X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=util%2Fcharmap.c;h=8e95b004ff7ff4678e9fe4c9fcc8df4028dee059;hb=c9ac021f1381269609f3654384698f398cf46b96;hp=112aba29641a6c889e3f9dc0589be7f970c06e58;hpb=003ed9e50b40a26c54bf79cab71bd26f7957b870;p=idzebra-moved-to-github.git diff --git a/util/charmap.c b/util/charmap.c index 112aba2..8e95b00 100644 --- a/util/charmap.c +++ b/util/charmap.c @@ -1,10 +1,44 @@ /* - * Copyright (C) 1994, Index Data I/S + * Copyright (C) 1996-1997, Index Data I/S * All rights reserved. * Sebastian Hammer, Adam Dickmeiss * * $Log: charmap.c,v $ - * Revision 1.1 1996-05-31 09:07:18 quinn + * Revision 1.12 1997-09-05 15:30:11 adam + * Changed prototype for chr_map_input - added const. + * Added support for C++, headers uses extern "C" for public definitions. + * + * Revision 1.11 1997/09/05 09:52:32 adam + * Extra argument added to function chr_read_maptab (tab path). + * + * Revision 1.10 1997/07/01 13:01:08 adam + * Bug fix in routine find_entry: didn't take into account the len arg. + * + * Revision 1.9 1996/10/29 13:48:14 adam + * Updated to use zebrautl.h instead of alexutil.h. + * + * Revision 1.8 1996/10/18 12:39:23 adam + * Uses LOG_DEBUG instead of LOG_WARN for "Character map overlap". + * + * Revision 1.7 1996/06/06 12:08:56 quinn + * Fixed bug. + * + * Revision 1.6 1996/06/04 13:28:00 quinn + * More work on charmapping + * + * Revision 1.5 1996/06/04 08:32:15 quinn + * Moved default keymap to keychars.c + * + * Revision 1.4 1996/06/03 16:32:13 quinn + * Temporary bug-fix + * + * Revision 1.3 1996/06/03 15:17:46 quinn + * Fixed bug. + * + * Revision 1.2 1996/06/03 10:15:09 quinn + * Fixed bug in mapping function. + * + * Revision 1.1 1996/05/31 09:07:18 quinn * Work on character-set handling * * @@ -16,18 +50,21 @@ */ #include +#include +#include -#include +#include #include #include #include +#define CHR_MAXSTR 1024 +#define CHR_MAXEQUIV 32 + const char *CHR_UNKNOWN = "\001"; const char *CHR_SPACE = "\002"; const char *CHR_BASE = "\003"; -extern char *data1_tabpath; - /* * Character map trie node. */ @@ -36,13 +73,22 @@ struct chr_t_entry chr_t_entry **children; /* array of children */ unsigned char *target; /* target for this node, if any */ unsigned char *equiv; /* equivalent to, or sumthin */ -} t_entry; +}; + +/* + * General argument structure for callback functions (internal use only) + */ +typedef struct chrwork +{ + chrmaptab *map; + char string[CHR_MAXSTR+1]; +} chrwork; /* * Add an entry to the character map. */ -static chr_t_entry *set_map_string(chr_t_entry *root, char *from, int len, - char *to) +static chr_t_entry *set_map_string(chr_t_entry *root, const char *from, + int len, char *to) { if (!root) { @@ -51,7 +97,13 @@ static chr_t_entry *set_map_string(chr_t_entry *root, char *from, int len, root->target = 0; } if (!len) - root->target = (unsigned char *) xstrdup(to); + { + if (!root->target || (char*) root->target == CHR_SPACE || + (char*) root->target == CHR_UNKNOWN) + root->target = (unsigned char *) xstrdup(to); + else if ((char*) to != CHR_SPACE) + logf(LOG_DEBUG, "Character map overlap"); + } else { if (!root->children) @@ -62,9 +114,10 @@ static chr_t_entry *set_map_string(chr_t_entry *root, char *from, int len, for (i = 0; i < 256; i++) root->children[i] = 0; } - root->children[(unsigned char) *from] = + if (!(root->children[(unsigned char) *from] = set_map_string(root->children[(unsigned char) *from], from + 1, - len - 1, to); + len - 1, to))) + return 0; } return root; } @@ -94,39 +147,38 @@ int chr_map_chrs(chr_t_entry *t, char **from, int len, int *read, char **to, return i; } -char **chr_map_input(chr_t_entry *t, char **from, int len) + +static chr_t_entry *find_entry(chr_t_entry *t, const char **from, int len) { - static char *buf[2] = {0, 0}, str[2] = {0, 0}; - char *start = *from; + chr_t_entry *res; - if (t) + if (len && t->children && t->children[(unsigned char) **from]) { - while (len && t->children && t->children[(unsigned char) **from]) - { - t = t->children[(unsigned char) **from]; - (*from)++; - len--; - } - buf[0] = (char*) t->target; - } - else /* null mapping */ - { - if (isalnum(**from)) - { - str[0] = **from; - buf[0] = str; - } - else if (isspace(**from)) - buf[0] = (char*) CHR_SPACE; - else - buf[0] = (char*) CHR_UNKNOWN; - } - if (start == *from) + const char *pos = *from; + (*from)++; + if ((res = find_entry(t->children[(unsigned char) *pos], + from, len - 1))) + return res; + /* no match */ + *from = pos; + } + /* no children match. use ourselves, if we have a target */ + return t->target ? t : 0; +} + +const char **chr_map_input(chr_t_entry *t, const char **from, int len) +{ + static const char *buf[2] = {0, 0}; + chr_t_entry *res; + + if (!(res = find_entry(t, from, len))) + abort(); + buf[0] = (char *) res->target; return buf; - /* return (char*) t->target; */ } + static unsigned char prim(char **s) { unsigned char c; @@ -160,7 +212,7 @@ static unsigned char prim(char **s) * Callback function. * Add an entry to the value space. */ -static void fun_addentry(char *s, void *data, int num) +static void fun_addentry(const char *s, void *data, int num) { chrmaptab *tab = data; char tmp[2]; @@ -174,16 +226,42 @@ static void fun_addentry(char *s, void *data, int num) * Callback function. * Add a space-entry to the value space. */ -static void fun_addspace(char *s, void *data, int num) +static void fun_addspace(const char *s, void *data, int num) { chrmaptab *tab = data; tab->input = set_map_string(tab->input, s, strlen(s), (char*) CHR_SPACE); } -static int scan_string(char *s, void (*fun)(char *c, void *data, int num), +/* + * Create a string containing the mapped characters provided. + */ +static void fun_mkstring(const char *s, void *data, int num) +{ + chrwork *arg = data; + const char **res, *p = s; + + res = chr_map_input(arg->map->input, &s, strlen(s)); + if (*res == (char*) CHR_UNKNOWN) + logf(LOG_WARN, "Map: '%s' has no mapping", p); + strncat(arg->string, *res, CHR_MAXSTR - strlen(arg->string)); + arg->string[CHR_MAXSTR] = '\0'; +} + +/* + * Add a map to the string contained in the argument. + */ +static void fun_addmap(const char *s, void *data, int num) +{ + chrwork *arg = data; + + assert(arg->map->input); + set_map_string(arg->map->input, s, strlen(s), arg->string); +} + +static int scan_string(char *s, void (*fun)(const char *c, void *data, int num), void *data, int *num) { - unsigned char c, str[1024], begin, end; + unsigned char c, str[1024], begin, end, *p; while (*s) { @@ -212,7 +290,22 @@ static int scan_string(char *s, void (*fun)(char *c, void *data, int num), } break; case '[': s++; abort(); break; - case '(': s++; abort(); break; + case '(': + p = (unsigned char*) ++s; + /* Find the end-marker, ignoring escapes */ + do + { + if (!(p = (unsigned char*) strchr((char*) p, ')'))) + { + logf(LOG_FATAL, "Missing ')' in string"); + return -1; + } + } + while (*(p - 1) == '\\'); + *p = 0; + (*fun)(s, data, num ? (*num)++ : 0); + s = (char*) p + 1; + break; default: c = prim(&s); str[0] = c; str[1] = '\0'; @@ -222,14 +315,14 @@ static int scan_string(char *s, void (*fun)(char *c, void *data, int num), return 0; } -chrmaptab *chr_read_maptab(char *name) +chrmaptab *chr_read_maptab(const char *tabpath, const char *name) { FILE *f; char line[512], *argv[50]; chrmaptab *res = xmalloc(sizeof(*res)); int argc, num = (int) *CHR_BASE, i; - if (!(f = yaz_path_fopen(data1_tabpath, name, "r"))) + if (!(f = yaz_path_fopen(tabpath, name, "r"))) { logf(LOG_WARN|LOG_ERRNO, "%s", name); return 0; @@ -238,27 +331,21 @@ chrmaptab *chr_read_maptab(char *name) res->input = xmalloc(sizeof(*res->input)); res->input->target = (unsigned char*) CHR_UNKNOWN; res->input->equiv = 0; -#if 0 +#if 1 res->input->children = xmalloc(sizeof(res->input) * 256); for (i = 0; i < 256; i++) { res->input->children[i] = xmalloc(sizeof(*res->input)); res->input->children[i]->children = 0; - res->input->children[i]->target = CHR_UNKNOWN; + res->input->children[i]->target = (unsigned char*) CHR_UNKNOWN; res->input->children[i]->equiv = 0; } #else res->input->children = 0; #endif res->query_equiv = 0; - for (i = 0; i < 256; i++) - { - char *t = xmalloc(2); - - t[0] = i; - t[1] = '\0'; - res->output[i] = (unsigned char*)t; - } + for (i = *CHR_BASE; i < 256; i++) + res->output[i] = 0; res->output[(int) *CHR_SPACE] = (unsigned char *) " "; res->output[(int) *CHR_UNKNOWN] = (unsigned char*) "@"; res->base_uppercase = 0; @@ -312,13 +399,38 @@ chrmaptab *chr_read_maptab(char *name) fclose(f); return 0; } - if (scan_string(argv[1], fun_addspace, res, 0)) + if (scan_string(argv[1], fun_addspace, res, 0) < 0) { logf(LOG_FATAL, "Bad space specification"); fclose(f); return 0; } } + else if (!yaz_matchstr(argv[0], "map")) + { + chrwork buf; + + if (argc != 3) + { + logf(LOG_FATAL, "charmap MAP directive requires 2 args"); + fclose(f); + return 0; + } + buf.map = res; + buf.string[0] = '\0'; + if (scan_string(argv[2], fun_mkstring, &buf, 0) < 0) + { + logf(LOG_FATAL, "Bad map target"); + fclose(f); + return 0; + } + if (scan_string(argv[1], fun_addmap, &buf, 0) < 0) + { + logf(LOG_FATAL, "Bad map source"); + fclose(f); + return 0; + } + } else { #if 0