X-Git-Url: http://git.indexdata.com/?p=idzebra-moved-to-github.git;a=blobdiff_plain;f=util%2Fcharmap.c;h=d38be4f6d62254142e54ccd49db52137f1774f2a;hp=96a390ab14b8f230a87e33d0ee8a73fc4f2326aa;hb=ecb3935e78cd9bcfdebafdee0834cfb1060d7b5e;hpb=d9b85731ab7b965b1ae9bc1c283e39faf10a644a diff --git a/util/charmap.c b/util/charmap.c index 96a390a..d38be4f 100644 --- a/util/charmap.c +++ b/util/charmap.c @@ -1,6 +1,6 @@ -/* $Id: charmap.c,v 1.29 2004-07-28 09:47:42 adam Exp $ - Copyright (C) 1995,1996,1997,1998,1999,2000,2001,2002,2003,2004 - Index Data Aps +/* $Id: charmap.c,v 1.40 2006-05-10 08:13:44 adam Exp $ + Copyright (C) 1995-2005 + Index Data ApS This file is part of the Zebra server. @@ -29,17 +29,20 @@ Free Software Foundation, 59 Temple Place - Suite 330, Boston, MA #include #include +#include #include typedef unsigned ucs4_t; -#include #include +#include #define CHR_MAXSTR 1024 #define CHR_MAXEQUIV 32 +const unsigned char CHR_FIELD_BEGIN = '^'; + const char *CHR_UNKNOWN = "\001"; const char *CHR_SPACE = "\002"; const char *CHR_CUT = "\003"; @@ -73,6 +76,15 @@ typedef struct chrwork } chrwork; /* + * Callback for equivalent stuff + */ +typedef struct +{ + NMEM nmem; + int no_eq; + char *eq[CHR_MAXEQUIV]; +} chr_equiv_work; +/* * Add an entry to the character map. */ static chr_t_entry *set_map_string(chr_t_entry *root, NMEM nmem, @@ -89,13 +101,14 @@ static chr_t_entry *set_map_string(chr_t_entry *root, NMEM nmem, } if (!len) { - if (!root->target || !root->target[0] || strcmp(root->target[0], to)) + if (!root->target || !root->target[0] || + strcmp((const char *) root->target[0], to)) { if (from_0 && root->target && root->target[0] && root->target[0][0] && - strcmp (root->target[0], CHR_UNKNOWN)) + strcmp((const char *) root->target[0], CHR_UNKNOWN)) { - yaz_log (LOG_WARN, "duplicate entry for charmap from '%s'", + yaz_log (YLOG_WARN, "duplicate entry for charmap from '%s'", from_0); } root->target = (unsigned char **) @@ -142,7 +155,7 @@ static chr_t_entry *find_entry(chr_t_entry *t, const char **from, int len) return t->target ? t : 0; } -static chr_t_entry *find_entry_x(chr_t_entry *t, const char **from, int *len) +static chr_t_entry *find_entry_x(chr_t_entry *t, const char **from, int *len, int first) { chr_t_entry *res; @@ -153,35 +166,49 @@ static chr_t_entry *find_entry_x(chr_t_entry *t, const char **from, int *len) from++; len++; } - if (*len > 0 && t->children && t->children[(unsigned char) **from]) + if (*len > 0 && t->children) { const char *old_from = *from; int old_len = *len; + + res = 0; + + if (first && t->children[CHR_FIELD_BEGIN]) + { + if ((res = find_entry_x(t->children[CHR_FIELD_BEGIN], from, len, 0)) && res != t->children[CHR_FIELD_BEGIN]) + return res; + else + res = 0; + /* otherwhise there was no match on beginning of field, move on */ + } - (*len)--; - (*from)++; - if ((res = find_entry_x(t->children[(unsigned char) *old_from], - from, len))) - return res; - /* no match */ - *len = old_len; - *from = old_from; + if (!res && t->children[(unsigned char) **from]) + { + (*len)--; + (*from)++; + if ((res = find_entry_x(t->children[(unsigned char) *old_from], + from, len, 0))) + return res; + /* no match */ + *len = old_len; + *from = old_from; + } } /* no children match. use ourselves, if we have a target */ return t->target ? t : 0; } -const char **chr_map_input_x(chrmaptab maptab, const char **from, int *len) +const char **chr_map_input_x(chrmaptab maptab, const char **from, int *len, int first) { chr_t_entry *t = maptab->input; chr_t_entry *res; - if (!(res = find_entry_x(t, from, len))) + if (!(res = find_entry_x(t, from, len, first))) abort(); return (const char **) (res->target); } -const char **chr_map_input(chrmaptab maptab, const char **from, int len) +const char **chr_map_input(chrmaptab maptab, const char **from, int len, int first) { chr_t_entry *t = maptab->input; chr_t_entry *res; @@ -189,11 +216,25 @@ const char **chr_map_input(chrmaptab maptab, const char **from, int len) len_tmp[0] = len; len_tmp[1] = -1; - if (!(res = find_entry_x(t, from, len_tmp))) + if (!(res = find_entry_x(t, from, len_tmp, first))) abort(); return (const char **) (res->target); } +const char **chr_map_q_input(chrmaptab maptab, + const char **from, int len, int first) +{ + chr_t_entry *t = maptab->q_input; + chr_t_entry *res; + int len_tmp[2]; + + len_tmp[0] = len; + len_tmp[1] = -1; + if (!(res = find_entry_x(t, from, len_tmp, first))) + return 0; + return (const char **) (res->target); +} + const char *chr_map_output(chrmaptab maptab, const char **from, int len) { unsigned char c = ** (unsigned char **) from; @@ -201,50 +242,6 @@ const char *chr_map_output(chrmaptab maptab, const char **from, int len) return (const char*) maptab->output[c]; } -unsigned char zebra_prim(char **s) -{ - unsigned char c; - unsigned int i = 0; - - yaz_log (LOG_DEBUG, "prim %.3s", *s); - if (**s == '\\') - { - (*s)++; - c = **s; - switch (c) - { - case '\\': c = '\\'; (*s)++; break; - case 'r': c = '\r'; (*s)++; break; - case 'n': c = '\n'; (*s)++; break; - case 't': c = '\t'; (*s)++; break; - case 's': c = ' '; (*s)++; break; - case 'x': sscanf(*s, "x%2x", &i); c = i; *s += 3; break; - case '0': - case '1': - case '2': - case '3': - case '4': - case '5': - case '6': - case '7': - case '8': - case '9': - sscanf(*s, "%3o", &i); - c = i; - *s += 3; - break; - default: - (*s)++; - } - } - else - { - c = **s; - ++(*s); - } - return c; -} - static int zebra_ucs4_strlen(ucs4_t *s) { int i = 0; @@ -259,8 +256,8 @@ ucs4_t zebra_prim_w(ucs4_t **s) ucs4_t i = 0; char fmtstr[8]; - yaz_log (LOG_DEBUG, "prim %.3s", (char *) *s); - if (**s == '\\') + yaz_log (YLOG_DEBUG, "prim_w %.3s", (char *) *s); + if (**s == '\\' && 1[*s]) { (*s)++; c = **s; @@ -325,7 +322,7 @@ ucs4_t zebra_prim_w(ucs4_t **s) c = **s; ++(*s); } - yaz_log (LOG_DEBUG, "out %d", c); + yaz_log (YLOG_DEBUG, "out %d", c); return c; } @@ -374,41 +371,38 @@ static void fun_mkstring(const char *s, void *data, int num) chrwork *arg = (chrwork *) data; const char **res, *p = s; - res = chr_map_input(arg->map, &s, strlen(s)); + res = chr_map_input(arg->map, &s, strlen(s), 0); if (*res == (char*) CHR_UNKNOWN) - logf(LOG_WARN, "Map: '%s' has no mapping", p); + yaz_log(YLOG_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. + * Create an unmodified string (scan_string handler). */ -static void fun_add_map(const char *s, void *data, int num) +static void fun_add_equivalent_string(const char *s, void *data, int num) { - chrwork *arg = (chrwork *) data; - - assert(arg->map->input); - logf (LOG_DEBUG, "set map %.*s", (int) strlen(s), s); - set_map_string(arg->map->input, arg->map->nmem, s, strlen(s), arg->string, - 0); - for (s = arg->string; *s; s++) - logf (LOG_DEBUG, " %3d", (unsigned char) *s); + chr_equiv_work *arg = (chr_equiv_work *) data; + + if (arg->no_eq == CHR_MAXEQUIV) + return; + arg->eq[arg->no_eq++] = nmem_strdup(arg->nmem, s); } /* - * Add a query map to the string contained in the argument. + * Add a map to the string contained in the argument. */ -static void fun_add_qmap(const char *s, void *data, int num) +static void fun_add_map(const char *s, void *data, int num) { chrwork *arg = (chrwork *) data; - assert(arg->map->q_input); - logf (LOG_DEBUG, "set qmap %.*s", (int) strlen(s), s); - set_map_string(arg->map->q_input, arg->map->nmem, s, - strlen(s), arg->string, 0); + assert(arg->map->input); + yaz_log (YLOG_DEBUG, "set map %.*s", (int) strlen(s), s); + set_map_string(arg->map->input, arg->map->nmem, s, strlen(s), arg->string, + 0); for (s = arg->string; *s; s++) - logf (LOG_DEBUG, " %3d", (unsigned char) *s); + yaz_log (YLOG_DEBUG, " %3d", (unsigned char) *s); } static int scan_to_utf8 (yaz_iconv_t t, ucs4_t *from, size_t inlen, @@ -425,9 +419,9 @@ static int scan_to_utf8 (yaz_iconv_t t, ucs4_t *from, size_t inlen, ret = yaz_iconv (t, &inbuf, &inbytesleft, &outbuf, &outbytesleft); if (ret == (size_t) (-1)) { - yaz_log(LOG_LOG, "from: %2X %2X %2X %2X", + yaz_log(YLOG_LOG, "from: %2X %2X %2X %2X", from[0], from[1], from[2], from[3]); - yaz_log (LOG_WARN|LOG_ERRNO, "bad unicode sequence"); + yaz_log (YLOG_WARN|YLOG_ERRNO, "bad unicode sequence"); return -1; } } @@ -443,6 +437,7 @@ static int scan_string(char *s_native, char str[1024]; ucs4_t arg[512]; + ucs4_t arg_prim[512]; ucs4_t *s0, *s = arg; ucs4_t c, begin, end; size_t i; @@ -477,14 +472,14 @@ static int scan_string(char *s_native, begin = zebra_prim_w(&s); if (*s != '-') { - logf(LOG_FATAL, "Bad range in char-map"); + yaz_log(YLOG_FATAL, "Bad range in char-map"); return -1; } s++; end = zebra_prim_w(&s); if (end <= begin) { - logf(LOG_FATAL, "Bad range in char-map"); + yaz_log(YLOG_FATAL, "Bad range in char-map"); return -1; } s++; @@ -495,14 +490,13 @@ static int scan_string(char *s_native, (*fun)(str, data, num ? (*num)++ : 0); } break; - case '[': s++; abort(); break; case '(': ++s; - s0 = s; - while (*s != ')' || s[-1] == '\\') - s++; - *s = 0; - if (scan_to_utf8 (t_utf8, s0, s - s0, str, sizeof(str)-1)) + s0 = s; i = 0; + while (*s != ')' || s[-1] == '\\') + arg_prim[i++] = zebra_prim_w(&s); + arg_prim[i] = 0; + if (scan_to_utf8 (t_utf8, arg_prim, zebra_ucs4_strlen(arg_prim), str, sizeof(str)-1)) return -1; (*fun)(str, data, num ? (*num)++ : 0); s++; @@ -517,8 +511,8 @@ static int scan_string(char *s_native, return 0; } -chrmaptab chrmaptab_create(const char *tabpath, const char *name, int map_only, - const char *tabroot) +chrmaptab chrmaptab_create(const char *tabpath, const char *name, + const char *tabroot) { FILE *f; char line[512], *argv[50]; @@ -532,17 +526,18 @@ chrmaptab chrmaptab_create(const char *tabpath, const char *name, int map_only, unsigned endian = 31; const char *ucs4_native = "UCS-4"; + yaz_log (YLOG_DEBUG, "maptab %s open", name); + if (!(f = yaz_fopen(tabpath, name, "r", tabroot))) + { + yaz_log(YLOG_WARN|YLOG_ERRNO, "%s", name); + return 0; + } + if (*(char*) &endian == 31) /* little endian? */ ucs4_native = "UCS-4LE"; t_utf8 = yaz_iconv_open ("UTF-8", ucs4_native); - logf (LOG_DEBUG, "maptab %s open", name); - if (!(f = yaz_fopen(tabpath, name, "r", tabroot))) - { - logf(LOG_WARN|LOG_ERRNO, "%s", name); - return 0; - } nmem = nmem_create (); res = (chrmaptab) nmem_malloc(nmem, sizeof(*res)); res->nmem = nmem; @@ -561,15 +556,7 @@ chrmaptab chrmaptab_create(const char *tabpath, const char *name, int map_only, res->input->children[i]->target = (unsigned char **) nmem_malloc (res->nmem, 2 * sizeof(unsigned char *)); res->input->children[i]->target[1] = 0; - if (map_only) - { - res->input->children[i]->target[0] = (unsigned char *) - nmem_malloc (res->nmem, 2 * sizeof(unsigned char)); - res->input->children[i]->target[0][0] = i; - res->input->children[i]->target[0][1] = 0; - } - else - res->input->children[i]->target[0] = (unsigned char*) CHR_UNKNOWN; + res->input->children[i]->target[0] = (unsigned char*) CHR_UNKNOWN; } res->q_input = (chr_t_entry *) nmem_malloc(res->nmem, sizeof(*res->q_input)); @@ -583,17 +570,17 @@ chrmaptab chrmaptab_create(const char *tabpath, const char *name, int map_only, res->base_uppercase = 0; while (!errors && (argc = readconf_line(f, &lineno, line, 512, argv, 50))) - if (!map_only && !yaz_matchstr(argv[0], "lowercase")) + if (!yaz_matchstr(argv[0], "lowercase")) { if (argc != 2) { - logf(LOG_FATAL, "Syntax error in charmap"); + yaz_log(YLOG_FATAL, "Syntax error in charmap"); ++errors; } if (scan_string(argv[1], t_unicode, t_utf8, fun_addentry, res, &num) < 0) { - logf(LOG_FATAL, "Bad value-set specification"); + yaz_log(YLOG_FATAL, "Bad value-set specification"); ++errors; } res->base_uppercase = num; @@ -601,50 +588,50 @@ chrmaptab chrmaptab_create(const char *tabpath, const char *name, int map_only, res->output[(int) *CHR_UNKNOWN + num] = (unsigned char*) "@"; num = (int) *CHR_BASE; } - else if (!map_only && !yaz_matchstr(argv[0], "uppercase")) + else if (!yaz_matchstr(argv[0], "uppercase")) { if (!res->base_uppercase) { - logf(LOG_FATAL, "Uppercase directive with no lowercase set"); + yaz_log(YLOG_FATAL, "Uppercase directive with no lowercase set"); ++errors; } if (argc != 2) { - logf(LOG_FATAL, "Missing arg for uppercase directive"); + yaz_log(YLOG_FATAL, "Missing arg for uppercase directive"); ++errors; } if (scan_string(argv[1], t_unicode, t_utf8, fun_addentry, res, &num) < 0) { - logf(LOG_FATAL, "Bad value-set specification"); + yaz_log(YLOG_FATAL, "Bad value-set specification"); ++errors; } } - else if (!map_only && !yaz_matchstr(argv[0], "space")) + else if (!yaz_matchstr(argv[0], "space")) { if (argc != 2) { - logf(LOG_FATAL, "Syntax error in charmap for space"); + yaz_log(YLOG_FATAL, "Syntax error in charmap for space"); ++errors; } if (scan_string(argv[1], t_unicode, t_utf8, fun_addspace, res, 0) < 0) { - logf(LOG_FATAL, "Bad space specification"); + yaz_log(YLOG_FATAL, "Bad space specification"); ++errors; } } - else if (!map_only && !yaz_matchstr(argv[0], "cut")) + else if (!yaz_matchstr(argv[0], "cut")) { if (argc != 2) { - logf(LOG_FATAL, "Syntax error in charmap for cut"); + yaz_log(YLOG_FATAL, "Syntax error in charmap for cut"); ++errors; } if (scan_string(argv[1], t_unicode, t_utf8, fun_addcut, res, 0) < 0) { - logf(LOG_FATAL, "Bad cut specification"); + yaz_log(YLOG_FATAL, "Bad cut specification"); ++errors; } } @@ -654,7 +641,7 @@ chrmaptab chrmaptab_create(const char *tabpath, const char *name, int map_only, if (argc != 3) { - logf(LOG_FATAL, "charmap directive map requires 2 args"); + yaz_log(YLOG_FATAL, "charmap directive map requires 2 args"); ++errors; } buf.map = res; @@ -662,39 +649,68 @@ chrmaptab chrmaptab_create(const char *tabpath, const char *name, int map_only, if (scan_string(argv[2], t_unicode, t_utf8, fun_mkstring, &buf, 0) < 0) { - logf(LOG_FATAL, "Bad map target"); + yaz_log(YLOG_FATAL, "Bad map target"); ++errors; } if (scan_string(argv[1], t_unicode, t_utf8, fun_add_map, &buf, 0) < 0) { - logf(LOG_FATAL, "Bad map source"); + yaz_log(YLOG_FATAL, "Bad map source"); ++errors; } } - else if (!yaz_matchstr(argv[0], "qmap")) + else if (!yaz_matchstr(argv[0], "equivalent")) { - chrwork buf; + chr_equiv_work w; - if (argc != 3) + if (argc != 2) { - logf(LOG_FATAL, "charmap directive qmap requires 2 args"); + yaz_log(YLOG_FATAL, "equivalent requires 1 argument"); ++errors; } - buf.map = res; - buf.string[0] = '\0'; - if (scan_string(argv[2], t_unicode, t_utf8, - fun_mkstring, &buf, 0) < 0) + w.nmem = res->nmem; + w.no_eq = 0; + if (scan_string(argv[1], t_unicode, t_utf8, + fun_add_equivalent_string, &w, 0) < 0) { - logf(LOG_FATAL, "Bad qmap target"); + yaz_log(YLOG_FATAL, "equivalent: invalid string"); ++errors; } - if (scan_string(argv[1], t_unicode, t_utf8, - fun_add_qmap, &buf, 0) < 0) + else if (w.no_eq == 0) { - logf(LOG_FATAL, "Bad qmap source"); + yaz_log(YLOG_FATAL, "equivalent: no strings"); ++errors; } + else + { + char *result_str; + int i, slen = 5; + + /* determine length of regular expression */ + for (i = 0; inmem, slen + 5); + + /* build the regular expression */ + *result_str = '\0'; + slen = 0; + for (i = 0; iq_input, res->nmem, + w.eq[i], strlen(w.eq[i]), + result_str, 0); + } + } } else if (!yaz_matchstr(argv[0], "encoding")) { @@ -723,7 +739,7 @@ chrmaptab chrmaptab_create(const char *tabpath, const char *name, int map_only, } else { - logf(LOG_WARN, "Syntax error at '%s' in %s", line, name); + yaz_log(YLOG_WARN, "Syntax error at '%s' in %s", line, name); } yaz_fclose(f); @@ -732,7 +748,7 @@ chrmaptab chrmaptab_create(const char *tabpath, const char *name, int map_only, chrmaptab_destroy(res); res = 0; } - logf (LOG_DEBUG, "maptab %s close %d errors", name, errors); + yaz_log (YLOG_DEBUG, "maptab %s close %d errors", name, errors); if (t_utf8 != 0) yaz_iconv_close(t_utf8); if (t_unicode != 0) @@ -747,3 +763,11 @@ void chrmaptab_destroy(chrmaptab tab) } +/* + * Local variables: + * c-basic-offset: 4 + * indent-tabs-mode: nil + * End: + * vim: shiftwidth=4 tabstop=8 expandtab + */ +