/** * Resolves a given uri depending on whether or not it is a fully * qualified IRI, a CURIE, or a short-form XHTML reserved word for * @rel or @rev as defined in the XHTML+RDFa Syntax Document. * * @param context the current processing context. * @param uri the URI part to process. * * @return the fully qualified IRI, or NULL if the conversion failed * due to the given URI not being a short-form XHTML reserved * word. The memory returned from this function MUST be freed. */ char* rdfa_resolve_relrev_curie(rdfacontext* context, const char* uri) { char* rval = NULL; int i = 0; const char* resource = uri; // check to make sure the URI doesn't have an empty prefix if(uri[0] == ':') { resource++; } // search all of the XHTML @rel/@rev reserved words for a // case-insensitive match against the given URI for(i = 0; i < XHTML_RELREV_RESERVED_WORDS_SIZE; i++) { if(strcasecmp(g_relrev_reserved_words[i], resource) == 0) { // since the URI is a reserved word for @rel/@rev, generate // the full IRI and stop the loop. rval = rdfa_join_string(XHTML_VOCAB_URI, g_relrev_reserved_words[i]); i = XHTML_RELREV_RESERVED_WORDS_SIZE; } } // if none of the XHTML @rel/@rev reserved words were found, // attempt to resolve the value as a standard CURIE if(rval == NULL) { rval = rdfa_resolve_curie(context, uri, CURIE_PARSE_RELREV); } return rval; }
rdfalist* rdfa_resolve_curie_list( rdfacontext* rdfa_context, const char* uris, curieparse_t mode) { rdfalist* rval = rdfa_create_list(3); char* working_uris = NULL; char* uptr = NULL; char* ctoken = NULL; working_uris = rdfa_replace_string(working_uris, uris); // go through each item in the list of CURIEs and resolve each ctoken = strtok_r(working_uris, RDFA_WHITESPACE, &uptr); while(ctoken != NULL) { char* resolved_curie = NULL; if((mode == CURIE_PARSE_INSTANCEOF_DATATYPE) || (mode == CURIE_PARSE_ABOUT_RESOURCE) || (mode == CURIE_PARSE_PROPERTY)) { resolved_curie = rdfa_resolve_curie(rdfa_context, ctoken, mode); } else if(mode == CURIE_PARSE_RELREV) { resolved_curie = rdfa_resolve_relrev_curie(rdfa_context, ctoken); } // add the CURIE if it was a valid one if(resolved_curie != NULL) { rdfa_add_item(rval, resolved_curie, RDFALIST_FLAG_TEXT); free(resolved_curie); } ctoken = strtok_r(NULL, RDFA_WHITESPACE, &uptr); } free(working_uris); return rval; }
/** * Handles the start_element call */ static void start_element(void *parser_context, const char* name, const char* prefix, const char* URI, int nb_namespaces, const char** namespaces, int nb_attributes, int nb_defaulted, const char** attributes) { rdfacontext* root_context = (rdfacontext*)parser_context; rdfalist* context_stack = (rdfalist*)root_context->context_stack; rdfacontext* context = rdfa_create_new_element_context(context_stack); char* xml_lang = NULL; const char* about_curie = NULL; char* about = NULL; const char* src_curie = NULL; char* src = NULL; const char* type_of_curie = NULL; rdfalist* type_of = NULL; const char* rel_curie = NULL; rdfalist* rel = NULL; const char* rev_curie = NULL; rdfalist* rev = NULL; const char* property_curie = NULL; rdfalist* property = NULL; const char* resource_curie = NULL; char* resource = NULL; const char* href_curie = NULL; char* href = NULL; char* content = NULL; const char* datatype_curie = NULL; char* datatype = NULL; #ifdef LIBRDFA_IN_RAPTOR if(1) { raptor_parser* rdf_parser = (raptor_parser*)context->callback_data; raptor_sax2_update_document_locator(context->sax2, &rdf_parser->locator); } #endif rdfa_push_item(context_stack, context, RDFALIST_FLAG_CONTEXT); #if defined(DEBUG) && DEBUG > 0 if(1) { int i; /* dump all arguments sent to this callback */ fprintf(stdout, "DEBUG: SAX.startElementNs(%s", (char *) name); if (prefix == NULL) fprintf(stdout, ", NULL"); else fprintf(stdout, ", %s", (char *) prefix); if (URI == NULL) fprintf(stdout, ", NULL"); else fprintf(stdout, ", '%s'", (char *) URI); fprintf(stdout, ", %d", nb_namespaces); /* dump all namespaces */ if (namespaces != NULL) { for (i = 0;i < nb_namespaces * 2;i++) { fprintf(stdout, ", xmlns"); if (namespaces[i] != NULL) fprintf(stdout, ":%s", namespaces[i]); i++; fprintf(stdout, "='%s'", namespaces[i]); } } /* dump all attributes */ fprintf(stdout, ", %d, %d", nb_attributes, nb_defaulted); if (attributes != NULL) { for (i = 0;i < nb_attributes * 5;i += 5) { if (attributes[i + 1] != NULL) fprintf( stdout, ", %s:%s='", attributes[i + 1], attributes[i]); else fprintf(stdout, ", %s='", attributes[i]); fprintf(stdout, "%.4s...', %d", attributes[i + 3], (int)(attributes[i + 4] - attributes[i + 3])); } } fprintf(stdout, ")\n"); } #endif /* start the XML Literal text */ if(context->xml_literal == NULL) { context->xml_literal = rdfa_replace_string(context->xml_literal, "<"); context->xml_literal_size = 1; } else { context->xml_literal = rdfa_n_append_string( context->xml_literal, &context->xml_literal_size, "<", 1); } context->xml_literal = rdfa_n_append_string( context->xml_literal, &context->xml_literal_size, name, strlen(name)); if(!context->xml_literal_namespaces_defined) { /* append namespaces to XML Literal */ #ifdef LIBRDFA_IN_RAPTOR raptor_namespace_stack* nstack = &context->sax2->namespaces; raptor_namespace* ns; raptor_namespace** ns_list = NULL; size_t ns_size; #else void** umap = context->uri_mappings; #endif char* umap_key = NULL; void* umap_value = NULL; /* if the namespaces are not defined, then neither is the xml:lang */ context->xml_literal_xml_lang_defined = 0; #ifdef LIBRDFA_IN_RAPTOR ns_size = 0; ns_list = raptor_namespace_stack_to_array(nstack, &ns_size); qsort((void*)ns_list, ns_size, sizeof(raptor_namespace*), raptor_nspace_compare); while(ns_size > 0) #else while(*umap != NULL) #endif { unsigned char insert_xmlns_definition = 1; const char* attr = NULL; /* get the next mapping to process */ #ifdef LIBRDFA_IN_RAPTOR ns=ns_list[--ns_size]; umap_key = (char*)raptor_namespace_get_prefix(ns); if(!umap_key) umap_key=(char*)XMLNS_DEFAULT_MAPPING; umap_value = (char*)raptor_uri_as_string(raptor_namespace_get_uri(ns)); #else rdfa_next_mapping(umap++, &umap_key, &umap_value); umap++; #endif /* check to make sure that the namespace isn't already * defined in the current element. */ if(attributes != NULL) { const char** attrs = attributes; while((*attrs != NULL) && insert_xmlns_definition) { attr = *attrs++; /* if the attribute is a umap_key, skip the definition * of the attribute. */ if(strcmp(attr, umap_key) == 0) { insert_xmlns_definition = 0; } } } /* if the namespace isn't already defined on the element, * copy it to the XML Literal string. */ if(insert_xmlns_definition) { /* append the namespace attribute to the XML Literal */ context->xml_literal = rdfa_n_append_string( context->xml_literal, &context->xml_literal_size, " xmlns", strlen(" xmlns")); /* check to see if we're dumping the standard XHTML namespace or * a user-defined XML namespace */ if(strcmp(umap_key, XMLNS_DEFAULT_MAPPING) != 0) { context->xml_literal = rdfa_n_append_string( context->xml_literal, &context->xml_literal_size, ":", 1); context->xml_literal = rdfa_n_append_string( context->xml_literal, &context->xml_literal_size, umap_key, strlen(umap_key)); } /* append the namespace value */ context->xml_literal = rdfa_n_append_string( context->xml_literal, &context->xml_literal_size, "=\"", 2); context->xml_literal = rdfa_n_append_string( context->xml_literal, &context->xml_literal_size, (const char*)umap_value, strlen((char*)umap_value)); context->xml_literal = rdfa_n_append_string( context->xml_literal, &context->xml_literal_size, "\"", 1); } } /* end while umap not NULL */ context->xml_literal_namespaces_defined = 1; #ifdef LIBRDFA_IN_RAPTOR if(ns_list) raptor_free_memory(ns_list); #endif } /* end if namespaces inserted */ #ifdef LIBRDFA_IN_RAPTOR /* Raptor namespace code does this already */ #else /* 3. For backward compatibility, RDFa Processors should also permit the * definition of mappings via @xmlns. In this case, the value to be mapped * is set by the XML namespace prefix, and the value to map is the value of * the attribute - an IRI. (Note that prefix mapping via @xmlns is * deprecated, and may be removed in a future version of this * specification.) When xmlns is supported, such mappings must be processed * before processing any mappings from @prefix on the same element. */ if(namespaces != NULL) { int ni; for(ni = 0; ni < nb_namespaces * 2; ni += 2) { const char* ns = namespaces[ni]; const char* value = namespaces[ni + 1]; /* Regardless of how the mapping is declared, the value to be mapped * must be converted to lower case, and the IRI is not processed in * any way; in particular if it is a relative path it must not be * resolved against the current base. */ char* lcns = NULL; if(ns != NULL) { /* convert the namespace string to lowercase */ unsigned int i; size_t ns_length = strlen(ns); lcns = (char*)malloc(ns_length + 1); for(i = 0; i <= ns_length; i++) { lcns[i] = tolower(ns[i]); } } /* update the URI mappings */ rdfa_update_uri_mappings(context, lcns, value); if(lcns != NULL) { free(lcns); } } } #endif /* detect the RDFa version of the document, if specified */ if(attributes != NULL) { int ci; /* search for a version attribute */ for(ci = 0; ci < nb_attributes * 5; ci += 5) { const char* attr; char* value; size_t value_length = 0; attr = attributes[ci]; value_length = attributes[ci + 4] - attributes[ci + 3] + 1; if(strcmp(attr, "version") == 0) { /* append the attribute-value pair to the XML literal */ value = (char*)malloc(value_length + 1); snprintf(value, value_length, "%s", attributes[ci + 3]); if(strstr(value, "RDFa 1.0") != NULL) { context->rdfa_version = RDFA_VERSION_1_0; } else if(strstr(value, "RDFa 1.1") != NULL) { context->rdfa_version = RDFA_VERSION_1_1; } free(value); } } } #ifdef LIBRDFA_IN_RAPTOR if(context->sax2) { /* Raptor handles xml:lang itself but not 'lang' */ xml_lang = (char*)raptor_sax2_inscope_xml_language(context->sax2); xml_lang = rdfa_replace_string(NULL, xml_lang); } #endif /* prepare all of the RDFa-specific attributes we are looking for. * scan all of the attributes for the RDFa-specific attributes */ if(attributes != NULL) { int ci; if(context->rdfa_version == RDFA_VERSION_1_1) { /* process all vocab and prefix attributes */ for(ci = 0; ci < nb_attributes * 5; ci += 5) { const char* attr; char* value; size_t value_length = 0; attr = attributes[ci]; value_length = attributes[ci + 4] - attributes[ci + 3] + 1; /* append the attribute-value pair to the XML literal */ value = (char*)malloc(value_length + 1); snprintf(value, value_length, "%s", attributes[ci + 3]); /* 2. Next the current element is examined for any change to the * default vocabulary via @vocab. */ if(strcmp(attr, "vocab") == 0) { if(strlen(value) < 1) { /* If the value is empty, then the local default vocabulary * must be reset to the Host Language defined default * (if any). */ free(context->default_vocabulary); context->default_vocabulary = NULL; } else { char* resolved_uri; rdftriple* triple; /* If @vocab is present and contains a value, the local * default vocabulary is updated according to the * section on CURIE and IRI Processing. */ resolved_uri = rdfa_resolve_uri(context, value); context->default_vocabulary = rdfa_replace_string( context->default_vocabulary, resolved_uri); /* The value of @vocab is used to generate a triple */ triple = rdfa_create_triple( context->base, "http://www.w3.org/ns/rdfa#usesVocabulary", resolved_uri, RDF_TYPE_IRI, NULL, NULL); context->default_graph_triple_callback( triple, context->callback_data); free(resolved_uri); } } else if(strcmp(attr, "prefix") == 0) { /* Mappings are defined via @prefix. */ char* working_string = NULL; char* atprefix = NULL; char* iri = NULL; char* saveptr = NULL; working_string = rdfa_replace_string(working_string, value); /* Values in this attribute are evaluated from beginning to * end (e.g., left to right in typical documents). */ atprefix = strtok_r(working_string, ":", &saveptr); while(atprefix != NULL) { /* find the prefix and IRI mappings while skipping whitespace */ while((*saveptr == ' ' || *saveptr == '\n' || *saveptr == '\r' || *saveptr == '\t' || *saveptr == '\f' || *saveptr == '\v') && *saveptr != '\0') { saveptr++; } iri = strtok_r(NULL, RDFA_WHITESPACE, &saveptr); /* update the prefix mappings */ rdfa_update_uri_mappings(context, atprefix, iri); if(!saveptr) break; while((*saveptr == ' ' || *saveptr == '\n' || *saveptr == '\r' || *saveptr == '\t' || *saveptr == '\f' || *saveptr == '\v') && *saveptr != '\0') { saveptr++; } /* get the next prefix to process */ atprefix = strtok_r(NULL, ":", &saveptr); } free(working_string); } else if(strcmp(attr, "inlist") == 0) { context->inlist_present = 1; } free(value); } } /* resolve all of the other RDFa values */ for(ci = 0; ci < nb_attributes * 5; ci += 5) { const char* attr; char* value; char* attrns; char* literal_text; size_t value_length = 0; attr = attributes[ci]; attrns = (char*)attributes[ci + 1]; value_length = attributes[ci + 4] - attributes[ci + 3] + 1; /* append the attribute-value pair to the XML literal */ value = (char*)malloc(value_length + 1); literal_text = (char*)malloc(strlen(attr) + value_length + 5); snprintf(value, value_length, "%s", attributes[ci + 3]); sprintf(literal_text, " %s=\"%s\"", attr, value); context->xml_literal = rdfa_n_append_string( context->xml_literal, &context->xml_literal_size, literal_text, strlen(literal_text)); free(literal_text); /* if xml:lang is defined, ensure that it is not overwritten */ if(attrns != NULL && strcmp(attrns, "xml") == 0 && strcmp(attr, "lang") == 0) { context->xml_literal_xml_lang_defined = 1; } /* process all of the RDFa attributes */ if(strcmp(attr, "about") == 0) { about_curie = value; about = rdfa_resolve_curie( context, about_curie, CURIE_PARSE_ABOUT_RESOURCE); } else if(strcmp(attr, "src") == 0) { src_curie = value; src = rdfa_resolve_curie(context, src_curie, CURIE_PARSE_HREF_SRC); } else if(strcmp(attr, "typeof") == 0) { type_of_curie = value; type_of = rdfa_resolve_curie_list( context, type_of_curie, CURIE_PARSE_INSTANCEOF_DATATYPE); } else if(strcmp(attr, "rel") == 0) { context->rel_present = 1; rel_curie = value; rel = rdfa_resolve_curie_list( context, rel_curie, CURIE_PARSE_RELREV); } else if(strcmp(attr, "rev") == 0) { context->rev_present = 1; rev_curie = value; rev = rdfa_resolve_curie_list( context, rev_curie, CURIE_PARSE_RELREV); } else if(strcmp(attr, "property") == 0) { property_curie = value; property = rdfa_resolve_curie_list( context, property_curie, CURIE_PARSE_PROPERTY); } else if(strcmp(attr, "resource") == 0) { resource_curie = value; resource = rdfa_resolve_curie( context, resource_curie, CURIE_PARSE_ABOUT_RESOURCE); } else if(strcmp(attr, "href") == 0) { href_curie = value; href = rdfa_resolve_curie(context, href_curie, CURIE_PARSE_HREF_SRC); } else if(strcmp(attr, "content") == 0) { content = rdfa_replace_string(content, value); } else if(strcmp(attr, "datatype") == 0) { datatype_curie = value; if(strlen(datatype_curie) == 0) { datatype = rdfa_replace_string(datatype, ""); } else { datatype = rdfa_resolve_curie(context, datatype_curie, CURIE_PARSE_INSTANCEOF_DATATYPE); } } else if((attrns == NULL && strcmp(attr, "lang") == 0) || (attrns != NULL && strcmp(attrns, "xml") == 0 && strcmp(attr, "lang") == 0)) { xml_lang = rdfa_replace_string(xml_lang, value); } free(value); } } /* The root element has an implicit @about declaration */ if(context->depth == 1 && about == NULL && resource == NULL && href == NULL && src == NULL) { about_curie = ""; about = rdfa_resolve_curie( context, about_curie, CURIE_PARSE_ABOUT_RESOURCE); } /* The HEAD and BODY element in XHTML and HTML has an implicit * about="" on it. */ if(about == NULL && resource == NULL && href == NULL && src == NULL && (context->parent_subject == NULL || type_of != NULL) && ((context->host_language == HOST_LANGUAGE_XHTML1 || context->host_language == HOST_LANGUAGE_HTML) && (strcasecmp(name, "head") == 0 || strcasecmp(name, "body") == 0))) { about_curie = ""; about = rdfa_resolve_curie( context, about_curie, CURIE_PARSE_ABOUT_RESOURCE); } /* check to see if we should append an xml:lang to the XML Literal * if one is defined in the context and does not exist on the * element. */ if((xml_lang == NULL) && (context->language != NULL) && !context->xml_literal_xml_lang_defined) { context->xml_literal = rdfa_n_append_string( context->xml_literal, &context->xml_literal_size, " xml:lang=\"", strlen(" xml:lang=\"")); context->xml_literal = rdfa_n_append_string( context->xml_literal, &context->xml_literal_size, context->language, strlen(context->language)); context->xml_literal = rdfa_n_append_string( context->xml_literal, &context->xml_literal_size, "\"", 1); /* ensure that the lang isn't set in a subtree (unless it's overwritten) */ context->xml_literal_xml_lang_defined = 1; } /* close the XML Literal value */ context->xml_literal = rdfa_n_append_string( context->xml_literal, &context->xml_literal_size, ">", 1); /* 3. The [current element] is also parsed for any language * information, and [language] is set in the [current * evaluation context]; */ rdfa_update_language(context, xml_lang); /***************** FOR DEBUGGING PURPOSES ONLY ******************/ #if defined(DEBUG) && DEBUG > 0 printf("DEBUG: depth = %u\n", context->depth); if(about != NULL) { printf("DEBUG: @about = %s\n", about); } if(src != NULL) { printf("DEBUG: @src = %s\n", src); } if(type_of != NULL) { printf("DEBUG: @type_of = "); rdfa_print_list(type_of); } if(context->inlist_present) { printf("DEBUG: @inlist = true\n"); } if(rel != NULL) { printf("DEBUG: @rel = "); rdfa_print_list(rel); } if(rev != NULL) { printf("DEBUG: @rev = "); rdfa_print_list(rev); } if(property != NULL) { printf("DEBUG: @property = "); rdfa_print_list(property); } if(resource != NULL) { printf("DEBUG: @resource = %s\n", resource); } if(href != NULL) { printf("DEBUG: @href = %s\n", href); } if(content != NULL) { printf("DEBUG: @content = %s\n", content); } if(datatype != NULL) { printf("DEBUG: @datatype = %s\n", datatype); } if(xml_lang != NULL) { printf("DEBUG: @xml:lang = %s\n", xml_lang); } #endif /* TODO: This isn't part of the processing model, it needs to be * included and is a correction for the last item in step #4. */ if((about == NULL) && (src == NULL) && (type_of == NULL) && (rel == NULL) && (rev == NULL) && (property == NULL) && (resource == NULL) && (href == NULL) && (context->default_vocabulary == NULL) && (prefix == NULL)) { context->skip_element = 1; } if((rel == NULL) && (rev == NULL)) { if(context->rdfa_version == RDFA_VERSION_1_0) { /* 4. If the [current element] contains no valid @rel or @rev * URI, obtained according to the section on CURIE and URI * Processing, then the next step is to establish a value for * [new subject]. Any of the attributes that can carry a * resource can set [new subject]; */ rdfa_establish_new_1_0_subject( context, name, about, src, resource, href, type_of); } else { rdfa_establish_new_1_1_subject( context, name, about, src, resource, href, type_of, property, content, datatype); } } else { if(context->rdfa_version == RDFA_VERSION_1_0) { /* 5. If the [current element] does contain a valid @rel or @rev * URI, obtained according to the section on CURIE and URI * Processing, then the next step is to establish both a value * for [new subject] and a value for [current object resource]: */ rdfa_establish_new_1_0_subject_with_relrev( context, name, about, src, resource, href, type_of); } else { rdfa_establish_new_1_1_subject_with_relrev( context, name, about, src, resource, href, type_of); } } if(context->new_subject != NULL) { #if defined(DEBUG) && DEBUG > 0 printf("DEBUG: new_subject = %s\n", context->new_subject); #endif /* RDFa 1.0: 6. If in any of the previous steps a [new subject] was set * to a non-null value, it is now used to provide a subject for * type values; */ /* RDFa 1.1: 7. If in any of the previous steps a typed resource was set * to a non-null value, it is now used to provide a subject for type * values; */ if(type_of != NULL) { rdfa_complete_type_triples(context, type_of); } /* Note that none of this block is executed if there is no * [new subject] value, i.e., [new subject] remains null. */ } if(context->current_object_resource != NULL) { /* If the element contains both the @inlist and the @rel attributes: * the @rel may contain one or more resources, obtained according to * the section on CURIE and IRI Processing each of which is used to * add an entry to the list mapping as follows: * if the local list mapping does not contain a list associated with * the IRI, instantiate a new list and add to local list mappings * add the current object resource to the list associated with the * resource in the local list mapping */ if(context->rdfa_version == RDFA_VERSION_1_1 && (rel != NULL) && context->inlist_present) { rdfresource_t object_type = RDF_TYPE_IRI; if((property != NULL) || (content != NULL)) { object_type = RDF_TYPE_PLAIN_LITERAL; if(datatype != NULL) { object_type = RDF_TYPE_TYPED_LITERAL; } } rdfa_establish_new_inlist_triples( context, rel, context->current_object_resource, object_type); } /* 7. If in any of the previous steps a [current object resource] * was set to a non-null value, it is now used to generate triples */ rdfa_complete_relrev_triples(context, rel, rev); } if((context->current_object_resource == NULL) && context->rdfa_version == RDFA_VERSION_1_1 && (rel != NULL) && context->inlist_present) { rdfa_save_incomplete_list_triples(context, rel); } else if((context->current_object_resource == NULL) && ((rel != NULL) || (rev != NULL))) { /* 8. If however [current object resource] was set to null, but * there are predicates present, then they must be stored as * [incomplete triple]s, pending the discovery of a subject that * can be used as the object. Also, [current object resource] * should be set to a newly created [bnode] */ rdfa_save_incomplete_triples(context, rel, rev); } /* Ensure to re-insert XML Literal namespace information from this * point on... */ if(property != NULL) { context->xml_literal_namespaces_defined = 0; } /* save these for processing steps #9 and #10 */ context->about = rdfa_replace_string(context->about, about); context->resource = rdfa_replace_string(context->resource, resource); context->href = rdfa_replace_string(context->href, href); context->src = rdfa_replace_string(context->src, src); context->content = rdfa_replace_string(context->content, content); context->datatype = rdfa_replace_string(context->datatype, datatype); context->property = property; /* free the resolved CURIEs */ free(about); free(src); rdfa_free_list(type_of); rdfa_free_list(rel); rdfa_free_list(rev); free(xml_lang); free(content); free(resource); free(href); free(datatype); }