X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=recctrl%2Frecgrs.c;h=d937600b4170457e5d7779d33916b7642afb97b5;hb=7ac37393aeb242f6eddf1a604923bd009baace41;hp=0b7feea2a0cc2dd4504c885d266164481091b3fb;hpb=12cdf00008c22bb5edf92632c7236b2676a273dd;p=idzebra-moved-to-github.git diff --git a/recctrl/recgrs.c b/recctrl/recgrs.c index 0b7feea..d937600 100644 --- a/recctrl/recgrs.c +++ b/recctrl/recgrs.c @@ -1,10 +1,37 @@ /* - * Copyright (C) 1994-1998, Index Data I/S + * Copyright (C) 1994-1999, Index Data * All rights reserved. * Sebastian Hammer, Adam Dickmeiss * * $Log: recgrs.c,v $ - * Revision 1.21 1998-07-01 09:16:10 adam + * Revision 1.30 1999-07-06 12:26:41 adam + * Retrieval handler obeys schema and handles XML transfer syntax. + * + * Revision 1.29 1999/05/26 07:49:14 adam + * C++ compilation. + * + * Revision 1.28 1999/05/21 12:00:17 adam + * Better diagnostics for extraction process. + * + * Revision 1.27 1999/05/20 12:57:18 adam + * Implemented TCL filter. Updated recctrl system. + * + * Revision 1.26 1999/03/02 16:15:44 quinn + * Added "tagsysno" and "tagrank" directives to zebra.cfg. + * + * Revision 1.25 1999/02/18 15:01:26 adam + * Minor changes. + * + * Revision 1.24 1999/02/02 14:51:28 adam + * Updated WIN32 code specific sections. Changed header. + * + * Revision 1.23 1998/10/18 07:51:10 adam + * Changed one logf call. + * + * Revision 1.22 1998/10/16 08:14:37 adam + * Updated record control system. + * + * Revision 1.21 1998/07/01 09:16:10 adam * Element localno only added when it's greater than 0. * * Revision 1.20 1998/05/20 10:12:26 adam @@ -160,7 +187,7 @@ #include #include #include -#ifndef WINDOWS +#ifndef WIN32 #include #endif @@ -172,19 +199,23 @@ #define GRS_MAX_WORD 512 -static data1_node *read_grs_type (struct grs_read_info *p, const char *type) +struct grs_handler { + RecTypeGrs type; + void *clientData; + int initFlag; + struct grs_handler *next; +}; + +struct grs_handlers { + struct grs_handler *handlers; +}; + +static int read_grs_type (struct grs_handlers *h, + struct grs_read_info *p, const char *type, + data1_node **root) { - static struct { - char *type; - data1_node *(*func)(struct grs_read_info *p); - } tab[] = { - { "sgml", grs_read_sgml }, - { "regx", grs_read_regx }, - { "marc", grs_read_marc }, - { NULL, NULL } - }; + struct grs_handler *gh = h->handlers; const char *cp = strchr (type, '.'); - int i; if (cp == NULL || cp == type) { @@ -193,16 +224,61 @@ static data1_node *read_grs_type (struct grs_read_info *p, const char *type) } else strcpy (p->type, cp+1); - for (i=0; tab[i].type; i++) + for (gh = h->handlers; gh; gh = gh->next) { - if (!memcmp (type, tab[i].type, cp-type)) - return (tab[i].func)(p); + if (!memcmp (type, gh->type->type, cp-type)) + { + if (!gh->initFlag) + { + gh->initFlag = 1; + gh->clientData = (*gh->type->init)(); + } + p->clientData = gh->clientData; + *root = (gh->type->read)(p); + gh->clientData = p->clientData; + return 0; + } } - return NULL; + return 1; +} + +static void grs_add_handler (struct grs_handlers *h, RecTypeGrs t) +{ + struct grs_handler *gh = (struct grs_handler *) malloc (sizeof(*gh)); + gh->next = h->handlers; + h->handlers = gh; + gh->initFlag = 0; + gh->clientData = 0; + gh->type = t; +} + +static void *grs_init(RecType recType) +{ + struct grs_handlers *h = (struct grs_handlers *) malloc (sizeof(*h)); + h->handlers = 0; + + grs_add_handler (h, recTypeGrs_sgml); + grs_add_handler (h, recTypeGrs_regx); +#if HAVE_TCL_H + grs_add_handler (h, recTypeGrs_tcl); +#endif + grs_add_handler (h, recTypeGrs_marc); + return h; } -static void grs_init(void) +static void grs_destroy(void *clientData) { + struct grs_handlers *h = (struct grs_handlers *) clientData; + struct grs_handler *gh = h->handlers, *gh_next; + while (gh) + { + gh_next = gh->next; + if (gh->initFlag) + (*gh->type->destroy)(gh->clientData); + free (gh); + gh = gh_next; + } + free (h); } static int dumpkeys(data1_node *n, struct recExtractCtrl *p, int level) @@ -330,13 +406,14 @@ int grs_extract_tree(struct recExtractCtrl *p, data1_node *n) return dumpkeys(n, p, 0); } -static int grs_extract(struct recExtractCtrl *p) +static int grs_extract(void *clientData, struct recExtractCtrl *p) { data1_node *n; NMEM mem; struct grs_read_info gri; oident oe; int oidtmp[OID_SIZE]; + struct grs_handlers *h = (struct grs_handlers *) clientData; mem = nmem_create (); gri.readf = p->readf; @@ -348,9 +425,10 @@ static int grs_extract(struct recExtractCtrl *p) gri.mem = mem; gri.dh = p->dh; - n = read_grs_type (&gri, p->subType); + if (read_grs_type (h, &gri, p->subType, &n)) + return RECCTRL_EXTRACT_ERROR; if (!n) - return -1; + return RECCTRL_EXTRACT_EOF; oe.proto = PROTO_Z3950; oe.oclass = CLASS_SCHEMA; @@ -361,11 +439,11 @@ static int grs_extract(struct recExtractCtrl *p) if (dumpkeys(n, p, 0) < 0) { data1_free_tree(p->dh, n); - return -2; + return RECCTRL_EXTRACT_ERROR; } data1_free_tree(p->dh, n); nmem_destroy(mem); - return 0; + return RECCTRL_EXTRACT_OK; } /* @@ -379,60 +457,61 @@ static int process_comp(data1_handle dh, data1_node *n, Z_RecordComposition *c) switch (c->which) { - case Z_RecordComp_simple: - if (c->u.simple->which != Z_ElementSetNames_generic) - return 26; /* only generic form supported. Fix this later */ - if (!(eset = data1_getesetbyname(dh, n->u.root.absyn, - c->u.simple->u.generic))) - { - logf(LOG_LOG, "Unknown esetname '%s'", c->u.simple->u.generic); + case Z_RecordComp_simple: + if (c->u.simple->which != Z_ElementSetNames_generic) + return 26; /* only generic form supported. Fix this later */ + if (!(eset = data1_getesetbyname(dh, n->u.root.absyn, + c->u.simple->u.generic))) + { + logf(LOG_LOG, "Unknown esetname '%s'", c->u.simple->u.generic); return 25; /* invalid esetname */ - } - logf(LOG_DEBUG, "Esetname '%s' in simple compspec", - c->u.simple->u.generic); - espec = eset->spec; - break; - case Z_RecordComp_complex: - if (c->u.complex->generic) + } + logf(LOG_DEBUG, "Esetname '%s' in simple compspec", + c->u.simple->u.generic); + espec = eset->spec; + break; + case Z_RecordComp_complex: + if (c->u.complex->generic) + { + /* insert check for schema */ + if ((p = c->u.complex->generic->elementSpec)) { - /* insert check for schema */ - if ((p = c->u.complex->generic->elementSpec)) - switch (p->which) + switch (p->which) + { + case Z_ElementSpec_elementSetName: + if (!(eset = + data1_getesetbyname(dh, n->u.root.absyn, + p->u.elementSetName))) + { + logf(LOG_LOG, "Unknown esetname '%s'", + p->u.elementSetName); + return 25; /* invalid esetname */ + } + logf(LOG_DEBUG, "Esetname '%s' in complex compspec", + p->u.elementSetName); + espec = eset->spec; + break; + case Z_ElementSpec_externalSpec: + if (p->u.externalSpec->which == Z_External_espec1) + { + logf(LOG_DEBUG, "Got Espec-1"); + espec = p->u.externalSpec-> u.espec1; + } + else { - case Z_ElementSpec_elementSetName: - if (!(eset = - data1_getesetbyname(dh, - n->u.root.absyn, - p->u.elementSetName))) - { - logf(LOG_LOG, "Unknown esetname '%s'", - p->u.elementSetName); - return 25; /* invalid esetname */ - } - logf(LOG_DEBUG, "Esetname '%s' in complex compspec", - p->u.elementSetName); - espec = eset->spec; - break; - case Z_ElementSpec_externalSpec: - if (p->u.externalSpec->which == Z_External_espec1) - { - logf(LOG_DEBUG, "Got Espec-1"); - espec = p->u.externalSpec-> u.espec1; - } - else - { - logf(LOG_LOG, "Unknown external espec."); - return 25; /* bad. what is proper diagnostic? */ - } - break; + logf(LOG_LOG, "Unknown external espec."); + return 25; /* bad. what is proper diagnostic? */ } + break; + } } - else - return 26; /* fix */ + } + else + return 26; /* fix */ } if (espec) { - logf (LOG_LOG, "Element: Espec-1 match"); + logf (LOG_DEBUG, "Element: Espec-1 match"); return data1_doespec1(dh, n, espec); } else @@ -442,7 +521,7 @@ static int process_comp(data1_handle dh, data1_node *n, Z_RecordComposition *c) } } -static int grs_retrieve(struct recRetrieveCtrl *p) +static int grs_retrieve(void *clientData, struct recRetrieveCtrl *p) { data1_node *node = 0, *onode = 0; data1_node *dnew; @@ -450,6 +529,9 @@ static int grs_retrieve(struct recRetrieveCtrl *p) int res, selected = 0; NMEM mem; struct grs_read_info gri; + char *tagname; + struct grs_handlers *h = (struct grs_handlers *) clientData; + int requested_schema = VAL_NONE; mem = nmem_create(); gri.readf = p->readf; @@ -462,7 +544,12 @@ static int grs_retrieve(struct recRetrieveCtrl *p) gri.dh = p->dh; logf (LOG_DEBUG, "grs_retrieve"); - node = read_grs_type (&gri, p->subType); + if (read_grs_type (h, &gri, p->subType, &node)) + { + p->diagnostic = 14; + nmem_destroy (mem); + return 0; + } if (!node) { p->diagnostic = 14; @@ -479,88 +566,125 @@ static int grs_retrieve(struct recRetrieveCtrl *p) dnew->u.data.len = strlen(dnew->u.data.data); } - logf (LOG_DEBUG, "grs_retrieve: score"); - if (p->score >= 0 && (dnew = - data1_insert_taggeddata(p->dh, node, - node, "rank", - mem))) + tagname = res_get_def(p->res, "tagrank", "rank"); + if (strcmp(tagname, "0") && p->score >= 0 && + (dnew = data1_insert_taggeddata(p->dh, node, node, tagname, mem))) { + logf (LOG_DEBUG, "grs_retrieve: %s", tagname); dnew->u.data.what = DATA1I_num; dnew->u.data.data = dnew->lbuf; sprintf(dnew->u.data.data, "%d", p->score); dnew->u.data.len = strlen(dnew->u.data.data); } - logf (LOG_DEBUG, "grs_retrieve: localControlNumber"); - if (p->localno > 0 && (dnew = data1_insert_taggeddata(p->dh, node, node, - "localControlNumber", mem))) + tagname = res_get_def(p->res, "tagsysno", "localControlNumber"); + if (strcmp(tagname, "0") && p->localno > 0 && + (dnew = data1_insert_taggeddata(p->dh, node, node, tagname, mem))) { + logf (LOG_DEBUG, "grs_retrieve: %s", tagname); dnew->u.data.what = DATA1I_text; dnew->u.data.data = dnew->lbuf; sprintf(dnew->u.data.data, "%d", p->localno); dnew->u.data.len = strlen(dnew->u.data.data); } + if (p->comp && p->comp->which == Z_RecordComp_complex && + p->comp->u.complex->generic && + p->comp->u.complex->generic->schema) + { + oident *oe = oid_getentbyoid (p->comp->u.complex->generic->schema); + if (oe) + requested_schema = oe->value; + } + + /* If schema has been specified, map if possible, then check that + * we got the right one + */ + if (requested_schema != VAL_NONE) + { + logf (LOG_DEBUG, "grs_retrieve: schema mapping"); + for (map = node->u.root.absyn->maptabs; map; map = map->next) + { + if (map->target_absyn_ref == requested_schema) + { + onode = node; + if (!(node = data1_map_record(p->dh, onode, map, mem))) + { + p->diagnostic = 14; + nmem_destroy (mem); + return 0; + } + break; + } + } + if (node->u.root.absyn && + requested_schema != node->u.root.absyn->reference) + { + p->diagnostic = 238; + nmem_destroy (mem); + return 0; + } + } + /* + * Does the requested format match a known syntax-mapping? (this reflects + * the overlap of schema and formatting which is inherent in the MARC + * family) + */ + logf (LOG_DEBUG, "grs_retrieve: syntax mapping"); + for (map = node->u.root.absyn->maptabs; map; map = map->next) + { + if (map->target_absyn_ref == p->input_format) + { + onode = node; + if (!(node = data1_map_record(p->dh, onode, map, mem))) + { + p->diagnostic = 14; + nmem_destroy (mem); + return 0; + } + break; + } + } logf (LOG_DEBUG, "grs_retrieve: schemaIdentifier"); - if (p->input_format == VAL_GRS1 && node->u.root.absyn && - node->u.root.absyn->reference != VAL_NONE) + if (node->u.root.absyn && + node->u.root.absyn->reference != VAL_NONE && + p->input_format == VAL_GRS1) { oident oe; Odr_oid *oid; int oidtmp[OID_SIZE]; - + oe.proto = PROTO_Z3950; oe.oclass = CLASS_SCHEMA; oe.value = node->u.root.absyn->reference; - + if ((oid = oid_ent_to_oid (&oe, oidtmp))) { char tmp[128]; data1_handle dh = p->dh; char *p = tmp; int *ii; - + for (ii = oid; *ii >= 0; ii++) { if (p != tmp) - *(p++) = '.'; + *(p++) = '.'; sprintf(p, "%d", *ii); p += strlen(p); } *(p++) = '\0'; - + if ((dnew = data1_insert_taggeddata(dh, node, node, - "schemaIdentifier", mem))) + "schemaIdentifier", mem))) { dnew->u.data.what = DATA1I_oid; - dnew->u.data.data = nmem_malloc(mem, p - tmp); + dnew->u.data.data = (char *) nmem_malloc(mem, p - tmp); memcpy(dnew->u.data.data, tmp, p - tmp); dnew->u.data.len = p - tmp; } } } - logf (LOG_DEBUG, "grs_retrieve: schema mapping"); - /* - * Does the requested format match a known schema-mapping? (this reflects - * the overlap of schema and formatting which is inherent in the MARC - * family) - * NOTE: This should look at the schema-specification in the compspec - * as well. - */ - for (map = node->u.root.absyn->maptabs; map; map = map->next) - if (map->target_absyn_ref == p->input_format) - { - onode = node; - if (!(node = data1_map_record(p->dh, onode, map, mem))) - { - p->diagnostic = 14; - nmem_destroy (mem); - return 0; - } - break; - } - logf (LOG_DEBUG, "grs_retrieve: element spec"); if (p->comp && (res = process_comp(p->dh, node, p->comp)) > 0) { @@ -576,65 +700,91 @@ static int grs_retrieve(struct recRetrieveCtrl *p) logf (LOG_DEBUG, "grs_retrieve: transfer syntax mapping"); switch (p->output_format = (p->input_format != VAL_NONE ? - p->input_format : VAL_SUTRS)) + p->input_format : VAL_SUTRS)) { data1_marctab *marctab; int dummy; - - case VAL_GRS1: - dummy = 0; - if (!(p->rec_buf = data1_nodetogr(p->dh, node, selected, - p->odr, &dummy))) - p->diagnostic = 238; /* not available in requested syntax */ - else - p->rec_len = -1; - break; - case VAL_EXPLAIN: - if (!(p->rec_buf = data1_nodetoexplain(p->dh, node, selected, - p->odr))) - p->diagnostic = 238; - else - p->rec_len = -1; - break; - case VAL_SUMMARY: - if (!(p->rec_buf = data1_nodetosummary(p->dh, node, selected, - p->odr))) - p->diagnostic = 238; - else - p->rec_len = -1; - break; - case VAL_SUTRS: - if (!(p->rec_buf = data1_nodetobuf(p->dh, node, selected, - (int*)&p->rec_len))) - p->diagnostic = 238; - break; - case VAL_SOIF: - if (!(p->rec_buf = data1_nodetosoif(p->dh, node, selected, - (int*)&p->rec_len))) - p->diagnostic = 238; + + case VAL_TEXT_XML: + if (!(p->rec_buf = data1_nodetoidsgml(p->dh, node, selected, + (int*)&p->rec_len))) + p->diagnostic = 238; + else + { + char *new_buf = (char*) odr_malloc (p->odr, p->rec_len); + memcpy (new_buf, p->rec_buf, p->rec_len); + p->rec_buf = new_buf; + } + break; + case VAL_GRS1: + dummy = 0; + if (!(p->rec_buf = data1_nodetogr(p->dh, node, selected, + p->odr, &dummy))) + p->diagnostic = 238; /* not available in requested syntax */ + else + p->rec_len = (size_t) (-1); + break; + case VAL_EXPLAIN: + if (!(p->rec_buf = data1_nodetoexplain(p->dh, node, selected, + p->odr))) + p->diagnostic = 238; + else + p->rec_len = (size_t) (-1); + break; + case VAL_SUMMARY: + if (!(p->rec_buf = data1_nodetosummary(p->dh, node, selected, + p->odr))) + p->diagnostic = 238; + else + p->rec_len = (size_t) (-1); + break; + case VAL_SUTRS: + if (!(p->rec_buf = data1_nodetobuf(p->dh, node, selected, + (int*)&p->rec_len))) + p->diagnostic = 238; + else + { + char *new_buf = (char*) odr_malloc (p->odr, p->rec_len); + memcpy (new_buf, p->rec_buf, p->rec_len); + p->rec_buf = new_buf; + } + break; + case VAL_SOIF: + if (!(p->rec_buf = data1_nodetosoif(p->dh, node, selected, + (int*)&p->rec_len))) + p->diagnostic = 238; + else + { + char *new_buf = (char*) odr_malloc (p->odr, p->rec_len); + memcpy (new_buf, p->rec_buf, p->rec_len); + p->rec_buf = new_buf; + } + break; + default: + if (!node->u.root.absyn) + { + p->diagnostic = 238; break; - default: - if (!node->u.root.absyn) - { - p->diagnostic = 238; - break; - } - for (marctab = node->u.root.absyn->marc; marctab; - marctab = marctab->next) - if (marctab->reference == p->input_format) - break; - if (!marctab) - { - p->diagnostic = 238; + } + for (marctab = node->u.root.absyn->marc; marctab; + marctab = marctab->next) + if (marctab->reference == p->input_format) break; - } - if (!(p->rec_buf = data1_nodetomarc(p->dh, marctab, node, + if (!marctab) + { + p->diagnostic = 238; + break; + } + if (!(p->rec_buf = data1_nodetomarc(p->dh, marctab, node, selected, - (int*)&p->rec_len))) - { - p->diagnostic = 238; - break; - } + (int*)&p->rec_len))) + p->diagnostic = 238; + else + { + char *new_buf = (char*) odr_malloc (p->odr, p->rec_len); + memcpy (new_buf, p->rec_buf, p->rec_len); + p->rec_buf = new_buf; + } } if (node) data1_free_tree(p->dh, node); @@ -648,6 +798,7 @@ static struct recType grs_type = { "grs", grs_init, + grs_destroy, grs_extract, grs_retrieve };