X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=recctrl%2Fxslt.c;h=d2cf67a4d3c5469689789ae573fef752d40541fd;hb=be32b4e5c154545c2b91b76144d6cbf1238f969a;hp=4aeb4557be0d5fa17a3ce2fe35d12c77e08a298c;hpb=aa9c7bcd6f5ec5a2523c40f929ebdc9d72c1b48a;p=idzebra-moved-to-github.git diff --git a/recctrl/xslt.c b/recctrl/xslt.c index 4aeb455..d2cf67a 100644 --- a/recctrl/xslt.c +++ b/recctrl/xslt.c @@ -1,4 +1,4 @@ -/* $Id: xslt.c,v 1.4 2005-05-01 07:17:46 adam Exp $ +/* $Id: xslt.c,v 1.10 2005-06-15 15:30:05 adam Exp $ Copyright (C) 1995-2005 Index Data ApS @@ -26,28 +26,48 @@ Free Software Foundation, 59 Temple Place - Suite 330, Boston, MA #include #include +#include #include -#ifdef LIBXML_READER_ENABLED +#include #include -#endif #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; + const char *include_snippet; xsltStylesheetPtr stylesheet_xsp; -#ifdef LIBXML_READER_ENABLED - xmlTextReaderPtr reader; -#endif +}; + +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" -#define ZEBRA_SCHEMA_IDENTITY_NS "http://indexdata.dk/zebra/identity/1" -static const char *zebra_index_ns = ZEBRA_INDEX_NS; +#define ZEBRA_SCHEMA_XSLT_NS "http://indexdata.dk/zebra/xslt/1" + +static const char *zebra_xslt_ns = ZEBRA_SCHEMA_XSLT_NS; + +static void set_param_xml(const char **params, const char *name, + const char *value, ODR odr) +{ + while (*params) + params++; + params[0] = name; + params[1] = value; + params[2] = 0; +} static void set_param_str(const char **params, const char *name, const char *value, ODR odr) @@ -74,16 +94,50 @@ static void set_param_int(const char **params, const char *name, } +int zebra_xmlInputMatchCallback (char const *filename) +{ + yaz_log(YLOG_LOG, "match %s", filename); + return 0; +} + + +void * zebra_xmlInputOpenCallback (char const *filename) +{ + return 0; +} + +int zebra_xmlInputReadCallback (void * context, char * buffer, int len) +{ + return 0; +} + +int zebra_xmlInputCloseCallback (void * context) +{ + return 0; +} + + + + + static void *filter_init_xslt(Res res, RecType recType) { struct filter_info *tinfo = (struct filter_info *) xmalloc(sizeof(*tinfo)); - tinfo->stylesheet_xsp = 0; -#ifdef LIBXML_READER_ENABLED tinfo->reader = 0; -#endif tinfo->fname = 0; - tinfo->split_depth = 0; + tinfo->split_level = 0; + tinfo->split_path = 0; tinfo->odr = odr_createmem(ODR_ENCODE); + tinfo->doc = 0; + tinfo->schemas = 0; + +#if 0 + xmlRegisterDefaultInputCallbacks(); + xmlRegisterInputCallbacks(zebra_xmlInputMatchCallback, + zebra_xmlInputOpenCallback, + zebra_xmlInputReadCallback, + zebra_xmlInputCloseCallback); +#endif return tinfo; } @@ -91,37 +145,134 @@ static void *filter_init_xslt1(Res res, RecType recType) { struct filter_info *tinfo = (struct filter_info *) filter_init_xslt(res, recType); - tinfo->split_depth = 1; + 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; + schema->include_snippet = 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); + attr_content(attr, "snippet", &schema->include_snippet); + } + 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); -#ifdef LIBXML_READER_ENABLED + destroy_schemas(tinfo); if (tinfo->reader) xmlFreeTextReader(tinfo->reader); -#endif - xfree(tinfo->fname); odr_destroy(tinfo->odr); xfree(tinfo); } @@ -137,12 +288,12 @@ static int ioclose_ex(void *context) return 0; } -static void index_field(struct filter_info *tinfo, struct recExtractCtrl *ctrl, +static void index_cdata(struct filter_info *tinfo, struct recExtractCtrl *ctrl, xmlNodePtr ptr, RecWord *recWord) { for(; ptr; ptr = ptr->next) { - index_field(tinfo, ctrl, ptr->children, recWord); + index_cdata(tinfo, ctrl, ptr->children, recWord); if (ptr->type != XML_TEXT_NODE) continue; recWord->term_buf = ptr->content; @@ -158,26 +309,26 @@ static void index_node(struct filter_info *tinfo, struct recExtractCtrl *ctrl, { index_node(tinfo, ctrl, ptr->children, recWord); if (ptr->type != XML_ELEMENT_NODE || !ptr->ns || - strcmp(ptr->ns->href, zebra_index_ns)) + strcmp(ptr->ns->href, zebra_xslt_ns)) continue; if (!strcmp(ptr->name, "index")) { - char *field_str = 0; + char *name_str = 0; const char *xpath_str = 0; struct _xmlAttr *attr; for (attr = ptr->properties; attr; attr = attr->next) { - if (!strcmp(attr->name, "field") + if (!strcmp(attr->name, "name") && attr->children && attr->children->type == XML_TEXT_NODE) - field_str = attr->children->content; + name_str = attr->children->content; if (!strcmp(attr->name, "xpath") && attr->children && attr->children->type == XML_TEXT_NODE) xpath_str = attr->children->content; } - if (field_str) + if (name_str) { - recWord->attrStr = field_str; - index_field(tinfo, ctrl, ptr->children, recWord); + recWord->attrStr = name_str; + index_cdata(tinfo, ctrl, ptr->children, recWord); } } } @@ -188,19 +339,21 @@ static int extract_doc(struct filter_info *tinfo, struct recExtractCtrl *p, { RecWord recWord; const char *params[10]; - params[0] = 0; xmlChar *buf_out; int len_out; - set_param_str(params, "schema", ZEBRA_INDEX_NS, tinfo->odr); + struct filter_schema *schema = lookup_schema(tinfo, zebra_xslt_ns); + + params[0] = 0; + set_param_str(params, "schema", zebra_xslt_ns, tinfo->odr); (*p->init)(p, &recWord); recWord.reg_type = 'w'; - if (tinfo->stylesheet_xsp) + if (schema && schema->stylesheet_xsp) { xmlDocPtr resDoc = - xsltApplyStylesheet(tinfo->stylesheet_xsp, + xsltApplyStylesheet(schema->stylesheet_xsp, doc, params); if (p->flagShowRecords) { @@ -221,10 +374,10 @@ static int extract_doc(struct filter_info *tinfo, struct recExtractCtrl *p, return RECCTRL_EXTRACT_OK; } -#ifdef LIBXML_READER_ENABLED static int extract_split(struct filter_info *tinfo, struct recExtractCtrl *p) { int ret; + int split_depth = 0; if (p->first_record) { if (tinfo->reader) @@ -238,15 +391,15 @@ static int extract_split(struct filter_info *tinfo, struct recExtractCtrl *p) if (!tinfo->reader) return RECCTRL_EXTRACT_ERROR_GENERIC; - if (!tinfo->stylesheet_xsp) - return RECCTRL_EXTRACT_ERROR_GENERIC; - + 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)) { xmlNodePtr ptr = xmlTextReaderExpand(tinfo->reader); xmlNodePtr ptr2 = xmlCopyNode(ptr, 1); @@ -262,7 +415,6 @@ static int extract_split(struct filter_info *tinfo, struct recExtractCtrl *p) tinfo->reader = 0; return RECCTRL_EXTRACT_EOF; } -#endif static int extract_full(struct filter_info *tinfo, struct recExtractCtrl *p) { @@ -288,16 +440,11 @@ static int filter_extract(void *clientData, struct recExtractCtrl *p) odr_reset(tinfo->odr); - if (tinfo->split_depth == 0) + if (tinfo->split_level == 0 && tinfo->split_path == 0) return extract_full(tinfo, p); else { -#ifdef LIBXML_READER_ENABLED return extract_split(tinfo, p); -#else - /* no xmlreader so we can't split it */ - return RECCTRL_EXTRACT_ERROR_GENERIC; -#endif } } @@ -312,13 +459,63 @@ static int ioclose_ret(void *context) return 0; } + +static const char *snippet_doc(struct recRetrieveCtrl *p, int text_mode, + int window_size) +{ + const char *xml_doc_str; + int ord = 0; + WRBUF wrbuf = wrbuf_alloc(); + zebra_snippets *res = + zebra_snippets_window(p->doc_snippet, p->hit_snippet, window_size); + zebra_snippet_word *w = zebra_snippets_list(res); + + if (text_mode) + wrbuf_printf(wrbuf, "\'"); + else + wrbuf_printf(wrbuf, "\n", zebra_xslt_ns); + for (; w; w = w->next) + { + if (ord == 0) + ord = w->ord; + else if (ord != w->ord) + + break; + if (text_mode) + wrbuf_printf(wrbuf, "%s%s%s ", + w->match ? "*" : "", + w->term, + w->match ? "*" : ""); + else + { + wrbuf_printf(wrbuf, " ", + w->ord, w->seqno, + (w->match ? "match='1'" : "")); + wrbuf_xmlputs(wrbuf, w->term); + wrbuf_printf(wrbuf, "\n"); + } + } + if (text_mode) + wrbuf_printf(wrbuf, "\'"); + else + wrbuf_printf(wrbuf, "\n"); + + xml_doc_str = odr_strdup(p->odr, wrbuf_buf(wrbuf)); + + zebra_snippets_destroy(res); + wrbuf_free(wrbuf, 1); + return xml_doc_str; +} + static int filter_retrieve (void *clientData, struct recRetrieveCtrl *p) { - const char *esn = ZEBRA_SCHEMA_IDENTITY_NS; + const char *esn = zebra_xslt_ns; const char *params[10]; struct filter_info *tinfo = clientData; xmlDocPtr resDoc; xmlDocPtr doc; + struct filter_schema *schema; + int window_size = -1; if (p->comp) { @@ -330,7 +527,17 @@ 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; + } + + if (schema->include_snippet) + window_size = atoi(schema->include_snippet); + params[0] = 0; set_param_str(params, "schema", esn, p->odr); if (p->fname) @@ -339,11 +546,9 @@ 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; - } + if (window_size >= 0) + set_param_xml(params, "snippet", snippet_doc(p, 1, window_size), + p->odr); doc = xmlReadIO(ioread_ret, ioclose_ret, p /* I/O handler */, 0 /* URL */, 0 /* encoding */, @@ -354,11 +559,18 @@ static int filter_retrieve (void *clientData, struct recRetrieveCtrl *p) return 0; } - if (!strcmp(esn, ZEBRA_SCHEMA_IDENTITY_NS)) + if (window_size >= 0) + { + xmlNodePtr node = xmlDocGetRootElement(doc); + const char *snippet_str = snippet_doc(p, 0, window_size); + xmlDocPtr snippet_doc = xmlParseMemory(snippet_str, strlen(snippet_str)); + xmlAddChild(node, xmlDocGetRootElement(snippet_doc)); + } + if (!schema->stylesheet_xsp) resDoc = doc; else { - resDoc = xsltApplyStylesheet(tinfo->stylesheet_xsp, + resDoc = xsltApplyStylesheet(schema->stylesheet_xsp, doc, params); xmlFreeDoc(doc); }