lib/libc/citrus/citrus_csmapper.c

   1 /* $NetBSD: citrus_csmapper.c,v 1.8 2008/02/09 14:56:20 junyoung Exp $ */
   2 /* $DragonFly: src/lib/libc/citrus/citrus_csmapper.c,v 1.2 2008/04/10 10:21:01 hasso 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 "namespace.h"
  31 #include <sys/types.h>
  32 #include <sys/endian.h>
  33 #include <sys/queue.h>
  34 #include <assert.h>
  35 #include <errno.h>
  36 #include <limits.h>
  37 #include <paths.h>
  38 #include <pthread.h>
  39 #include <stdio.h>
  40 #include <stdlib.h>
  41 #include <string.h>
  42 #include "un-namespace.h"
  43
  44 #include "libc_private.h"
  45
  46 #include "citrus_namespace.h"
  47 #include "citrus_types.h"
  48 #include "citrus_bcs.h"
  49 #include "citrus_region.h"
  50 #include "citrus_memstream.h"
  51 #include "citrus_mmap.h"
  52 #include "citrus_module.h"
  53 #include "citrus_hash.h"
  54 #include "citrus_mapper.h"
  55 #include "citrus_csmapper.h"
  56 #include "citrus_pivot_file.h"
  57 #include "citrus_db.h"
  58 #include "citrus_db_hash.h"
  59 #include "citrus_lookup.h"
  60
  61 static pthread_mutex_t lock = PTHREAD_MUTEX_INITIALIZER;
  62 static struct _citrus_mapper_area *maparea = NULL;
  63
  64 #define CS_ALIAS        _PATH_CSMAPPER "/charset.alias"
  65 #define CS_PIVOT        _PATH_CSMAPPER "/charset.pivot"
  66
  67
  68 /* ---------------------------------------------------------------------- */
  69
  70 static int
  71 get32(struct _region *r, uint32_t *rval)
  72 {
  73         if (_region_size(r) != 4)
  74                 return EFTYPE;
  75
  76         memcpy(rval, _region_head(r), 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         int ret;
  86         struct _region r;
  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         int ret, num, i;
 105         struct _region fr, r1, r2;
 106         struct _citrus_db *db1, *db2, *db3;
 107         char buf[LINE_MAX];
 108         unsigned long norm;
 109         uint32_t val32;
 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                         goto quit4;
 144                 /* r2: norm among pivot and dst */
 145                 ret = get32(&r2, &val32);
 146                 if (ret)
 147                         goto quit4;
 148                 norm += val32;
 149                 /* judge minimum norm */
 150                 if (norm < *rnorm) {
 151                         *rnorm = norm;
 152                         strlcpy(pivot, buf, pvlen);
 153                 }
 154 quit4:
 155                 _db_close(db3);
 156                 if (ret)
 157                         goto quit3;
 158         }
 159 quit3:
 160         _db_close(db2);
 161 quit2:
 162         _db_close(db1);
 163 quit1:
 164         _unmap_file(&fr);
 165         if (ret)
 166                 return ret;
 167
 168         if (*rnorm == ULONG_MAX)
 169                 return ENOENT;
 170
 171         return 0;
 172 }
 173
 174 /* ---------------------------------------------------------------------- */
 175
 176 struct zone {
 177         const char *begin, *end;
 178 };
 179
 180 struct parse_arg {
 181         char dst[PATH_MAX];
 182         unsigned long norm;
 183 };
 184
 185 static int
 186 parse_line(struct parse_arg *pa, struct _region *r)
 187 {
 188         char buf[20];
 189         struct zone z1, z2;
 190         size_t len;
 191
 192         len = _region_size(r);
 193         z1.begin = _bcs_skip_ws_len(_region_head(r), &len);
 194         if (len == 0)
 195                 return EFTYPE;
 196         z1.end = _bcs_skip_nonws_len(z1.begin, &len);
 197         if (len == 0)
 198                 return EFTYPE;
 199         z2.begin = _bcs_skip_ws_len(z1.end, &len);
 200         if (len == 0)
 201                 return EFTYPE;
 202         z2.end = _bcs_skip_nonws_len(z2.begin, &len);
 203
 204         /* z1 : dst name, z2 : norm */
 205         snprintf(pa->dst, sizeof(pa->dst),
 206                  "%.*s", (int)(z1.end-z1.begin), z1.begin);
 207         snprintf(buf, sizeof(buf),
 208                  "%.*s", (int)(z2.end-z2.begin), z2.begin);
 209         pa->norm = strtoul(buf, NULL, 0);
 210
 211         return 0;
 212 }
 213
 214 static int
 215 find_dst(struct parse_arg *pasrc, const char *dst)
 216 {
 217         int ret;
 218         struct parse_arg padst;
 219         struct _lookup *cl;
 220         struct _region data;
 221
 222         ret = _lookup_seq_open(&cl, CS_PIVOT, _LOOKUP_CASE_IGNORE);
 223         if (ret)
 224                 return ret;
 225
 226         ret = _lookup_seq_lookup(cl, pasrc->dst, &data);
 227         while (ret == 0) {
 228                 ret = parse_line(&padst, &data);
 229                 if (ret)
 230                         break;
 231                 if (strcmp(dst, padst.dst) == 0) {
 232                         pasrc->norm += padst.norm;
 233                         break;
 234                 }
 235                 ret = _lookup_seq_next(cl, NULL, &data);
 236         }
 237         _lookup_seq_close(cl);
 238
 239         return ret;
 240 }
 241
 242 static int
 243 find_best_pivot_lookup(const char *src, const char *dst, char *pivot,
 244                        size_t pvlen, unsigned long *rnorm)
 245 {
 246         int ret;
 247         struct _lookup *cl;
 248         struct _region data;
 249         struct parse_arg pa;
 250         unsigned long norm_min;
 251         char pivot_min[PATH_MAX];
 252
 253         ret = _lookup_seq_open(&cl, CS_PIVOT, _LOOKUP_CASE_IGNORE);
 254         if (ret)
 255                 return ret;
 256
 257         norm_min = ULONG_MAX;
 258
 259         /* find pivot code */
 260         ret = _lookup_seq_lookup(cl, src, &data);
 261         while (ret == 0) {
 262                 ret = parse_line(&pa, &data);
 263                 if (ret)
 264                         break;
 265                 ret = find_dst(&pa, dst);
 266                 if (ret)
 267                         break;
 268                 if (pa.norm < norm_min) {
 269                         norm_min = pa.norm;
 270                         strlcpy(pivot_min, pa.dst, sizeof(pivot_min));
 271                 }
 272                 ret = _lookup_seq_next(cl, NULL, &data);
 273         }
 274         _lookup_seq_close(cl);
 275
 276         if (ret != ENOENT)
 277                 return ret;
 278         if (norm_min == ULONG_MAX)
 279                 return ENOENT;
 280         strlcpy(pivot, pivot_min, pvlen);
 281         if (rnorm)
 282                 *rnorm = norm_min;
 283
 284         return 0;
 285 }
 286
 287 static int
 288 find_best_pivot(const char *src, const char *dst, char *pivot, size_t pvlen,
 289                 unsigned long *rnorm)
 290 {
 291         int ret;
 292
 293         ret = find_best_pivot_pvdb(src, dst, pivot, pvlen, rnorm);
 294         if (ret == NO_SUCH_FILE)
 295                 ret = find_best_pivot_lookup(src, dst, pivot, pvlen, rnorm);
 296
 297         return ret;
 298 }
 299
 300 static __inline int
 301 open_serial_mapper(struct _citrus_mapper_area *__restrict ma,
 302                    struct _citrus_mapper * __restrict * __restrict rcm,
 303                    const char *src, const char *pivot, const char *dst)
 304 {
 305         char buf[PATH_MAX];
 306
 307         snprintf(buf, sizeof(buf), "%s/%s,%s/%s", src, pivot, pivot, dst);
 308
 309         return _mapper_open_direct(ma, rcm, "mapper_serial", buf);
 310 }
 311
 312 static struct _citrus_csmapper *csm_none = NULL;
 313 static int
 314 get_none(struct _citrus_mapper_area *__restrict ma,
 315          struct _citrus_csmapper *__restrict *__restrict rcsm)
 316 {
 317         int ret;
 318
 319         if (__isthreaded)
 320                 _pthread_mutex_lock(&lock);
 321
 322         if (csm_none) {
 323                 *rcsm = csm_none;
 324                 ret = 0;
 325                 goto quit;
 326         }
 327
 328         ret = _mapper_open_direct(ma, &csm_none, "mapper_none", "");
 329         if (ret)
 330                 goto quit;
 331         _mapper_set_persistent(csm_none);
 332
 333         *rcsm = csm_none;
 334         ret = 0;
 335 quit:
 336         if (__isthreaded)
 337                 _pthread_mutex_unlock(&lock);
 338
 339         return ret;
 340 }
 341
 342 int
 343 _citrus_csmapper_open(struct _citrus_csmapper * __restrict * __restrict rcsm,
 344                       const char * __restrict src, const char * __restrict dst,
 345                       uint32_t flags, unsigned long *rnorm)
 346 {
 347         int ret;
 348         char buf1[PATH_MAX], buf2[PATH_MAX], key[PATH_MAX], pivot[PATH_MAX];
 349         const char *realsrc, *realdst;
 350         unsigned long norm;
 351
 352         norm = 0;       /* XXX gcc */
 353
 354         ret = _citrus_mapper_create_area(&maparea, _PATH_CSMAPPER);
 355         if (ret)
 356                 return ret;
 357
 358         realsrc = _lookup_alias(CS_ALIAS, src, buf1, sizeof(buf1),
 359                                 _LOOKUP_CASE_IGNORE);
 360         realdst = _lookup_alias(CS_ALIAS, dst, buf2, sizeof(buf2),
 361                                 _LOOKUP_CASE_IGNORE);
 362         if (!strcmp(realsrc, realdst)) {
 363                 ret = get_none(maparea, rcsm);
 364                 if (ret == 0 && rnorm != NULL)
 365                         *rnorm = 0;
 366                 return ret;
 367         }
 368
 369         snprintf(key, sizeof(key), "%s/%s", realsrc, realdst);
 370
 371         ret = _mapper_open(maparea, rcsm, key);
 372         if (ret == 0) {
 373                 if (rnorm != NULL)
 374                         *rnorm = 0;
 375                 return 0;
 376         }
 377         if (ret != ENOENT || (flags & _CSMAPPER_F_PREVENT_PIVOT)!=0)
 378                 return ret;
 379
 380         ret = find_best_pivot(realsrc, realdst, pivot, sizeof(pivot), &norm);
 381         if (ret)
 382                 return ret;
 383
 384         ret = open_serial_mapper(maparea, rcsm, realsrc, pivot, realdst);
 385         if (ret == 0 && rnorm != NULL)
 386                 *rnorm = norm;
 387
 388         return ret;
 389 }