X-Git-Url: http://git.indexdata.com/?p=idzebra-moved-to-github.git;a=blobdiff_plain;f=data1%2Fd1_absyn.c;h=d737d35df8c93ba4e25c3ac4774f5ef12ad33b1a;hp=83d44ae0c75cb4081a88c11a3678b790c8825f67;hb=d513d15e315601b730b0b3a6126c3163d00223fb;hpb=22963c0b23a178529bc71e5879c7e4f9c3c9b7f9 diff --git a/data1/d1_absyn.c b/data1/d1_absyn.c index 83d44ae..d737d35 100644 --- a/data1/d1_absyn.c +++ b/data1/d1_absyn.c @@ -1,8 +1,5 @@ -/* $Id: d1_absyn.c,v 1.12 2004-09-28 10:15:02 adam Exp $ - Copyright (C) 1995,1996,1997,1998,1999,2000,2001,2002,2003,2004 - Index Data Aps - -This file is part of the Zebra server. +/* This file is part of the Zebra server. + Copyright (C) 1994-2011 Index Data Zebra is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free @@ -15,26 +12,117 @@ FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. You should have received a copy of the GNU General Public License -along with Zebra; see the file LICENSE.zebra. If not, write to the -Free Software Foundation, 59 Temple Place - Suite 330, Boston, MA -02111-1307, USA. +along with this program; if not, write to the Free Software +Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA + */ +#if HAVE_CONFIG_H +#include +#endif #include #include #include #include -#include #include +#include #include +#include #include #include #define D1_MAX_NESTING 128 -#if PRIVATE_DATA1_ABSYN -#endif +struct data1_hash_table { + NMEM nmem; + int size; + struct data1_hash_entry **ar; +}; + +struct data1_hash_entry { + void *clientData; + char *str; + struct data1_hash_entry *next; +}; + +unsigned data1_hash_calc(struct data1_hash_table *ht, const char *str) +{ + unsigned v = 0; + assert(str); + while (*str) + { + if (*str >= 'a' && *str <= 'z') + v = v*65509 + *str -'a'+10; + else if (*str >= 'A' && *str <= 'Z') + v = v*65509 + *str -'A'+10; + else if (*str >= '0' && *str <= '9') + v = v*65509 + *str -'0'; + str++; + } + return v % ht->size; +} + +struct data1_hash_table *data1_hash_open(int size, NMEM nmem) +{ + int i; + struct data1_hash_table *ht = nmem_malloc(nmem, sizeof(*ht)); + ht->nmem = nmem; + ht->size = size; + if (ht->size <= 0) + ht->size = 29; + ht->ar = nmem_malloc(nmem, sizeof(*ht->ar) * ht->size); + for (i = 0; isize; i++) + ht->ar[i] = 0; + return ht; +} + +void data1_hash_insert(struct data1_hash_table *ht, const char *str, + void *clientData, int copy) +{ + char *dstr = copy ? nmem_strdup(ht->nmem, str) : (char*) str; + if (strchr(str, '?') || strchr(str, '.')) + { + int i; + for (i = 0; isize; i++) + { + struct data1_hash_entry **he = &ht->ar[i]; + for (; *he && strcmp(str, (*he)->str); he = &(*he)->next) + ; + if (!*he) + { + *he = nmem_malloc(ht->nmem, sizeof(**he)); + (*he)->str = dstr; + (*he)->next = 0; + } + (*he)->clientData = clientData; + } + } + else + { + struct data1_hash_entry **he = &ht->ar[data1_hash_calc(ht, str)]; + for (; *he && strcmp(str, (*he)->str); he = &(*he)->next) + ; + if (!*he) + { + *he = nmem_malloc(ht->nmem, sizeof(**he)); + (*he)->str = dstr; + (*he)->next = 0; + } + (*he)->clientData = clientData; + } +} + +void *data1_hash_lookup(struct data1_hash_table *ht, const char *str) +{ + struct data1_hash_entry **he = &ht->ar[data1_hash_calc(ht, str)]; + + for (; *he && yaz_matchstr(str, (*he)->str); he = &(*he)->next) + ; + if (*he) + return (*he)->clientData; + return 0; +} struct data1_systag { char *name; @@ -56,6 +144,18 @@ struct data1_attset_cache_info data1_attset_cache next; }; +data1_element *data1_mk_element(data1_handle dh) +{ + data1_element *e = nmem_malloc(data1_nmem_get(dh), sizeof(*e)); + e->name = 0; + e->tag = 0; + e->termlists = 0; + e->next = e->children = 0; + e->sub_name = 0; + e->hash = 0; + return e; +} + data1_absyn *data1_absyn_search (data1_handle dh, const char *name) { data1_absyn_cache p = *data1_absyn_cache_get (dh); @@ -66,7 +166,7 @@ data1_absyn *data1_absyn_search (data1_handle dh, const char *name) return p->absyn; p = p->next; } - return NULL; + return 0; } /* *ostrich* We need to destroy DFAs, in xp_element (xelm) definitions @@ -84,10 +184,11 @@ void data1_absyn_destroy (data1_handle dh) { data1_xpelement *xpe = abs->xp_elements; while (xpe) { - logf (LOG_DEBUG,"Destroy xp element %s",xpe->xpath_expr); - if (xpe->dfa) { dfa_delete (&xpe->dfa); } + yaz_log (YLOG_DEBUG,"Destroy xp element %s",xpe->xpath_expr); + if (xpe->dfa) + dfa_delete (&xpe->dfa); xpe = xpe->next; - } + } } p = p->next; } @@ -106,7 +207,11 @@ void data1_absyn_trav (data1_handle dh, void *handle, } } -data1_absyn *data1_absyn_add (data1_handle dh, const char *name) +static data1_absyn *data1_read_absyn(data1_handle dh, const char *file, + enum DATA1_XPATH_INDEXING en); + +static data1_absyn *data1_absyn_add(data1_handle dh, const char *name, + enum DATA1_XPATH_INDEXING en) { char fname[512]; NMEM mem = data1_nmem_get (dh); @@ -114,20 +219,21 @@ data1_absyn *data1_absyn_add (data1_handle dh, const char *name) data1_absyn_cache p = (data1_absyn_cache)nmem_malloc (mem, sizeof(*p)); data1_absyn_cache *pp = data1_absyn_cache_get (dh); - sprintf(fname, "%s.abs", name); - p->absyn = data1_read_absyn (dh, fname, 0); - p->name = nmem_strdup (mem, name); + sprintf(fname, "%.500s.abs", name); + p->absyn = data1_read_absyn(dh, fname, en); + p->name = nmem_strdup(mem, name); p->next = *pp; *pp = p; return p->absyn; } -data1_absyn *data1_get_absyn (data1_handle dh, const char *name) +data1_absyn *data1_get_absyn (data1_handle dh, const char *name, + enum DATA1_XPATH_INDEXING en) { data1_absyn *absyn; if (!(absyn = data1_absyn_search (dh, name))) - absyn = data1_absyn_add (dh, name); + absyn = data1_absyn_add (dh, name, en); return absyn; } @@ -141,47 +247,37 @@ data1_attset *data1_attset_search_name (data1_handle dh, const char *name) return p->attset; p = p->next; } - return NULL; + return 0; } -data1_attset *data1_attset_search_id (data1_handle dh, int id) +data1_attset *data1_attset_search_id(data1_handle dh, const Odr_oid *oid) { data1_attset_cache p = *data1_attset_cache_get (dh); while (p) { - if (id == p->attset->reference) + if (p->attset->oid && !oid_oidcmp(oid, p->attset->oid)) return p->attset; p = p->next; } - return NULL; + return 0; } data1_attset *data1_attset_add (data1_handle dh, const char *name) { - char fname[512], aname[512]; NMEM mem = data1_nmem_get (dh); data1_attset *attset; - - strcpy (aname, name); - sprintf(fname, "%s.att", name); - attset = data1_read_attset (dh, fname); - if (!attset) - { - char *cp; - attset = data1_read_attset (dh, name); - if (attset && (cp = strrchr (aname, '.'))) - *cp = '\0'; - } + + attset = data1_read_attset (dh, name); if (!attset) - yaz_log (LOG_WARN|LOG_ERRNO, "Couldn't load attribute set %s", name); + yaz_log (YLOG_WARN|YLOG_ERRNO, "Couldn't load attribute set %s", name); else { data1_attset_cache p = (data1_attset_cache) nmem_malloc (mem, sizeof(*p)); data1_attset_cache *pp = data1_attset_cache_get (dh); - attset->name = p->name = nmem_strdup (mem, aname); + attset->name = p->name = nmem_strdup(mem, name); p->attset = attset; p->next = *pp; *pp = p; @@ -209,11 +305,15 @@ data1_esetname *data1_getesetbyname(data1_handle dh, data1_absyn *a, return 0; } +/* we have multiple versions of data1_getelementbyname */ +#define DATA1_GETELEMENTBYTAGNAME_VERSION 1 + data1_element *data1_getelementbytagname (data1_handle dh, data1_absyn *abs, data1_element *parent, const char *tagname) { data1_element *r; + struct data1_hash_table *ht; /* It's now possible to have a data1 tree with no abstract syntax */ if ( !abs ) @@ -224,6 +324,27 @@ data1_element *data1_getelementbytagname (data1_handle dh, data1_absyn *abs, else r = parent->children; +#if DATA1_GETELEMENTBYTAGNAME_VERSION==1 + /* using hash search */ + if (!r) + return 0; + + ht = r->hash; + if (!ht) + { + /* build hash table (the first time) */ + ht = r->hash = data1_hash_open(29, data1_nmem_get(dh)); + for (; r; r = r->next) + { + data1_name *n; + + for (n = r->tag->names; n; n = n->next) + data1_hash_insert(ht, n->name, r, 0); + } + } + return data1_hash_lookup(ht, tagname); +#else + /* using linear search */ for (; r; r = r->next) { data1_name *n; @@ -233,6 +354,7 @@ data1_element *data1_getelementbytagname (data1_handle dh, data1_absyn *abs, return r; } return 0; +#endif } data1_element *data1_getelementbyname (data1_handle dh, data1_absyn *absyn, @@ -271,7 +393,7 @@ void fix_element_ref (data1_handle dh, data1_absyn *absyn, data1_element *e) if (sub_e) e->children = sub_e->elements; else - yaz_log (LOG_WARN, "Unresolved reference to sub-elements %s", + yaz_log (YLOG_WARN, "Unresolved reference to sub-elements %s", e->sub_name); } } @@ -297,78 +419,81 @@ void fix_element_ref (data1_handle dh, data1_absyn *absyn, data1_element *e) */ -const char * mk_xpath_regexp (data1_handle dh, char *expr) +static const char * mk_xpath_regexp (data1_handle dh, const char *expr) { - char *p = expr; - char *pp; - char *s; + const char *p = expr; int abs = 1; - int i; - int j; - int e=0; - int is_predicate = 0; + int e = 0; + char *stack[32]; + char *res_p, *res = 0; + size_t res_size = 1; - static char *stack[32]; - static char res[1024]; - char *r = ""; - - if (*p != '/') { return (""); } + if (*p != '/') + return (""); p++; - if (*p == '/') { abs=0; p++; } - - while (*p) { - i=0; - while (*p && !strchr("/",*p)) { - i++; p++; - } - stack[e] = (char *) nmem_malloc (data1_nmem_get (dh), i+1); + if (*p == '/') + { + abs =0; + p++; + } + while (*p) + { + int is_predicate = 0; + char *s; + int i, j; + for (i = 0; *p && !strchr("/",*p); i++, p++) + ; + res_size += (i+3); /* we'll add / between later .. */ + stack[e] = (char *) nmem_malloc(data1_nmem_get(dh), i+1); s = stack[e]; - for (j=0; j< i; j++) { - pp = p-i+j; - if (*pp == '[') { - is_predicate=1; - } - else if (*pp == ']') { - is_predicate=0; - } - else { - if (!is_predicate) { - if (*pp == '*') - *s++ = '.'; - *s++ = *pp; + for (j = 0; j < i; j++) + { + const char *pp = p-i+j; + if (*pp == '[') + is_predicate=1; + else if (*pp == ']') + is_predicate=0; + else + { + if (!is_predicate) { + if (*pp == '*') + *s++ = '.'; + *s++ = *pp; + } } - } } *s = 0; e++; - if (*p) {p++;} + if (*p) + p++; } - e--; p = &res[0]; i=0; - sprintf (p, "^"); p++; - while (e >= 0) { - /* !!! res size is not checked !!! */ - sprintf (p, "%s/",stack[e]); - p += strlen(stack[e]) + 1; - e--; + res_p = res = nmem_malloc(data1_nmem_get(dh), res_size + 10); + + if (stack[e-1][0] == '@') /* path/@attr spec (leaf is attribute) */ + strcpy(res_p, "/"); + else + strcpy(res_p, "[^@]*/"); /* path .. (index all cdata below it) */ + res_p = res_p + strlen(res_p); + while (--e >= 0) { + sprintf(res_p, "%s/", stack[e]); + res_p += strlen(stack[e]) + 1; + } + if (!abs) + { + sprintf(res_p, ".*"); + res_p += 2; } - if (!abs) { sprintf (p, ".*"); p+=2; } - sprintf (p, "$"); p++; - r = nmem_strdup (data1_nmem_get (dh), res); - yaz_log(LOG_DEBUG,"Got regexp: %s",r); - return (r); + sprintf (res_p, "$"); + res_p++; + yaz_log(YLOG_DEBUG, "Got regexp: %s", res); + return res; } -/* *ostrich* - - added arg xpelement... when called from xelm context, it's 1, saying - that ! means xpath, not element name as attribute name... - - pop, 2002-12-13 - */ -static int parse_termlists (data1_handle dh, data1_termlist ***tpp, - char *cp, const char *file, int lineno, - const char *element_name, data1_absyn *res, - int xpelement) +static int parse_termlists(data1_handle dh, data1_termlist ***tpp, + char *cp, const char *file, int lineno, + const char *element_name, data1_absyn *res, + int xpelement, + data1_attset *attset) { data1_termlist **tp = *tpp; while(1) @@ -386,7 +511,7 @@ static int parse_termlists (data1_handle dh, data1_termlist ***tpp, if (i == 0) { if (*cp) - yaz_log(LOG_WARN, + yaz_log(YLOG_WARN, "%s:%d: Syntax error in termlistspec '%s'", file, lineno, cp); break; @@ -419,22 +544,24 @@ static int parse_termlists (data1_handle dh, data1_termlist ***tpp, nmem_malloc(data1_nmem_get(dh), sizeof(**tp)); (*tp)->next = 0; - if (!xpelement) { - if (*attname == '!') + if (*attname == '!') + { + if (!xpelement && element_name) strcpy(attname, element_name); - } - if (!((*tp)->att = data1_getattbyname(dh, res->attset, - attname))) { - if ((!xpelement) || (*attname != '!')) { - yaz_log(LOG_WARN, - "%s:%d: Couldn't find att '%s' in attset", - file, lineno, attname); - return -1; - } else { - (*tp)->att = 0; + else if (xpelement) + strcpy(attname, ZEBRA_XPATH_CDATA); + } + if (attset) + { + if (!data1_getattbyname(dh, attset, attname)) + { + yaz_log(YLOG_WARN, "Index '%s' not found in attset(s)", + attname); } - } - + } + + (*tp)->index_name = nmem_strdup(data1_nmem_get(dh), attname); + assert (*(*tp)->index_name != '!'); if (r == 2 && (source = strchr(structure, ':'))) *source++ = '\0'; /* cut off structure .. */ else @@ -454,6 +581,33 @@ static int parse_termlists (data1_handle dh, data1_termlist ***tpp, return 0; } +/* quinn + * Converts a 'melm' field[$subfield] pattern to a simple xpath + */ +static int melm2xpath(char *melm, char *buf) +{ + char *dollar; + char *field = melm; + char *subfield; + char *fieldtype; + if ((dollar = strchr(melm, '$'))) { + *dollar = '\0'; + subfield = ++dollar; + } else + subfield = ""; + if (field[0] == '0' && field[1] == '0') + fieldtype = "controlfield"; + else + fieldtype = "datafield"; + sprintf(buf, "/*/%s[@tag=\"%s\"]", fieldtype, field); + if (*subfield) + sprintf(buf + strlen(buf), "/subfield[@code=\"%s\"]", subfield); + else if (field[0] != '0' || field[1] != '0') + strcat(buf, "/subfield"); + yaz_log(YLOG_DEBUG, "Created xpath: '%s'", buf); + return 0; +} + const char *data1_systag_lookup(data1_absyn *absyn, const char *tag, const char *default_value) { @@ -505,22 +659,28 @@ int read_absyn_line(FILE *f, int *lineno, char *line, int len, return argc; } -data1_marctab *data1_absyn_getmarctab(data1_handle dh, data1_absyn *absyn) +data1_marctab *data1_absyn_getmarctab(data1_handle dh, data1_node *root) { - return absyn->marc; + if (root->u.root.absyn) + return root->u.root.absyn->marc; + return 0; } -YAZ_EXPORT data1_element *data1_absyn_getelements(data1_handle dh, - data1_absyn *absyn) +data1_element *data1_absyn_getelements(data1_handle dh, + data1_node *root) { - return absyn->main_elements; + if (root->u.root.absyn) + return root->u.root.absyn->main_elements; + return 0; } -data1_absyn *data1_read_absyn (data1_handle dh, const char *file, - int file_must_exist) +static data1_absyn *data1_read_absyn(data1_handle dh, const char *file, + enum DATA1_XPATH_INDEXING default_xpath) { data1_sub_elements *cur_elements = NULL; - data1_xpelement *cur_xpelement = NULL; + data1_xpelement **cur_xpelement = NULL; + data1_attset *attset_list = data1_empty_attset(dh); + data1_attset_child **attset_childp = &attset_list->children; data1_absyn *res = 0; FILE *f; @@ -529,7 +689,6 @@ data1_absyn *data1_read_absyn (data1_handle dh, const char *file, data1_maptab **maptabp; data1_marctab **marcp; data1_termlist *all = 0; - data1_attset_child **attset_childp; data1_tagset **tagset_childp; struct data1_systag **systagsp; int level = 0; @@ -537,26 +696,19 @@ data1_absyn *data1_read_absyn (data1_handle dh, const char *file, int argc; char *argv[50], line[512]; - if (!(f = data1_path_fopen(dh, file, "r"))) - { - yaz_log(LOG_WARN|LOG_ERRNO, "Couldn't open %s", file); - if (file_must_exist) - return 0; - } + f = data1_path_fopen(dh, file, "r"); res = (data1_absyn *) nmem_malloc(data1_nmem_get(dh), sizeof(*res)); res->name = 0; - res->reference = VAL_NONE; + res->oid = 0; res->tagset = 0; res->encoding = 0; - res->enable_xpath_indexing = (f ? 0 : 1); + res->xpath_indexing = + (f ? DATA1_XPATH_INDEXING_DISABLE : default_xpath); res->systags = 0; systagsp = &res->systags; tagset_childp = &res->tagset; - res->attset = data1_empty_attset (dh); - attset_childp = &res->attset->children; - res->varset = 0; res->esetnames = 0; esetpp = &res->esetnames; @@ -567,7 +719,8 @@ data1_absyn *data1_read_absyn (data1_handle dh, const char *file, res->sub_elements = NULL; res->main_elements = NULL; res->xp_elements = NULL; - + cur_xpelement = &res->xp_elements; + while (f && (argc = read_absyn_line(f, &lineno, line, 512, argv, 50))) { char *cmd = *argv; @@ -581,7 +734,7 @@ data1_absyn *data1_read_absyn (data1_handle dh, const char *file, if (argc < 4) { - yaz_log(LOG_WARN, "%s:%d: Bad # of args to elm", file, lineno); + yaz_log(YLOG_WARN, "%s:%d: Bad # of args to elm", file, lineno); continue; } path = argv[1]; @@ -612,17 +765,12 @@ data1_absyn *data1_read_absyn (data1_handle dh, const char *file, } if (i > level+1) { - yaz_log(LOG_WARN, "%s:%d: Bad level increase", file, lineno); + yaz_log(YLOG_WARN, "%s:%d: Bad level increase", file, lineno); fclose(f); return 0; } level = i; - new_element = *ppl[level-1] = (data1_element *) - nmem_malloc(data1_nmem_get(dh), sizeof(*new_element)); - new_element->next = new_element->children = 0; - new_element->tag = 0; - new_element->termlists = 0; - new_element->sub_name = 0; + new_element = *ppl[level-1] = data1_mk_element(dh); tp = &new_element->termlists; ppl[level-1] = &new_element->next; @@ -640,14 +788,14 @@ data1_absyn *data1_read_absyn (data1_handle dh, const char *file, { if (!res->tagset) { - yaz_log(LOG_WARN, "%s:%d: No tagset loaded", file, lineno); + yaz_log(YLOG_WARN, "%s:%d: No tagset loaded", file, lineno); fclose(f); return 0; } if (!(new_element->tag = data1_gettagbynum (dh, res->tagset, type, value))) { - yaz_log(LOG_WARN, "%s:%d: Couldn't find tag %s in tagset", + yaz_log(YLOG_WARN, "%s:%d: Couldn't find tag %s in tagset", file, lineno, p); fclose(f); return 0; @@ -673,7 +821,7 @@ data1_absyn *data1_read_absyn (data1_handle dh, const char *file, } else { - yaz_log(LOG_WARN, "%s:%d: Bad element", file, lineno); + yaz_log(YLOG_WARN, "%s:%d: Bad element", file, lineno); fclose(f); return 0; } @@ -681,9 +829,8 @@ data1_absyn *data1_read_absyn (data1_handle dh, const char *file, p = termlists; if (*p != '-') { - assert (res->attset); - - if (parse_termlists (dh, &tp, p, file, lineno, name, res, 0)) + if (parse_termlists (dh, &tp, p, file, lineno, name, res, 0, + attset_list)) { fclose (f); return 0; @@ -706,73 +853,92 @@ data1_absyn *data1_read_absyn (data1_handle dh, const char *file, pop, 2003-01-17 */ - else if (!strcmp(cmd, "xelm")) { + else if (!strcmp(cmd, "xelm") || !strcmp(cmd, "melm")) { int i; char *p, *xpath_expr, *termlists; const char *regexp; - struct DFA *dfa = dfa = dfa_init(); + struct DFA *dfa = 0; data1_termlist **tp; + char melm_xpath[128]; + data1_xpelement *xp_ele = 0; + data1_xpelement *last_match = 0; - if (argc < 3) + if (argc != 3) { - yaz_log(LOG_WARN, "%s:%d: Bad # of args to xelm", file, lineno); + yaz_log(YLOG_WARN, "%s:%d: Bad # of args to %s", + file, lineno, cmd); continue; } - xpath_expr = argv[1]; + + if (!strcmp(cmd, "melm")) { + if (melm2xpath(argv[1], melm_xpath) < 0) + continue; + xpath_expr = melm_xpath; + } else { + xpath_expr = argv[1]; + } termlists = argv[2]; regexp = mk_xpath_regexp(dh, xpath_expr); - i = dfa_parse (dfa, ®exp); - if (i || *regexp) { - yaz_log(LOG_WARN, "%s:%d: Bad xpath to xelm", file, lineno); - dfa_delete (&dfa); - continue; - } - - if (!cur_xpelement) - { - cur_xpelement = (data1_xpelement *) - nmem_malloc(data1_nmem_get(dh), sizeof(*cur_xpelement)); - res->xp_elements = cur_xpelement; - } else { - cur_xpelement->next = (data1_xpelement *) - nmem_malloc(data1_nmem_get(dh), sizeof(*cur_xpelement)); - cur_xpelement = cur_xpelement->next; - } - cur_xpelement->next = NULL; - cur_xpelement->xpath_expr = nmem_strdup(data1_nmem_get (dh), - xpath_expr); - - dfa_mkstate (dfa); - cur_xpelement->dfa = dfa; -#ifdef ENHANCED_XELM - cur_xpelement->xpath_len = - zebra_parse_xpath_str(xpath_expr, - cur_xpelement->xpath, XPATH_STEP_COUNT, - data1_nmem_get(dh)); +#if OPTIMIZE_MELM + /* get last of existing regulars with same regexp */ + for (xp_ele = res->xp_elements; xp_ele; xp_ele = xp_ele->next) + if (!strcmp(xp_ele->regexp, regexp)) + last_match = xp_ele; +#endif + if (!last_match) + { + /* new regular expression . Parse + generate */ + const char *regexp_ptr = regexp; + + dfa = dfa_init(); + i = dfa_parse (dfa, ®exp_ptr); + if (i || *regexp_ptr) { + yaz_log(YLOG_WARN, "%s:%d: Bad xpath to xelm", file, lineno); + dfa_delete (&dfa); + continue; + } + } + *cur_xpelement = (data1_xpelement *) + nmem_malloc(data1_nmem_get(dh), sizeof(**cur_xpelement)); + (*cur_xpelement)->next = 0; + (*cur_xpelement)->match_next = 0; + if (last_match) + last_match->match_next = *cur_xpelement; +#if OPTIMIZE_MELM + (*cur_xpelement)->regexp = regexp; +#endif + (*cur_xpelement)->xpath_expr = nmem_strdup(data1_nmem_get (dh), + xpath_expr); + + if (dfa) + dfa_mkstate (dfa); + (*cur_xpelement)->dfa = dfa; - /* - dump_xp_steps(cur_xpelement->xpath,cur_xpelement->xpath_len); - */ +#ifdef ENHANCED_XELM + (*cur_xpelement)->xpath_len = + zebra_parse_xpath_str( + xpath_expr, + (*cur_xpelement)->xpath, XPATH_STEP_COUNT, + data1_nmem_get(dh)); #endif - cur_xpelement->termlists = 0; - tp = &cur_xpelement->termlists; + (*cur_xpelement)->termlists = 0; + tp = &(*cur_xpelement)->termlists; /* parse termList definitions */ p = termlists; if (*p != '-') { - assert (res->attset); - if (parse_termlists (dh, &tp, p, file, lineno, - xpath_expr, res, 1)) + xpath_expr, res, 1, attset_list)) { fclose (f); return 0; } *tp = all; /* append any ALL entries to the list */ } + cur_xpelement = &(*cur_xpelement)->next; } else if (!strcmp(cmd, "section")) { @@ -780,7 +946,7 @@ data1_absyn *data1_read_absyn (data1_handle dh, const char *file, if (argc < 2) { - yaz_log(LOG_WARN, "%s:%d: Bad # of args to section", + yaz_log(YLOG_WARN, "%s:%d: Bad # of args to section", file, lineno); continue; } @@ -800,17 +966,17 @@ data1_absyn *data1_read_absyn (data1_handle dh, const char *file, { if (argc != 2) { - yaz_log(LOG_WARN, "%s:%d: Bad # of args to 'xpath' directive", + yaz_log(YLOG_WARN, "%s:%d: Bad # of args to 'xpath' directive", file, lineno); continue; } if (!strcmp(argv[1], "enable")) - res->enable_xpath_indexing = 1; + res->xpath_indexing = DATA1_XPATH_INDEXING_ENABLE; else if (!strcmp (argv[1], "disable")) - res->enable_xpath_indexing = 0; + res->xpath_indexing = DATA1_XPATH_INDEXING_DISABLE; else { - yaz_log(LOG_WARN, "%s:%d: Expecting disable/enable " + yaz_log(YLOG_WARN, "%s:%d: Expecting disable/enable " "after 'xpath' directive", file, lineno); } } @@ -819,17 +985,18 @@ data1_absyn *data1_read_absyn (data1_handle dh, const char *file, data1_termlist **tp = &all; if (all) { - yaz_log(LOG_WARN, "%s:%d: Too many 'all' directives - ignored", + yaz_log(YLOG_WARN, "%s:%d: Too many 'all' directives - ignored", file, lineno); continue; } if (argc != 2) { - yaz_log(LOG_WARN, "%s:%d: Bad # of args to 'all' directive", + yaz_log(YLOG_WARN, "%s:%d: Bad # of args to 'all' directive", file, lineno); continue; } - if (parse_termlists (dh, &tp, argv[1], file, lineno, 0, res, 0)) + if (parse_termlists (dh, &tp, argv[1], file, lineno, 0, res, 0, + attset_list)) { fclose (f); return 0; @@ -839,7 +1006,7 @@ data1_absyn *data1_read_absyn (data1_handle dh, const char *file, { if (argc != 2) { - yaz_log(LOG_WARN, "%s:%d: Bad # of args to name directive", + yaz_log(YLOG_WARN, "%s:%d: Bad # of args to name directive", file, lineno); continue; } @@ -851,41 +1018,44 @@ data1_absyn *data1_read_absyn (data1_handle dh, const char *file, if (argc != 2) { - yaz_log(LOG_WARN, "%s:%d: Bad # of args to reference", + yaz_log(YLOG_WARN, "%s:%d: Bad # of args to reference", file, lineno); continue; } name = argv[1]; - if ((res->reference = oid_getvalbyname(name)) == VAL_NONE) + res->oid = yaz_string_to_oid_nmem(yaz_oid_std(), + CLASS_SCHEMA, name, + data1_nmem_get(dh)); + if (!res->oid) { - yaz_log(LOG_WARN, "%s:%d: Unknown tagset ref '%s'", + yaz_log(YLOG_WARN, "%s:%d: Unknown tagset ref '%s'", file, lineno, name); continue; } } else if (!strcmp(cmd, "attset")) { - char *name; - data1_attset *attset; - - if (argc != 2) - { - yaz_log(LOG_WARN, "%s:%d: Bad # of args to attset", - file, lineno); - continue; - } - name = argv[1]; - if (!(attset = data1_get_attset (dh, name))) - { - yaz_log(LOG_WARN, "%s:%d: Couldn't find attset %s", - file, lineno, name); - continue; - } - *attset_childp = (data1_attset_child *) - nmem_malloc (data1_nmem_get(dh), sizeof(**attset_childp)); - (*attset_childp)->child = attset; - (*attset_childp)->next = 0; - attset_childp = &(*attset_childp)->next; + char *name; + data1_attset *attset; + + if (argc != 2) + { + yaz_log(YLOG_WARN, "%s:%d: Bad # of args to attset", + file, lineno); + continue; + } + name = argv[1]; + if (!(attset = data1_get_attset (dh, name))) + { + yaz_log(YLOG_WARN, "%s:%d: Couldn't find attset %s", + file, lineno, name); + continue; + } + *attset_childp = (data1_attset_child *) + nmem_malloc (data1_nmem_get(dh), sizeof(**attset_childp)); + (*attset_childp)->child = attset; + (*attset_childp)->next = 0; + attset_childp = &(*attset_childp)->next; } else if (!strcmp(cmd, "tagset")) { @@ -893,7 +1063,7 @@ data1_absyn *data1_read_absyn (data1_handle dh, const char *file, int type = 0; if (argc < 2) { - yaz_log(LOG_WARN, "%s:%d: Bad # of args to tagset", + yaz_log(YLOG_WARN, "%s:%d: Bad # of args to tagset", file, lineno); continue; } @@ -903,7 +1073,7 @@ data1_absyn *data1_read_absyn (data1_handle dh, const char *file, *tagset_childp = data1_read_tagset (dh, name, type); if (!(*tagset_childp)) { - yaz_log(LOG_WARN, "%s:%d: Couldn't load tagset %s", + yaz_log(YLOG_WARN, "%s:%d: Couldn't load tagset %s", file, lineno, name); continue; } @@ -915,14 +1085,14 @@ data1_absyn *data1_read_absyn (data1_handle dh, const char *file, if (argc != 2) { - yaz_log(LOG_WARN, "%s:%d: Bad # of args in varset", + yaz_log(YLOG_WARN, "%s:%d: Bad # of args in varset", file, lineno); continue; } name = argv[1]; if (!(res->varset = data1_read_varset (dh, name))) { - yaz_log(LOG_WARN, "%s:%d: Couldn't load Varset %s", + yaz_log(YLOG_WARN, "%s:%d: Couldn't load Varset %s", file, lineno, name); continue; } @@ -933,7 +1103,7 @@ data1_absyn *data1_read_absyn (data1_handle dh, const char *file, if (argc != 3) { - yaz_log(LOG_WARN, "%s:%d: Bad # of args in esetname", + yaz_log(YLOG_WARN, "%s:%d: Bad # of args in esetname", file, lineno); continue; } @@ -948,7 +1118,7 @@ data1_absyn *data1_read_absyn (data1_handle dh, const char *file, (*esetpp)->spec = 0; else if (!((*esetpp)->spec = data1_read_espec1 (dh, fname))) { - yaz_log(LOG_WARN, "%s:%d: Espec-1 read failed for %s", + yaz_log(YLOG_WARN, "%s:%d: Espec-1 read failed for %s", file, lineno, fname); continue; } @@ -960,14 +1130,14 @@ data1_absyn *data1_read_absyn (data1_handle dh, const char *file, if (argc != 2) { - yaz_log(LOG_WARN, "%s:%d: Bad # of args for maptab", + yaz_log(YLOG_WARN, "%s:%d: Bad # of args for maptab", file, lineno); continue; } name = argv[1]; if (!(*maptabp = data1_read_maptab (dh, name))) { - yaz_log(LOG_WARN, "%s:%d: Couldn't load maptab %s", + yaz_log(YLOG_WARN, "%s:%d: Couldn't load maptab %s", file, lineno, name); continue; } @@ -979,14 +1149,14 @@ data1_absyn *data1_read_absyn (data1_handle dh, const char *file, if (argc != 2) { - yaz_log(LOG_WARN, "%s:%d: Bad # or args for marc", + yaz_log(YLOG_WARN, "%s:%d: Bad # or args for marc", file, lineno); continue; } name = argv[1]; if (!(*marcp = data1_read_marctab (dh, name))) { - yaz_log(LOG_WARN, "%s:%d: Couldn't read marctab %s", + yaz_log(YLOG_WARN, "%s:%d: Couldn't read marctab %s", file, lineno, name); continue; } @@ -996,7 +1166,7 @@ data1_absyn *data1_read_absyn (data1_handle dh, const char *file, { if (argc != 2) { - yaz_log(LOG_WARN, "%s:%d: Bad # or args for encoding", + yaz_log(YLOG_WARN, "%s:%d: Bad # or args for encoding", file, lineno); continue; } @@ -1006,7 +1176,7 @@ data1_absyn *data1_read_absyn (data1_handle dh, const char *file, { if (argc != 3) { - yaz_log(LOG_WARN, "%s:%d: Bad # or args for systag", + yaz_log(YLOG_WARN, "%s:%d: Bad # or args for systag", file, lineno); continue; } @@ -1018,7 +1188,7 @@ data1_absyn *data1_read_absyn (data1_handle dh, const char *file, } else { - yaz_log(LOG_WARN, "%s:%d: Unknown directive '%s'", file, + yaz_log(YLOG_WARN, "%s:%d: Unknown directive '%s'", file, lineno, cmd); continue; } @@ -1034,6 +1204,15 @@ data1_absyn *data1_read_absyn (data1_handle dh, const char *file, fix_element_ref (dh, res, cur_elements->elements); } *systagsp = 0; - yaz_log (LOG_DEBUG, "%s: data1_read_absyn end", file); return res; } + +/* + * Local variables: + * c-basic-offset: 4 + * c-file-style: "Stroustrup" + * indent-tabs-mode: nil + * End: + * vim: shiftwidth=4 tabstop=8 expandtab + */ +