[855] | 1 | /**************************************************************************
|
---|
| 2 | *
|
---|
| 3 | * UCArray.cpp -- vector based string class
|
---|
| 4 | * Copyright (C) 1999 Rodger McNab
|
---|
| 5 | *
|
---|
| 6 | * This program is free software; you can redistribute it and/or modify
|
---|
| 7 | * it under the terms of the GNU General Public License as published by
|
---|
| 8 | * the Free Software Foundation; either version 2 of the License, or
|
---|
| 9 | * (at your option) any later version.
|
---|
| 10 | *
|
---|
| 11 | * This program is distributed in the hope that it will be useful,
|
---|
| 12 | * but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
| 13 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
| 14 | * GNU General Public License for more details.
|
---|
| 15 | *
|
---|
| 16 | * You should have received a copy of the GNU General Public License
|
---|
| 17 | * along with this program; if not, write to the Free Software
|
---|
| 18 | * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
|
---|
| 19 | *
|
---|
| 20 | * $Id: UCArray.cpp 1279 2000-07-12 22:21:53Z sjboddie $
|
---|
| 21 | *
|
---|
| 22 | **************************************************************************/
|
---|
| 23 |
|
---|
| 24 | #include "UCArray.h"
|
---|
| 25 | #include "netorder.h" /* [RPAP - Jan 97: Endian Ordering] */
|
---|
| 26 |
|
---|
| 27 |
|
---|
| 28 | void SetCStr (UCArray &text, const char *cStr) {
|
---|
| 29 | text.erase(text.begin(), text.end());
|
---|
| 30 |
|
---|
| 31 | while (*cStr != '\0') {
|
---|
| 32 | text.push_back (*cStr);
|
---|
| 33 | cStr++;
|
---|
| 34 | }
|
---|
| 35 | }
|
---|
| 36 |
|
---|
[1279] | 37 | char * GetCStr(UCArray text) {
|
---|
[855] | 38 |
|
---|
[1279] | 39 | char *cstr = new char[text.size()+1];
|
---|
| 40 | UCArray::const_iterator here = text.begin();
|
---|
| 41 | UCArray::const_iterator end = text.end();
|
---|
| 42 |
|
---|
| 43 | int i = 0;
|
---|
| 44 | while (here != end) {
|
---|
| 45 | cstr[i] = (char)*here;
|
---|
| 46 | here++; i++;
|
---|
| 47 | }
|
---|
| 48 | cstr[i]='\0';
|
---|
| 49 | return cstr;
|
---|
| 50 | }
|
---|
| 51 |
|
---|
[855] | 52 | ostream &operator<<(ostream &s, const UCArray &a) {
|
---|
| 53 | UCArray::const_iterator here = a.begin();
|
---|
| 54 | UCArray::const_iterator end = a.end();
|
---|
| 55 | while (here != end) {
|
---|
| 56 | s << *here;
|
---|
| 57 | here++;
|
---|
| 58 | }
|
---|
| 59 |
|
---|
| 60 | return s;
|
---|
| 61 | }
|
---|
| 62 |
|
---|
| 63 |
|
---|
| 64 | bool ReadVarLenUL (FILE *f, unsigned long &n) {
|
---|
| 65 | register unsigned long temp = 0;
|
---|
| 66 | register unsigned int bitPos = 0;
|
---|
| 67 | unsigned char b = 0;
|
---|
| 68 |
|
---|
| 69 | do {
|
---|
| 70 | b = fgetc (f);
|
---|
| 71 | if (feof(f)) return false;
|
---|
| 72 | temp |= (b & 0x7f) << bitPos;
|
---|
| 73 | bitPos += 7;
|
---|
| 74 | } while (b >= 0x80 && bitPos < 32);
|
---|
| 75 |
|
---|
| 76 | n = temp;
|
---|
| 77 |
|
---|
| 78 | return true;
|
---|
| 79 | }
|
---|
| 80 |
|
---|
| 81 | bool WriteVarLenUL (FILE *f, unsigned long n) {
|
---|
| 82 | register unsigned long temp = n;
|
---|
| 83 | register unsigned char b = 0;
|
---|
| 84 | do {
|
---|
| 85 | b = static_cast<unsigned char> (temp & 0x7f);
|
---|
| 86 | if (temp >= 0x80) b |= 0x80;
|
---|
| 87 | fputc (b, f);
|
---|
| 88 | if (ferror (f) != 0) return false;
|
---|
| 89 | } while ((temp = temp >> 7) > 0);
|
---|
| 90 |
|
---|
| 91 | return true;
|
---|
| 92 | }
|
---|
| 93 |
|
---|
| 94 |
|
---|
| 95 | bool ReadUL (FILE *f, unsigned long &n) {
|
---|
| 96 | if (fread (&n, sizeof (unsigned long), 1, f) <= 0) return false;
|
---|
| 97 | NTOHUL (n);
|
---|
| 98 | return true;
|
---|
| 99 | }
|
---|
| 100 |
|
---|
| 101 |
|
---|
| 102 | bool WriteUL (FILE *f, unsigned long n) {
|
---|
| 103 | HTONUL (n);
|
---|
| 104 | return (fwrite (&n, sizeof (unsigned long), 1, f) > 0);
|
---|
| 105 | }
|
---|
| 106 |
|
---|
| 107 | bool ReadF (FILE *f, float &n) {
|
---|
| 108 | if (fread (&n, sizeof (float), 1, f) <= 0) return false;
|
---|
| 109 | NTOHF(n);
|
---|
| 110 | return true;
|
---|
| 111 | }
|
---|
| 112 |
|
---|
| 113 | bool WriteF (FILE *f, float n) {
|
---|
| 114 | HTONF(n);
|
---|
| 115 | return (fwrite (&n, sizeof (float), 1, f) > 0);
|
---|
| 116 | }
|
---|
| 117 |
|
---|
| 118 | bool ReadD (FILE *f, double &n) {
|
---|
| 119 | if (fread (&n, sizeof (double), 1, f) <= 0) return false;
|
---|
| 120 | NTOHD(n);
|
---|
| 121 | return true;
|
---|
| 122 | }
|
---|
| 123 |
|
---|
| 124 | bool WriteD (FILE *f, double n) {
|
---|
| 125 | HTOND(n);
|
---|
| 126 | return (fwrite (&n, sizeof (double), 1, f) > 0);
|
---|
| 127 | }
|
---|
| 128 |
|
---|
| 129 | bool ReadUCArray (FILE *f, UCArray &a) {
|
---|
| 130 | // clear the array in preparation
|
---|
| 131 | a.erase (a.begin(), a.end());
|
---|
| 132 |
|
---|
| 133 | // read in the array size
|
---|
| 134 | unsigned long arraySize = 0;
|
---|
| 135 | if (!ReadVarLenUL (f, arraySize)) return false;
|
---|
| 136 |
|
---|
| 137 | // read in the array
|
---|
| 138 | unsigned char b = 0;
|
---|
| 139 | while (arraySize > 0) {
|
---|
| 140 | b = fgetc (f);
|
---|
| 141 | if (feof(f)) return false;
|
---|
| 142 | a.push_back (b);
|
---|
| 143 |
|
---|
| 144 | arraySize--;
|
---|
| 145 | }
|
---|
| 146 |
|
---|
| 147 | return true;
|
---|
| 148 | }
|
---|
| 149 |
|
---|
| 150 | bool WriteUCArray (FILE *f, const UCArray &a) {
|
---|
| 151 | // write out the array size
|
---|
| 152 | if (!WriteVarLenUL (f, a.size())) return false;
|
---|
| 153 |
|
---|
| 154 | UCArray::const_iterator here = a.begin();
|
---|
| 155 | UCArray::const_iterator end = a.end();
|
---|
| 156 | while (here != end) {
|
---|
| 157 | fputc (*here, f);
|
---|
| 158 | if (ferror (f) != 0) return false;
|
---|
| 159 |
|
---|
| 160 | here++;
|
---|
| 161 | }
|
---|
| 162 |
|
---|
| 163 | return true;
|
---|
| 164 | }
|
---|
| 165 |
|
---|
| 166 | /*
|
---|
| 167 | * This array is designed for mapping upper and lower case letter
|
---|
| 168 | * together for a case independent comparison. The mappings are
|
---|
| 169 | * based upon ascii character sequences.
|
---|
| 170 | */
|
---|
| 171 | static unsigned char casecharmap[] = {
|
---|
| 172 | '\000', '\001', '\002', '\003', '\004', '\005', '\006', '\007',
|
---|
| 173 | '\010', '\011', '\012', '\013', '\014', '\015', '\016', '\017',
|
---|
| 174 | '\020', '\021', '\022', '\023', '\024', '\025', '\026', '\027',
|
---|
| 175 | '\030', '\031', '\032', '\033', '\034', '\035', '\036', '\037',
|
---|
| 176 | '\040', '\041', '\042', '\043', '\044', '\045', '\046', '\047',
|
---|
| 177 | '\050', '\051', '\052', '\053', '\054', '\055', '\056', '\057',
|
---|
| 178 | '\060', '\061', '\062', '\063', '\064', '\065', '\066', '\067',
|
---|
| 179 | '\070', '\071', '\072', '\073', '\074', '\075', '\076', '\077',
|
---|
| 180 | '\100', '\141', '\142', '\143', '\144', '\145', '\146', '\147',
|
---|
| 181 | '\150', '\151', '\152', '\153', '\154', '\155', '\156', '\157',
|
---|
| 182 | '\160', '\161', '\162', '\163', '\164', '\165', '\166', '\167',
|
---|
| 183 | '\170', '\171', '\172', '\133', '\134', '\135', '\136', '\137',
|
---|
| 184 | '\140', '\141', '\142', '\143', '\144', '\145', '\146', '\147',
|
---|
| 185 | '\150', '\151', '\152', '\153', '\154', '\155', '\156', '\157',
|
---|
| 186 | '\160', '\161', '\162', '\163', '\164', '\165', '\166', '\167',
|
---|
| 187 | '\170', '\171', '\172', '\173', '\174', '\175', '\176', '\177',
|
---|
| 188 | '\200', '\201', '\202', '\203', '\204', '\205', '\206', '\207',
|
---|
| 189 | '\210', '\211', '\212', '\213', '\214', '\215', '\216', '\217',
|
---|
| 190 | '\220', '\221', '\222', '\223', '\224', '\225', '\226', '\227',
|
---|
| 191 | '\230', '\231', '\232', '\233', '\234', '\235', '\236', '\237',
|
---|
| 192 | '\240', '\241', '\242', '\243', '\244', '\245', '\246', '\247',
|
---|
| 193 | '\250', '\251', '\252', '\253', '\254', '\255', '\256', '\257',
|
---|
| 194 | '\260', '\261', '\262', '\263', '\264', '\265', '\266', '\267',
|
---|
| 195 | '\270', '\271', '\272', '\273', '\274', '\275', '\276', '\277',
|
---|
| 196 | '\300', '\341', '\342', '\343', '\344', '\345', '\346', '\347',
|
---|
| 197 | '\350', '\351', '\352', '\353', '\354', '\355', '\356', '\357',
|
---|
| 198 | '\360', '\361', '\362', '\363', '\364', '\365', '\366', '\367',
|
---|
| 199 | '\370', '\371', '\372', '\333', '\334', '\335', '\336', '\337',
|
---|
| 200 | '\340', '\341', '\342', '\343', '\344', '\345', '\346', '\347',
|
---|
| 201 | '\350', '\351', '\352', '\353', '\354', '\355', '\356', '\357',
|
---|
| 202 | '\360', '\361', '\362', '\363', '\364', '\365', '\366', '\367',
|
---|
| 203 | '\370', '\371', '\372', '\373', '\374', '\375', '\376', '\377',
|
---|
| 204 | };
|
---|
| 205 |
|
---|
| 206 |
|
---|
| 207 | int DictCompare (const UCArray &a1, const UCArray &a2) {
|
---|
| 208 | unsigned int l1 = a1.size();
|
---|
| 209 | unsigned int l2 = a2.size();
|
---|
| 210 | unsigned int l = (l1 < l2) ? l1 : l2;
|
---|
| 211 | int pos = 0;
|
---|
| 212 | register int diff = 0;
|
---|
| 213 |
|
---|
| 214 | UCArray::const_iterator a1Here = a1.begin();
|
---|
| 215 | UCArray::const_iterator a2Here = a2.begin();
|
---|
| 216 |
|
---|
| 217 | while (l--) {
|
---|
| 218 | if ((diff = casecharmap[*a1Here] - casecharmap[*a2Here]) != 0)
|
---|
| 219 | return diff;
|
---|
| 220 | if (pos == 0 && (diff = *a1Here - *a2Here) != 0)
|
---|
| 221 | pos = diff;
|
---|
| 222 |
|
---|
| 223 | a1Here++;
|
---|
| 224 | a2Here++;
|
---|
| 225 | }
|
---|
| 226 |
|
---|
| 227 | return ((l1 - l2) ? (l1 - l2) : (pos));
|
---|
| 228 | }
|
---|
| 229 |
|
---|
[1279] | 230 | /* comparison for browse index - items match if the smaller word
|
---|
| 231 | is a prefix of the larger word, case independent
|
---|
| 232 | */
|
---|
| 233 | int BrowseCompare (const UCArray &a1, const UCArray &a2) {
|
---|
| 234 | unsigned int l1 = a1.size();
|
---|
| 235 | unsigned int l2 = a2.size();
|
---|
| 236 | unsigned int l = (l1 < l2) ? l1 : l2; // l is the shorter of the two
|
---|
| 237 | int diff = 0;
|
---|
[855] | 238 |
|
---|
[1279] | 239 | UCArray::const_iterator a1Here = a1.begin();
|
---|
| 240 | UCArray::const_iterator a2Here = a2.begin();
|
---|
| 241 |
|
---|
| 242 | while(l--) {
|
---|
| 243 | if ((diff = casecharmap[*a1Here] - casecharmap[*a2Here]) !=0)
|
---|
| 244 | return diff;
|
---|
| 245 | a1Here++;
|
---|
| 246 | a2Here++;
|
---|
| 247 | }
|
---|
| 248 | return 0;
|
---|
| 249 |
|
---|
| 250 | }
|
---|
| 251 |
|
---|
[855] | 252 | unsigned long PrefixLen (const UCArray &a1, const UCArray &a2) {
|
---|
| 253 | unsigned long l = (a1.size() < a2.size()) ? a1.size() : a2.size();
|
---|
| 254 | unsigned long i = 0;
|
---|
| 255 |
|
---|
| 256 | UCArray::const_iterator a1Here = a1.begin();
|
---|
| 257 | UCArray::const_iterator a2Here = a2.begin();
|
---|
| 258 |
|
---|
| 259 | while (i < l && *a1Here == *a2Here) {
|
---|
| 260 | i++; a1Here++; a2Here++;
|
---|
| 261 | }
|
---|
| 262 |
|
---|
| 263 | return i;
|
---|
| 264 | }
|
---|
| 265 |
|
---|
| 266 | bool WritePreSufStr (FILE *f, const UCArray *prev, const UCArray &a) {
|
---|
| 267 | unsigned char preLen;
|
---|
| 268 | unsigned char sufLen;
|
---|
| 269 |
|
---|
| 270 | if (prev != NULL) preLen = PrefixLen (*prev, a);
|
---|
| 271 | else preLen = 0;
|
---|
| 272 | sufLen = a.size() - preLen;
|
---|
| 273 |
|
---|
| 274 | // output the prefix length, suffix length, and the suffix
|
---|
| 275 | fputc (preLen, f);
|
---|
| 276 | if (ferror(f) != 0) return false;
|
---|
| 277 | fputc (sufLen, f);
|
---|
| 278 | if (ferror(f) != 0) return false;
|
---|
| 279 | return (fwrite ((char *)a.begin()+preLen, sizeof (char), sufLen, f) == sufLen);
|
---|
| 280 | }
|
---|
| 281 |
|
---|
| 282 | // a also used for prev
|
---|
| 283 | bool ReadPreSufStr (FILE *f, UCArray &a) {
|
---|
| 284 | unsigned char preLen = 0;
|
---|
| 285 | unsigned char sufLen = 0;
|
---|
| 286 |
|
---|
| 287 | preLen = fgetc(f);
|
---|
| 288 | sufLen = fgetc(f);
|
---|
| 289 |
|
---|
| 290 | if (a.size () > preLen) a.erase (a.begin()+preLen, a.end());
|
---|
| 291 | while (sufLen > 0) {
|
---|
| 292 | unsigned char c = fgetc (f);
|
---|
| 293 | a.push_back (c);
|
---|
| 294 | sufLen--;
|
---|
| 295 | }
|
---|
| 296 |
|
---|
| 297 | return true;
|
---|
| 298 | }
|
---|
| 299 |
|
---|