X-Git-Url: http://git.indexdata.com/?p=idzebra-moved-to-github.git;a=blobdiff_plain;f=data1%2Fd1_absyn.c;h=5adb241359536656b8fbdd59e5754c29293c0bac;hp=1cee7f622b46d2fcd3b7c969ab9b58d6b8b6ea38;hb=396e9aaedfbed7534e329b42475cd7abe2fd3814;hpb=02266b68e2e2c86d8a8467ee56721926b365d2d0 diff --git a/data1/d1_absyn.c b/data1/d1_absyn.c index 1cee7f6..5adb241 100644 --- a/data1/d1_absyn.c +++ b/data1/d1_absyn.c @@ -1,6 +1,6 @@ -/* $Id: d1_absyn.c,v 1.2 2002-10-22 13:19:50 adam Exp $ - Copyright (C) 1995,1996,1997,1998,1999,2000,2001,2002 - Index Data Aps +/* $Id: d1_absyn.c,v 1.29 2006-09-28 18:38:44 adam Exp $ + Copyright (C) 1995-2006 + Index Data ApS This file is part of the Zebra server. @@ -15,9 +15,9 @@ 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 + */ #include @@ -25,12 +25,111 @@ Free Software Foundation, 59 Temple Place - Suite 330, Boston, MA #include #include -#include #include -#include +#include +#include +#include +#include +#include #define D1_MAX_NESTING 128 +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; + char *value; + struct data1_systag *next; +}; + struct data1_absyn_cache_info { char *name; @@ -45,18 +144,56 @@ 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); while (p) { - if (!strcmp (name, p->name)) + if (!yaz_matchstr (name, p->name)) return p->absyn; p = p->next; } - return NULL; + return 0; } +/* *ostrich* + We need to destroy DFAs, in xp_element (xelm) definitions + pop, 2002-12-13 +*/ + +void data1_absyn_destroy (data1_handle dh) +{ + data1_absyn_cache p = *data1_absyn_cache_get (dh); + + while (p) + { + data1_absyn *abs = p->absyn; + if (abs) + { + data1_xpelement *xpe = abs->xp_elements; + while (xpe) { + yaz_log (YLOG_DEBUG,"Destroy xp element %s",xpe->xpath_expr); + if (xpe->dfa) + dfa_delete (&xpe->dfa); + xpe = xpe->next; + } + } + p = p->next; + } +} + void data1_absyn_trav (data1_handle dh, void *handle, void (*fh)(data1_handle dh, void *h, data1_absyn *a)) @@ -70,7 +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); @@ -78,20 +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; } @@ -101,11 +243,11 @@ data1_attset *data1_attset_search_name (data1_handle dh, const char *name) while (p) { - if (!strcmp (name, p->name)) + if (!yaz_matchstr (name, p->name)) return p->attset; p = p->next; } - return NULL; + return 0; } data1_attset *data1_attset_search_id (data1_handle dh, int id) @@ -118,34 +260,24 @@ data1_attset *data1_attset_search_id (data1_handle dh, int id) 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; @@ -173,11 +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 ) @@ -188,6 +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; @@ -197,6 +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, @@ -235,45 +393,175 @@ 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); } } } +/* *ostrich* + + New function, a bit dummy now... I've seen it in zrpn.c... We should build + more clever regexps... + + + //a -> ^a/.*$ + //a/b -> ^b/a/.*$ + /a -> ^a/$ + /a/b -> ^b/a/$ + + / -> none + + pop, 2002-12-13 + + Now [] predicates are supported + pop, 2003-01-17 -static int parse_termlists (data1_handle dh, data1_termlist ***tpp, - char *p, const char *file, int lineno, - const char *element_name, data1_absyn *res) + */ + +static const char * mk_xpath_regexp (data1_handle dh, const char *expr) +{ + const char *p = expr; + int abs = 1; + int e = 0; + char *stack[32]; + char *res_p, *res = 0; + size_t res_size = 1; + + if (*p != '/') + return (""); + p++; + 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++) + { + 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++; + } + 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; + } + sprintf (res_p, "$"); + res_p++; + yaz_log(YLOG_DEBUG, "Got regexp: %s", res); + return res; +} + +static int parse_termlists(data1_handle dh, data1_termlist ***tpp, + char *cp, const char *file, int lineno, + const char *element_name, data1_absyn *res, + int xpelement, + data1_attset *attset) { data1_termlist **tp = *tpp; - do + while(1) { char attname[512], structure[512]; char *source; - int r; - - if (!(r = sscanf(p, "%511[^:,]:%511[^,]", attname, - structure))) + int r, i; + int level = 0; + structure[0] = '\0'; + for (i = 0; cp[i] && inext = 0; - if (!((*tp)->att = data1_getattbyname(dh, res->attset, - attname))) - { - yaz_log(LOG_WARN, - "%s:%d: Couldn't find att '%s' in attset", - file, lineno, attname); - return -1; - } + + if (*attname == '!') + { + if (!xpelement && element_name) + strcpy(attname, element_name); + else if (xpelement) + strcpy(attname, ZEBRA_XPATH_CDATA); + } + if (attset) + { + if (!data1_getattbyname(dh, attset, attname)) + { + yaz_log(YLOG_WARN, "Index '%s' not found in attset(s)", + attname); + } + } + + (*tp)->index_name = nmem_strdup(data1_nmem_get(dh), attname); + assert (*(*tp)->index_name != '!'); if (r == 2 && (source = strchr(structure, ':'))) *source++ = '\0'; /* cut off structure .. */ else @@ -288,15 +576,108 @@ static int parse_termlists (data1_handle dh, data1_termlist ***tpp, nmem_strdup (data1_nmem_get (dh), structure); tp = &(*tp)->next; } - while ((p = strchr(p, ',')) && *(++p)); + *tpp = tp; return 0; } -data1_absyn *data1_read_absyn (data1_handle dh, const char *file, - int file_must_exist) +/* quinn + * Converts a 'melm' field[$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) +{ + struct data1_systag *p = absyn->systags; + for (; p; p = p->next) + if (!strcmp(p->name, tag)) + return p->value; + return default_value; +} + +#define l_isspace(c) ((c) == '\t' || (c) == ' ' || (c) == '\n' || (c) == '\r') + +int read_absyn_line(FILE *f, int *lineno, char *line, int len, + char *argv[], int num) +{ + char *p; + int argc; + int quoted = 0; + + while ((p = fgets(line, len, f))) + { + (*lineno)++; + while (*p && l_isspace(*p)) + p++; + if (*p && *p != '#') + break; + } + if (!p) + return 0; + + for (argc = 0; *p ; argc++) + { + if (*p == '#') /* trailing comment */ + break; + argv[argc] = p; + while (*p && !(l_isspace(*p) && !quoted)) { + if (*p =='"') quoted = 1 - quoted; + if (*p =='[') quoted = 1; + if (*p ==']') quoted = 0; + p++; + } + if (*p) + { + *(p++) = '\0'; + while (*p && l_isspace(*p)) + p++; + } + } + return argc; +} + +data1_marctab *data1_absyn_getmarctab(data1_handle dh, data1_absyn *absyn) +{ + return absyn->marc; +} + +YAZ_EXPORT data1_element *data1_absyn_getelements(data1_handle dh, + data1_absyn *absyn) +{ + return absyn->main_elements; +} + +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_attset *attset_list = data1_empty_attset(dh); + data1_attset_child **attset_childp = &attset_list->children; + data1_absyn *res = 0; FILE *f; data1_element **ppl[D1_MAX_NESTING]; @@ -304,31 +685,26 @@ 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; int lineno = 0; 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->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; @@ -336,11 +712,11 @@ data1_absyn *data1_read_absyn (data1_handle dh, const char *file, maptabp = &res->maptabs; res->marc = 0; marcp = &res->marc; - res->sub_elements = NULL; res->main_elements = NULL; - - while (f && (argc = readconf_line(f, &lineno, line, 512, argv, 50))) + res->xp_elements = NULL; + + while (f && (argc = read_absyn_line(f, &lineno, line, 512, argv, 50))) { char *cmd = *argv; if (!strcmp(cmd, "elm") || !strcmp(cmd, "element")) @@ -353,7 +729,7 @@ 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]; @@ -384,17 +760,12 @@ 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; @@ -412,14 +783,14 @@ 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; @@ -445,7 +816,7 @@ 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; } @@ -453,9 +824,8 @@ 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)) + if (parse_termlists (dh, &tp, p, file, lineno, name, res, 0, + attset_list)) { fclose (f); return 0; @@ -464,14 +834,118 @@ data1_absyn *data1_read_absyn (data1_handle dh, const char *file, } new_element->name = nmem_strdup(data1_nmem_get (dh), name); } + /* *ostrich* + New code to support xelm directive + for each xelm a dfa is built. xelms are stored in res->xp_elements + + maybe we should use a simple sscanf instead of dfa? + + pop, 2002-12-13 + + Now [] predicates are supported. regexps and xpath structure is + a bit redundant, however it's comfortable later... + + pop, 2003-01-17 + */ + + else if (!strcmp(cmd, "xelm") || !strcmp(cmd, "melm")) { + + int i; + char *p, *xpath_expr, *termlists; + const char *regexp; + struct DFA *dfa = 0; + data1_termlist **tp; + char melm_xpath[128]; + data1_xpelement *xp_old = 0; + + if (argc < 3) + { + yaz_log(YLOG_WARN, "%s:%d: Bad # of args to xelm", file, lineno); + continue; + } + + 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); + +#if OPTIMIZE_MELM + for (xp_old = res->xp_elements; xp_old; xp_old = xp_old->next) + if (!strcmp(xp_old->regexp, regexp)) + break; +#endif + if (!xp_old) + { + 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; + } + } + 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; + } +#if OPTIMIZE_MELM + cur_xpelement->regexp = regexp; +#endif + cur_xpelement->next = NULL; + cur_xpelement->xpath_expr = nmem_strdup(data1_nmem_get (dh), + xpath_expr); + + if (dfa) + 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)); + + /* + dump_xp_steps(cur_xpelement->xpath,cur_xpelement->xpath_len); + */ +#endif + cur_xpelement->termlists = 0; + tp = &cur_xpelement->termlists; + + /* parse termList definitions */ + p = termlists; + if (*p != '-') + { + if (parse_termlists (dh, &tp, p, file, lineno, + xpath_expr, res, 1, attset_list)) + { + fclose (f); + return 0; + } + *tp = all; /* append any ALL entries to the list */ + } + } else if (!strcmp(cmd, "section")) { char *name; if (argc < 2) { - yaz_log(LOG_WARN, "%s:%d: Bad # of args to section", - file, lineno); + yaz_log(YLOG_WARN, "%s:%d: Bad # of args to section", + file, lineno); continue; } name = argv[1]; @@ -490,17 +964,17 @@ 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); } } @@ -509,17 +983,18 @@ 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)) + if (parse_termlists (dh, &tp, argv[1], file, lineno, 0, res, 0, + attset_list)) { fclose (f); return 0; @@ -529,7 +1004,7 @@ 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; } @@ -541,41 +1016,41 @@ 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) { - 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")) { @@ -583,7 +1058,7 @@ 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; } @@ -593,7 +1068,7 @@ 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; } @@ -605,14 +1080,14 @@ 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; } @@ -623,7 +1098,7 @@ 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; } @@ -638,7 +1113,7 @@ 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; } @@ -650,14 +1125,14 @@ 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; } @@ -669,14 +1144,14 @@ 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; } @@ -686,15 +1161,29 @@ 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; } res->encoding = nmem_strdup (data1_nmem_get(dh), argv[1]); } + else if (!strcmp(cmd, "systag")) + { + if (argc != 3) + { + yaz_log(YLOG_WARN, "%s:%d: Bad # or args for systag", + file, lineno); + continue; + } + *systagsp = nmem_malloc (data1_nmem_get(dh), sizeof(**systagsp)); + + (*systagsp)->name = nmem_strdup(data1_nmem_get(dh), argv[1]); + (*systagsp)->value = nmem_strdup(data1_nmem_get(dh), argv[2]); + systagsp = &(*systagsp)->next; + } else { - yaz_log(LOG_WARN, "%s:%d: Unknown directive '%s'", file, + yaz_log(YLOG_WARN, "%s:%d: Unknown directive '%s'", file, lineno, cmd); continue; } @@ -709,6 +1198,14 @@ data1_absyn *data1_read_absyn (data1_handle dh, const char *file, res->main_elements = cur_elements->elements; fix_element_ref (dh, res, cur_elements->elements); } - yaz_log (LOG_DEBUG, "%s: data1_read_absyn end", file); + *systagsp = 0; return res; } +/* + * Local variables: + * c-basic-offset: 4 + * indent-tabs-mode: nil + * End: + * vim: shiftwidth=4 tabstop=8 expandtab + */ +