NULL;
michael@0: 
michael@0:     // read record count
michael@0:     if (fread(&c, 1, 2, fin) < 2) return fail(MSG_FORMAT, filename);
michael@0: 
michael@0:     if (key) {
michael@0:         c[0] ^= *enc;
michael@0:         if (*(++enc) == '\0') enc = key;
michael@0:         c[1] ^= *enc;
michael@0:     }        
michael@0:     
michael@0:     n = ((int) c[0] << 8) + c[1];
michael@0:     dec = (struct bit *) malloc(BASEBITREC * sizeof(struct bit));
michael@0:     if (!dec) return fail(MSG_MEMORY, filename);
michael@0:     dec[0].v[0] = 0;
michael@0:     dec[0].v[1] = 0;
michael@0: 
michael@0:     // read codes
michael@0:     for (i = 0; i < n; i++) {
michael@0:         unsigned char l;
michael@0:         if (fread(c, 1, 2, fin) < 2) return fail(MSG_FORMAT, filename);
michael@0:         if (key) {
michael@0:             if (*(++enc) == '\0') enc = key;
michael@0:             c[0] ^= *enc;
michael@0:             if (*(++enc) == '\0') enc = key;            
michael@0:             c[1] ^= *enc;
michael@0:         }        
michael@0:         if (fread(&l, 1, 1, fin) < 1) return fail(MSG_FORMAT, filename);
michael@0:         if (key) {
michael@0:             if (*(++enc) == '\0') enc = key;
michael@0:             l ^= *enc;
michael@0:         }
michael@0:         if (fread(in, 1, l/8+1, fin) < (size_t) l/8+1) return fail(MSG_FORMAT, filename);
michael@0:         if (key) for (j = 0; j <= l/8; j++) {
michael@0:             if (*(++enc) == '\0') enc = key;
michael@0:             in[j] ^= *enc;
michael@0:         }
michael@0:         p = 0;
michael@0:         for (j = 0; j < l; j++) {
michael@0:             int b = (in[j/8] & (1 << (7 - (j % 8)))) ? 1 : 0;
michael@0:             int oldp = p;
michael@0:             p = dec[p].v[b];
michael@0:             if (p == 0) {
michael@0:                 lastbit++;
michael@0:                 if (lastbit == allocatedbit) {
michael@0:                     allocatedbit += BASEBITREC;
michael@0:                     dec = (struct bit *) realloc(dec, allocatedbit * sizeof(struct bit));
michael@0:                 }
michael@0:                 dec[lastbit].v[0] = 0;
michael@0:                 dec[lastbit].v[1] = 0;
michael@0:                 dec[oldp].v[b] = lastbit;
michael@0:                 p = lastbit;
michael@0:             }
michael@0:         }
michael@0:         dec[p].c[0] = c[0];
michael@0:         dec[p].c[1] = c[1];
michael@0:     }
michael@0:     return 0;
michael@0: }
michael@0: 
michael@0: Hunzip::~Hunzip()
michael@0: {
michael@0:     if (dec) free(dec);
michael@0:     if (fin) fclose(fin);
michael@0:     if (filename) free(filename);
michael@0: }
michael@0: 
michael@0: int Hunzip::getbuf() {
michael@0:     int p = 0;
michael@0:     int o = 0;
michael@0:     do {
michael@0:         if (inc == 0) inbits = fread(in, 1, BUFSIZE, fin) * 8;
michael@0:         for (; inc < inbits; inc++) {
michael@0:             int b = (in[inc / 8] & (1 << (7 - (inc % 8)))) ? 1 : 0;
michael@0:             int oldp = p;
michael@0:             p = dec[p].v[b];
michael@0:             if (p == 0) {
michael@0:                 if (oldp == lastbit) {
michael@0:                     fclose(fin);
michael@0:                     fin = NULL;
michael@0:                     // add last odd byte
michael@0:                     if (dec[lastbit].c[0]) out[o++]  = dec[lastbit].c[1];
michael@0:                     return o;
michael@0:                 }
michael@0:                 out[o++] = dec[oldp].c[0];
michael@0:                 out[o++] = dec[oldp].c[1];
michael@0:                 if (o == BUFSIZE) return o;
michael@0:                 p = dec[p].v[b];
michael@0:             }
michael@0:         }
michael@0:         inc = 0;
michael@0:     } while (inbits == BUFSIZE * 8);
michael@0:     return fail(MSG_FORMAT, filename);
michael@0: }
michael@0: 
michael@0: const char * Hunzip::getline() {
michael@0:     char linebuf[BUFSIZE];
michael@0:     int l = 0, eol = 0, left = 0, right = 0;
michael@0:     if (bufsiz == -1) return NULL;
michael@0:     while (l < bufsiz && !eol) {
michael@0:         linebuf[l++] = out[outc];
michael@0:         switch (out[outc]) {
michael@0:             case '\t': break;
michael@0:             case 31: { // escape
michael@0:                 if (++outc == bufsiz) {
michael@0:                     bufsiz = getbuf();
michael@0:                     outc = 0;
michael@0:                 }
michael@0:                 linebuf[l - 1] = out[outc];
michael@0:                 break;
michael@0:             }
michael@0:             case ' ': break;
michael@0:             default: if (((unsigned char) out[outc]) < 47) {
michael@0:                 if (out[outc] > 32) {
michael@0:                     right = out[outc] - 31;
michael@0:                     if (++outc == bufsiz) {
michael@0:                         bufsiz = getbuf();
michael@0:                         outc = 0;
michael@0:                     }
michael@0:                 }
michael@0:                 if (out[outc] == 30) left = 9; else left = out[outc];
michael@0:                 linebuf[l-1] = '\n';
michael@0:                 eol = 1;
michael@0:             }
michael@0:         }
michael@0:         if (++outc == bufsiz) {
michael@0:             outc = 0;
michael@0:             bufsiz = fin ? getbuf(): -1;
michael@0:         }
michael@0:     }
michael@0:     if (right) strcpy(linebuf + l - 1, line + strlen(line) - right - 1);
michael@0:     else linebuf[l] = '\0';
michael@0:     strcpy(line + left, linebuf);
michael@0:     return line;
michael@0: }