lib/libc/citrus/citrus_csmapper.c

   1 /* $FreeBSD: head/lib/libc/iconv/citrus_csmapper.c 263986 2014-04-01 10:36:11Z tijl $ */
   2 /*      $NetBSD: citrus_csmapper.c,v 1.11 2011/11/20 07:43:52 tnozaki Exp $     */
   3
   4 /*-
   5  * Copyright (c)2003 Citrus Project,
   6  * All rights reserved.
   7  *
   8  * Redistribution and use in source and binary forms, with or without
   9  * modification, are permitted provided that the following conditions
  10  * are met:
  11  * 1. Redistributions of source code must retain the above copyright
  12  *    notice, this list of conditions and the following disclaimer.
  13  * 2. Redistributions in binary form must reproduce the above copyright
  14  *    notice, this list of conditions and the following disclaimer in the
  15  *    documentation and/or other materials provided with the distribution.
  16  *
  17  * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
  18  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  19  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  20  * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
  21  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  22  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
  23  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  24  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  25  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  26  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  27  * SUCH DAMAGE.
  28  */
  29
  30 #include <sys/cdefs.h>
  31 #include <sys/endian.h>
  32 #include <sys/types.h>
  33 #include <sys/queue.h>
  34
  35 #include <assert.h>
  36 #include <errno.h>
  37 #include <limits.h>
  38 #include <paths.h>
  39 #include <stdio.h>
  40 #include <stdlib.h>
  41 #include <string.h>
  42
  43 #include "citrus_namespace.h"
  44 #include "citrus_types.h"
  45 #include "citrus_bcs.h"
  46 #include "citrus_region.h"
  47 #include "citrus_lock.h"
  48 #include "citrus_memstream.h"
  49 #include "citrus_mmap.h"
  50 #include "citrus_module.h"
  51 #include "citrus_hash.h"
  52 #include "citrus_mapper.h"
  53 #include "citrus_csmapper.h"
  54 #include "citrus_pivot_file.h"
  55 #include "citrus_db.h"
  56 #include "citrus_db_hash.h"
  57 #include "citrus_lookup.h"
  58
  59 static struct _citrus_mapper_area       *maparea = NULL;
  60
  61 static pthread_rwlock_t                 ma_lock = PTHREAD_RWLOCK_INITIALIZER;
  62
  63 #define CS_ALIAS        _PATH_CSMAPPER "/charset.alias"
  64 #define CS_PIVOT        _PATH_CSMAPPER "/charset.pivot"
  65
  66
  67 /* ---------------------------------------------------------------------- */
  68
  69 static int
  70 get32(struct _region *r, uint32_t *rval)
  71 {
  72
  73         if (_region_size(r) != 4)
  74                 return (EFTYPE);
  75
  76         memcpy(rval, _region_head(r), (size_t)4);
  77         *rval = be32toh(*rval);
  78
  79         return (0);
  80 }
  81
  82 static int
  83 open_subdb(struct _citrus_db **subdb, struct _citrus_db *db, const char *src)
  84 {
  85         struct _region r;
  86         int ret;
  87
  88         ret = _db_lookup_by_s(db, src, &r, NULL);
  89         if (ret)
  90                 return (ret);
  91         ret = _db_open(subdb, &r, _CITRUS_PIVOT_SUB_MAGIC, _db_hash_std, NULL);
  92         if (ret)
  93                 return (ret);
  94
  95         return (0);
  96 }
  97
  98
  99 #define NO_SUCH_FILE    EOPNOTSUPP
 100 static int
 101 find_best_pivot_pvdb(const char *src, const char *dst, char *pivot,
 102     size_t pvlen, unsigned long *rnorm)
 103 {
 104         struct _citrus_db *db1, *db2, *db3;
 105         struct _region fr, r1, r2;
 106         char buf[LINE_MAX];
 107         uint32_t val32;
 108         unsigned long norm;
 109         int i, num, ret;
 110
 111         ret = _map_file(&fr, CS_PIVOT ".pvdb");
 112         if (ret) {
 113                 if (ret == ENOENT)
 114                         ret = NO_SUCH_FILE;
 115                 return (ret);
 116         }
 117         ret = _db_open(&db1, &fr, _CITRUS_PIVOT_MAGIC, _db_hash_std, NULL);
 118         if (ret)
 119                 goto quit1;
 120         ret = open_subdb(&db2, db1, src);
 121         if (ret)
 122                 goto quit2;
 123
 124         num = _db_get_num_entries(db2);
 125         *rnorm = ULONG_MAX;
 126         for (i = 0; i < num; i++) {
 127                 /* iterate each pivot */
 128                 ret = _db_get_entry(db2, i, &r1, &r2);
 129                 if (ret)
 130                         goto quit3;
 131                 /* r1:pivot name, r2:norm among src and pivot */
 132                 ret = get32(&r2, &val32);
 133                 if (ret)
 134                         goto quit3;
 135                 norm = val32;
 136                 snprintf(buf, sizeof(buf), "%.*s",
 137                          (int)_region_size(&r1), (char *)_region_head(&r1));
 138                 /* buf: pivot name */
 139                 ret = open_subdb(&db3, db1, buf);
 140                 if (ret)
 141                         goto quit3;
 142                 if (_db_lookup_by_s(db3, dst, &r2, NULL) != 0)
 143                         /* don't break the loop, test all src/dst pairs. */
 144                         goto quit4;
 145                 /* r2: norm among pivot and dst */
 146                 ret = get32(&r2, &val32);
 147                 if (ret)
 148                         goto quit4;
 149                 norm += val32;
 150                 /* judge minimum norm */
 151                 if (norm < *rnorm) {
 152                         *rnorm = norm;
 153                         strlcpy(pivot, buf, pvlen);
 154                 }
 155 quit4:
 156                 _db_close(db3);
 157                 if (ret)
 158                         goto quit3;
 159         }
 160 quit3:
 161         _db_close(db2);
 162 quit2:
 163         _db_close(db1);
 164 quit1:
 165         _unmap_file(&fr);
 166         if (ret)
 167                 return (ret);
 168
 169         if (*rnorm == ULONG_MAX)
 170                 return (ENOENT);
 171
 172         return (0);
 173 }
 174
 175 /* ---------------------------------------------------------------------- */
 176
 177 struct zone {
 178         const char *begin, *end;
 179 };
 180
 181 struct parse_arg {
 182         char dst[PATH_MAX];
 183         unsigned long norm;
 184 };
 185
 186 static int
 187 parse_line(struct parse_arg *pa, struct _region *r)
 188 {
 189         struct zone z1, z2;
 190         char buf[20];
 191         size_t len;
 192
 193         len = _region_size(r);
 194         z1.begin = _bcs_skip_ws_len(_region_head(r), &len);
 195         if (len == 0)
 196                 return (EFTYPE);
 197         z1.end = _bcs_skip_nonws_len(z1.begin, &len);
 198         if (len == 0)
 199                 return (EFTYPE);
 200         z2.begin = _bcs_skip_ws_len(z1.end, &len);
 201         if (len == 0)
 202                 return (EFTYPE);
 203         z2.end = _bcs_skip_nonws_len(z2.begin, &len);
 204
 205         /* z1 : dst name, z2 : norm */
 206         snprintf(pa->dst, sizeof(pa->dst),
 207             "%.*s", (int)(z1.end-z1.begin), z1.begin);
 208         snprintf(buf, sizeof(buf),
 209             "%.*s", (int)(z2.end-z2.begin), z2.begin);
 210         pa->norm = _bcs_strtoul(buf, NULL, 0);
 211
 212         return (0);
 213 }
 214
 215 static int
 216 find_dst(struct parse_arg *pasrc, const char *dst)
 217 {
 218         struct _lookup *cl;
 219         struct parse_arg padst;
 220         struct _region data;
 221         int ret;
 222
 223         ret = _lookup_seq_open(&cl, CS_PIVOT, _LOOKUP_CASE_IGNORE);
 224         if (ret)
 225                 return (ret);
 226
 227         ret = _lookup_seq_lookup(cl, pasrc->dst, &data);
 228         while (ret == 0) {
 229                 ret = parse_line(&padst, &data);
 230                 if (ret)
 231                         break;
 232                 if (strcmp(dst, padst.dst) == 0) {
 233                         pasrc->norm += padst.norm;
 234                         break;
 235                 }
 236                 ret = _lookup_seq_next(cl, NULL, &data);
 237         }
 238         _lookup_seq_close(cl);
 239
 240         return (ret);
 241 }
 242
 243 static int
 244 find_best_pivot_lookup(const char *src, const char *dst, char *pivot,
 245     size_t pvlen, unsigned long *rnorm)
 246 {
 247         struct _lookup *cl;
 248         struct _region data;
 249         struct parse_arg pa;
 250         char pivot_min[PATH_MAX];
 251         unsigned long norm_min;
 252         int ret;
 253
 254         ret = _lookup_seq_open(&cl, CS_PIVOT, _LOOKUP_CASE_IGNORE);
 255         if (ret)
 256                 return (ret);
 257
 258         norm_min = ULONG_MAX;
 259
 260         /* find pivot code */
 261         ret = _lookup_seq_lookup(cl, src, &data);
 262         while (ret == 0) {
 263                 ret = parse_line(&pa, &data);
 264                 if (ret)
 265                         break;
 266                 ret = find_dst(&pa, dst);
 267                 if (ret)
 268                         break;
 269                 if (pa.norm < norm_min) {
 270                         norm_min = pa.norm;
 271                         strlcpy(pivot_min, pa.dst, sizeof(pivot_min));
 272                 }
 273                 ret = _lookup_seq_next(cl, NULL, &data);
 274         }
 275         _lookup_seq_close(cl);
 276
 277         if (ret != ENOENT)
 278                 return (ret);
 279         if (norm_min == ULONG_MAX)
 280                 return (ENOENT);
 281         strlcpy(pivot, pivot_min, pvlen);
 282         if (rnorm)
 283                 *rnorm = norm_min;
 284
 285         return (0);
 286 }
 287
 288 static int
 289 find_best_pivot(const char *src, const char *dst, char *pivot, size_t pvlen,
 290     unsigned long *rnorm)
 291 {
 292         int ret;
 293
 294         ret = find_best_pivot_pvdb(src, dst, pivot, pvlen, rnorm);
 295         if (ret == NO_SUCH_FILE)
 296                 ret = find_best_pivot_lookup(src, dst, pivot, pvlen, rnorm);
 297
 298         return (ret);
 299 }
 300
 301 static __inline int
 302 open_serial_mapper(struct _citrus_mapper_area *__restrict ma,
 303     struct _citrus_mapper * __restrict * __restrict rcm,
 304     const char *src, const char *pivot, const char *dst)
 305 {
 306         char buf[PATH_MAX];
 307
 308         snprintf(buf, sizeof(buf), "%s/%s,%s/%s", src, pivot, pivot, dst);
 309
 310         return (_mapper_open_direct(ma, rcm, "mapper_serial", buf));
 311 }
 312
 313 static struct _citrus_csmapper *csm_none = NULL;
 314 static int
 315 get_none(struct _citrus_mapper_area *__restrict ma,
 316     struct _citrus_csmapper *__restrict *__restrict rcsm)
 317 {
 318         int ret;
 319
 320         WLOCK(&ma_lock);
 321         if (csm_none) {
 322                 *rcsm = csm_none;
 323                 ret = 0;
 324                 goto quit;
 325         }
 326
 327         ret = _mapper_open_direct(ma, &csm_none, "mapper_none", "");
 328         if (ret)
 329                 goto quit;
 330         _mapper_set_persistent(csm_none);
 331
 332         *rcsm = csm_none;
 333         ret = 0;
 334 quit:
 335         UNLOCK(&ma_lock);
 336         return (ret);
 337 }
 338
 339 int
 340 _citrus_csmapper_open(struct _citrus_csmapper * __restrict * __restrict rcsm,
 341     const char * __restrict src, const char * __restrict dst, uint32_t flags,
 342     unsigned long *rnorm)
 343 {
 344         const char *realsrc, *realdst;
 345         char buf1[PATH_MAX], buf2[PATH_MAX], key[PATH_MAX], pivot[PATH_MAX];
 346         unsigned long norm;
 347         int ret;
 348
 349         norm = 0;
 350
 351         ret = _citrus_mapper_create_area(&maparea, _PATH_CSMAPPER);
 352         if (ret)
 353                 return (ret);
 354
 355         realsrc = _lookup_alias(CS_ALIAS, src, buf1, sizeof(buf1),
 356             _LOOKUP_CASE_IGNORE);
 357         realdst = _lookup_alias(CS_ALIAS, dst, buf2, sizeof(buf2),
 358             _LOOKUP_CASE_IGNORE);
 359         if (!strcmp(realsrc, realdst)) {
 360                 ret = get_none(maparea, rcsm);
 361                 if (ret == 0 && rnorm != NULL)
 362                         *rnorm = 0;
 363                 return (ret);
 364         }
 365
 366         snprintf(key, sizeof(key), "%s/%s", realsrc, realdst);
 367
 368         ret = _mapper_open(maparea, rcsm, key);
 369         if (ret == 0) {
 370                 if (rnorm != NULL)
 371                         *rnorm = 0;
 372                 return (0);
 373         }
 374         if (ret != ENOENT || (flags & _CSMAPPER_F_PREVENT_PIVOT)!=0)
 375                 return (ret);
 376
 377         ret = find_best_pivot(realsrc, realdst, pivot, sizeof(pivot), &norm);
 378         if (ret)
 379                 return (ret);
 380
 381         ret = open_serial_mapper(maparea, rcsm, realsrc, pivot, realdst);
 382         if (ret == 0 && rnorm != NULL)
 383                 *rnorm = norm;
 384
 385         return (ret);
 386 }