X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=recctrl%2Fxslt.c;h=9fe383a161c7fbb576c21bb0929973a28e71334b;hb=b912fcbe609de53861c5cf9b5196986dd97a1462;hp=421a307330480c94ebce38543b1c52ccd08d8e52;hpb=fd5356062a1e7ed5d750b8ffc97d0052c1c6af85;p=idzebra-moved-to-github.git diff --git a/recctrl/xslt.c b/recctrl/xslt.c index 421a307..9fe383a 100644 --- a/recctrl/xslt.c +++ b/recctrl/xslt.c @@ -1,4 +1,4 @@ -/* $Id: xslt.c,v 1.2 2005-04-28 12:34:21 adam Exp $ +/* $Id: xslt.c,v 1.7 2005-06-01 07:32:46 adam Exp $ Copyright (C) 1995-2005 Index Data ApS @@ -25,18 +25,32 @@ Free Software Foundation, 59 Temple Place - Suite 330, Boston, MA #include #include +#include +#include +#include #include #include #include #include -struct filter_info { +struct filter_schema { + const char *name; + const char *identifier; + const char *stylesheet; + struct filter_schema *next; + const char *default_schema; xsltStylesheetPtr stylesheet_xsp; - xmlTextReaderPtr reader; +}; + +struct filter_info { + xmlDocPtr doc; char *fname; - int split_depth; + const char *split_level; + const char *split_path; ODR odr; + struct filter_schema *schemas; + xmlTextReaderPtr reader; }; #define ZEBRA_INDEX_NS "http://indexdata.dk/zebra/indexing/1" @@ -59,6 +73,8 @@ static void set_param_int(const char **params, const char *name, zint value, ODR odr) { char *quoted = odr_malloc(odr, 30); /* 25 digits enough for 2^64 */ + while (*params) + params++; sprintf(quoted, "'" ZINT_FORMAT "'", value); params[0] = name; params[1] = quoted; @@ -66,40 +82,149 @@ static void set_param_int(const char **params, const char *name, } -static void *filter_init (Res res, RecType recType) +static void *filter_init_xslt(Res res, RecType recType) { struct filter_info *tinfo = (struct filter_info *) xmalloc(sizeof(*tinfo)); - tinfo->stylesheet_xsp = 0; tinfo->reader = 0; tinfo->fname = 0; - tinfo->split_depth = 1; + tinfo->split_level = 0; + tinfo->split_path = 0; tinfo->odr = odr_createmem(ODR_ENCODE); + tinfo->doc = 0; + tinfo->schemas = 0; + return tinfo; +} + +static void *filter_init_xslt1(Res res, RecType recType) +{ + struct filter_info *tinfo = (struct filter_info *) + filter_init_xslt(res, recType); + tinfo->split_level = "1"; return tinfo; } +static int attr_content(struct _xmlAttr *attr, const char *name, + const char **dst_content) +{ + if (!strcmp(attr->name, name) && attr->children && + attr->children->type == XML_TEXT_NODE) + { + *dst_content = attr->children->content; + return 1; + } + return 0; +} + +static void destroy_schemas(struct filter_info *tinfo) +{ + struct filter_schema *schema = tinfo->schemas; + while (schema) + { + struct filter_schema *schema_next = schema->next; + if (schema->stylesheet_xsp) + xsltFreeStylesheet(schema->stylesheet_xsp); + xfree(schema); + schema = schema_next; + } + tinfo->schemas = 0; + xfree(tinfo->fname); + if (tinfo->doc) + xmlFreeDoc(tinfo->doc); + tinfo->doc = 0; +} + +static ZEBRA_RES create_schemas(struct filter_info *tinfo, const char *fname) +{ + xmlNodePtr ptr; + tinfo->fname = xstrdup(fname); + tinfo->doc = xmlParseFile(tinfo->fname); + if (!tinfo->doc) + return ZEBRA_FAIL; + ptr = xmlDocGetRootElement(tinfo->doc); + if (!ptr || ptr->type != XML_ELEMENT_NODE || + strcmp(ptr->name, "schemaInfo")) + return ZEBRA_FAIL; + for (ptr = ptr->children; ptr; ptr = ptr->next) + { + if (ptr->type != XML_ELEMENT_NODE) + continue; + if (!strcmp(ptr->name, "schema")) + { + struct _xmlAttr *attr; + struct filter_schema *schema = xmalloc(sizeof(*schema)); + schema->name = 0; + schema->identifier = 0; + schema->stylesheet = 0; + schema->default_schema = 0; + schema->next = tinfo->schemas; + schema->stylesheet_xsp = 0; + tinfo->schemas = schema; + for (attr = ptr->properties; attr; attr = attr->next) + { + attr_content(attr, "identifier", &schema->identifier); + attr_content(attr, "name", &schema->name); + attr_content(attr, "stylesheet", &schema->stylesheet); + attr_content(attr, "default", &schema->default_schema); + } + if (schema->stylesheet) + schema->stylesheet_xsp = + xsltParseStylesheetFile( + (const xmlChar*) schema->stylesheet); + } + else if (!strcmp(ptr->name, "split")) + { + struct _xmlAttr *attr; + for (attr = ptr->properties; attr; attr = attr->next) + { + attr_content(attr, "level", &tinfo->split_level); + attr_content(attr, "path", &tinfo->split_path); + } + } + else + { + yaz_log(YLOG_WARN, "Bad element %s in %s", ptr->name, fname); + return ZEBRA_FAIL; + } + } + return ZEBRA_OK; +} + +static struct filter_schema *lookup_schema(struct filter_info *tinfo, + const char *est) +{ + struct filter_schema *schema; + for (schema = tinfo->schemas; schema; schema = schema->next) + { + if (est) + { + if (schema->identifier && !strcmp(schema->identifier, est)) + return schema; + if (schema->name && !strcmp(schema->name, est)) + return schema; + } + if (schema->default_schema) + return schema; + } + return 0; +} + static void filter_config(void *clientData, Res res, const char *args) { struct filter_info *tinfo = clientData; if (!args || !*args) - args = "default.xsl"; - if (!tinfo->fname || strcmp(args, tinfo->fname)) - { - /* different filename so must reread stylesheet */ - xfree(tinfo->fname); - tinfo->fname = xstrdup(args); - if (tinfo->stylesheet_xsp) - xsltFreeStylesheet(tinfo->stylesheet_xsp); - tinfo->stylesheet_xsp = - xsltParseStylesheetFile((const xmlChar*) tinfo->fname); - } + args = "xsltfilter.xml"; + if (tinfo->fname && !strcmp(args, tinfo->fname)) + return; + destroy_schemas(tinfo); + create_schemas(tinfo, args); } static void filter_destroy(void *clientData) { struct filter_info *tinfo = clientData; - if (tinfo->stylesheet_xsp) - xsltFreeStylesheet(tinfo->stylesheet_xsp); - xfree(tinfo->fname); + destroy_schemas(tinfo); + if (tinfo->reader) + xmlFreeTextReader(tinfo->reader); odr_destroy(tinfo->odr); xfree(tinfo); } @@ -161,18 +286,50 @@ static void index_node(struct filter_info *tinfo, struct recExtractCtrl *ctrl, } } -static int filter_extract(void *clientData, struct recExtractCtrl *p) +static int extract_doc(struct filter_info *tinfo, struct recExtractCtrl *p, + xmlDocPtr doc) { - const char *params[10]; - struct filter_info *tinfo = clientData; RecWord recWord; - int ret; + const char *params[10]; + xmlChar *buf_out; + int len_out; - params[0] = 0; + struct filter_schema *schema = lookup_schema(tinfo, ZEBRA_INDEX_NS); - odr_reset(tinfo->odr); + params[0] = 0; set_param_str(params, "schema", ZEBRA_INDEX_NS, tinfo->odr); + (*p->init)(p, &recWord); + recWord.reg_type = 'w'; + + if (schema && schema->stylesheet_xsp) + { + xmlDocPtr resDoc = + xsltApplyStylesheet(schema->stylesheet_xsp, + doc, params); + if (p->flagShowRecords) + { + xmlDocDumpMemory(resDoc, &buf_out, &len_out); + fwrite(buf_out, len_out, 1, stdout); + xmlFree(buf_out); + } + index_node(tinfo, p, xmlDocGetRootElement(resDoc), &recWord); + xmlFreeDoc(resDoc); + } + xmlDocDumpMemory(doc, &buf_out, &len_out); + if (p->flagShowRecords) + fwrite(buf_out, len_out, 1, stdout); + (*p->setStoreData)(p, buf_out, len_out); + xmlFree(buf_out); + + xmlFreeDoc(doc); + return RECCTRL_EXTRACT_OK; +} + +static int extract_split(struct filter_info *tinfo, struct recExtractCtrl *p) +{ + int ret; + int split_depth = 0; if (p->first_record) { if (tinfo->reader) @@ -186,50 +343,23 @@ static int filter_extract(void *clientData, struct recExtractCtrl *p) if (!tinfo->reader) return RECCTRL_EXTRACT_ERROR_GENERIC; - if (!tinfo->stylesheet_xsp) - return RECCTRL_EXTRACT_ERROR_GENERIC; - - (*p->init)(p, &recWord); - recWord.reg_type = 'w'; - + if (tinfo->split_level) + split_depth = atoi(tinfo->split_level); ret = xmlTextReaderRead(tinfo->reader); while (ret == 1) { int type = xmlTextReaderNodeType(tinfo->reader); int depth = xmlTextReaderDepth(tinfo->reader); - if (tinfo->split_depth == 0 || - (type == XML_READER_TYPE_ELEMENT && tinfo->split_depth == depth)) + if (split_depth == 0 || + (split_depth > 0 && + type == XML_READER_TYPE_ELEMENT && split_depth == depth)) { - xmlChar *buf_out; - int len_out; - xmlNodePtr ptr = xmlTextReaderExpand(tinfo->reader); xmlNodePtr ptr2 = xmlCopyNode(ptr, 1); xmlDocPtr doc = xmlNewDoc("1.0"); xmlDocSetRootElement(doc, ptr2); - - if (tinfo->stylesheet_xsp) - { - xmlDocPtr resDoc = - xsltApplyStylesheet(tinfo->stylesheet_xsp, - doc, params); - if (p->flagShowRecords) - { - xmlDocDumpMemory(resDoc, &buf_out, &len_out); - fwrite(buf_out, len_out, 1, stdout); - xmlFree(buf_out); - } - index_node(tinfo, p, xmlDocGetRootElement(resDoc), &recWord); - xmlFreeDoc(resDoc); - } - xmlDocDumpMemory(doc, &buf_out, &len_out); - if (p->flagShowRecords) - fwrite(buf_out, len_out, 1, stdout); - (*p->setStoreData)(p, buf_out, len_out); - xmlFree(buf_out); - xmlFreeDoc(doc); - return RECCTRL_EXTRACT_OK; + return extract_doc(tinfo, p, doc); } ret = xmlTextReaderRead(tinfo->reader); } @@ -238,6 +368,38 @@ static int filter_extract(void *clientData, struct recExtractCtrl *p) return RECCTRL_EXTRACT_EOF; } +static int extract_full(struct filter_info *tinfo, struct recExtractCtrl *p) +{ + if (p->first_record) /* only one record per stream */ + { + xmlDocPtr doc = xmlReadIO(ioread_ex, ioclose_ex, p /* I/O handler */, + 0 /* URL */, + 0 /* encoding */, + XML_PARSE_XINCLUDE); + if (!doc) + { + return RECCTRL_EXTRACT_ERROR_GENERIC; + } + return extract_doc(tinfo, p, doc); + } + else + return RECCTRL_EXTRACT_EOF; +} + +static int filter_extract(void *clientData, struct recExtractCtrl *p) +{ + struct filter_info *tinfo = clientData; + + odr_reset(tinfo->odr); + + if (tinfo->split_level == 0 && tinfo->split_path == 0) + return extract_full(tinfo, p); + else + { + return extract_split(tinfo, p); + } +} + static int ioread_ret(void *context, char *buffer, int len) { struct recRetrieveCtrl *p = context; @@ -249,6 +411,7 @@ static int ioclose_ret(void *context) return 0; } + static int filter_retrieve (void *clientData, struct recRetrieveCtrl *p) { const char *esn = ZEBRA_SCHEMA_IDENTITY_NS; @@ -256,6 +419,7 @@ static int filter_retrieve (void *clientData, struct recRetrieveCtrl *p) struct filter_info *tinfo = clientData; xmlDocPtr resDoc; xmlDocPtr doc; + struct filter_schema *schema; if (p->comp) { @@ -267,7 +431,14 @@ static int filter_retrieve (void *clientData, struct recRetrieveCtrl *p) } esn = p->comp->u.simple->u.generic; } - + schema = lookup_schema(tinfo, esn); + if (!schema) + { + p->diagnostic = + YAZ_BIB1_SPECIFIED_ELEMENT_SET_NAME_NOT_VALID_FOR_SPECIFIED_; + return 0; + } + params[0] = 0; set_param_str(params, "schema", esn, p->odr); if (p->fname) @@ -276,11 +447,6 @@ static int filter_retrieve (void *clientData, struct recRetrieveCtrl *p) set_param_int(params, "score", p->score, p->odr); set_param_int(params, "size", p->recordSize, p->odr); - if (!tinfo->stylesheet_xsp) - { - p->diagnostic = YAZ_BIB1_SYSTEM_ERROR_IN_PRESENTING_RECORDS; - return 0; - } doc = xmlReadIO(ioread_ret, ioclose_ret, p /* I/O handler */, 0 /* URL */, 0 /* encoding */, @@ -291,11 +457,11 @@ static int filter_retrieve (void *clientData, struct recRetrieveCtrl *p) return 0; } - if (!strcmp(esn, ZEBRA_SCHEMA_IDENTITY_NS)) + if (!schema->stylesheet_xsp) resDoc = doc; else { - resDoc = xsltApplyStylesheet(tinfo->stylesheet_xsp, + resDoc = xsltApplyStylesheet(schema->stylesheet_xsp, doc, params); xmlFreeDoc(doc); } @@ -337,10 +503,20 @@ static int filter_retrieve (void *clientData, struct recRetrieveCtrl *p) return 0; } -static struct recType filter_type = { +static struct recType filter_type_xslt = { 0, "xslt", - filter_init, + filter_init_xslt, + filter_config, + filter_destroy, + filter_extract, + filter_retrieve +}; + +static struct recType filter_type_xslt1 = { + 0, + "xslt1", + filter_init_xslt1, filter_config, filter_destroy, filter_extract, @@ -355,6 +531,9 @@ idzebra_filter #endif [] = { - &filter_type, + &filter_type_xslt, +#ifdef LIBXML_READER_ENABLED + &filter_type_xslt1, +#endif 0, };