示例#1
0
文件: rdfa.c 项目: Distrotech/raptor
static void rdfa_report_error(void* parser_context, char* msg, ...)
{
   char error[1024];
   char* eptr;
   va_list args;
   rdfacontext* context = (rdfacontext*)parser_context;

   /* format the error message */
   va_start(args, msg);
   vsprintf(error, msg, args);
   va_end(args);

   /* Remove any newlines from the libxml2 error */
   eptr = error;
   while(*eptr != '\0')
   {
      if(*eptr == '\n')
      {
         *eptr = '.';
      }
      eptr++;
   }

   /* Generate the processor error */
   rdfa_processor_triples(context, RDFA_PROCESSOR_ERROR, error);
}
示例#2
0
/**
 * Attempts to update the uri mappings in the given context using the
 * given attribute/value pair.
 *
 * @param attribute the attribute, which must start with xmlns.
 * @param value the value of the attribute
 */
void rdfa_update_uri_mappings(
   rdfacontext* context, const char* attr, const char* value)
{
#ifdef LIBRDFA_IN_RAPTOR
  raptor_namespace_stack* nstack;
  nstack = &context->sax2->namespaces;
#endif

   /* * the [current element] is parsed for [URI mappings] and these
    * are added to the [list of URI mappings]. Note that a [URI
    * mapping] will simply overwrite any current mapping in the list
    * that has the same name; */

   /* Mappings are provided by @xmlns. The value to be mapped is set
    * by the XML namespace prefix, and the value to map is the value
    * of the attribute -- a URI. Note that the URI is not processed
    * in any way; in particular if it is a relative path it is not
    * resolved against the [current base]. Authors are advised to
    * follow best practice for using namespaces, which includes not
    * using relative paths. */

   if(attr == NULL)
   {
#ifdef LIBRDFA_IN_RAPTOR
      raptor_namespaces_start_namespace_full(nstack,
                                             NULL,
                                             (const unsigned char*)value,
                                             0);
#else
      rdfa_update_mapping(
         context->uri_mappings, XMLNS_DEFAULT_MAPPING, value,
         (update_mapping_value_fp)rdfa_replace_string);
#endif
   }
   else if(strcmp(attr, "_") == 0)
   {
#define FORMAT_1 "The underscore character must not be declared as a prefix " \
         "because it conflicts with the prefix for blank node identifiers. " \
         "The occurrence of this prefix declaration is being ignored."
#ifdef LIBRDFA_IN_RAPTOR
      raptor_parser_warning((raptor_parser*)context->callback_data, 
                            FORMAT_1);
#else
      rdfa_processor_triples(context,
         RDFA_PROCESSOR_WARNING,
         FORMAT_1);
#endif
   }
   else if(attr[0] == ':' || attr[0] == '_' ||
      (attr[0] >= 'A' && attr[0] <= 'Z') ||
      (attr[0] >= 'a' && attr[0] <= 'z') ||
      ((unsigned char)attr[0] >= 0xc0 && (unsigned char)attr[0] <= 0xd6) ||
      ((unsigned char)attr[0] >= 0xd8 && (unsigned char)attr[0] <= 0xf6) || (unsigned char)attr[0] >= 0xf8)
   {
#ifdef LIBRDFA_IN_RAPTOR
     raptor_namespaces_start_namespace_full(nstack,
                                            (const unsigned char*)attr,
                                            (const unsigned char*)value,
                                            0);
#else
      rdfa_generate_namespace_triple(context, attr, value);
      rdfa_update_mapping(context->uri_mappings, attr, value,
         (update_mapping_value_fp)rdfa_replace_string);
#endif
   }
   else
   {
      /* allowable characters for CURIEs:
       * ":" | [A-Z] | "_" | [a-z] | [#xC0-#xD6] | [#xD8-#xF6] | [#xF8-#x2FF] |
       * [#x370-#x37D] | [#x37F-#x1FFF] | [#x200C-#x200D] | [#x2070-#x218F] |
       * [#x2C00-#x2FEF] | [#x3001-#xD7FF] | [#xF900-#xFDCF] | [#xFDF0-#xFFFD]
       * | [#x10000-#xEFFFF]
       */

      /* Generate the processor warning if this is an invalid prefix */
#define FORMAT_2 "The declaration of the '%s' prefix is invalid " \
         "because it starts with an invalid character. Please see " \
         "http://www.w3.org/TR/REC-xml/#NT-NameStartChar for a " \
         "full explanation of valid first characters for declaring " \
         "prefixes."
#ifdef LIBRDFA_IN_RAPTOR
      raptor_parser_warning((raptor_parser*)context->callback_data, 
                            FORMAT_2, attr);
#else
      char msg[1024];
      snprintf(msg, 1024, FORMAT_1);
      rdfa_processor_triples(context, RDFA_PROCESSOR_WARNING, msg);
#endif
   }

#ifdef LIBRDFA_IN_RAPTOR
#else
   /* print the current mapping */
   if(DEBUG)
   {
      printf("DEBUG: PREFIX MAPPINGS:");
      rdfa_print_mapping(context->uri_mappings,
         (print_mapping_value_fp)rdfa_print_string);
   }
#endif
}
示例#3
0
char* rdfa_resolve_curie(
   rdfacontext* context, const char* uri, curieparse_t mode)
{
   char* rval = NULL;
   curie_t ctype = rdfa_get_curie_type(uri);

   if(!uri)
      return NULL;

   if(ctype == CURIE_TYPE_INVALID)
   {
      rval = NULL;
   }
   else if((ctype == CURIE_TYPE_IRI_OR_UNSAFE) &&
           ((mode == CURIE_PARSE_HREF_SRC) ||
            (context->rdfa_version == RDFA_VERSION_1_0 &&
               mode == CURIE_PARSE_ABOUT_RESOURCE)))
   {
      /* If we are parsing something that can take either a CURIE or a
       * URI, and the type is either IRI or UNSAFE, assume that it is
       * an IRI */
      rval = rdfa_resolve_uri(context, uri);
   }

   /*
    * Check to see if the value is a term.
    */
   if(ctype == CURIE_TYPE_IRI_OR_UNSAFE && mode == CURIE_PARSE_PROPERTY)
   {
      const char* term_iri;
      term_iri = (const char*)rdfa_get_mapping(context->term_mappings, uri);
      if(term_iri != NULL)
      {
         rval = strdup(term_iri);
      }
      else if(context->default_vocabulary == NULL && strstr(uri, ":") == NULL)
      {
         /* Generate the processor warning if this is a missing term */
#define FORMAT_1 "The use of the '%s' term was unrecognized by the RDFa processor because it is not a valid term for the current Host Language."

#ifdef LIBRDFA_IN_RAPTOR
         raptor_parser_warning((raptor_parser*)context->callback_data, 
                               FORMAT_1, uri);
#else
         char msg[1024];
         snprintf(msg, 1024, FORMAT_1, uri);

         rdfa_processor_triples(context, RDFA_PROCESSOR_WARNING, msg);
#endif
      }
   }

   /* if we are processing a safe CURIE OR
    * if we are parsing an unsafe CURIE that is an @type_of,
    * @datatype, @property, @rel, or @rev attribute, treat the curie
    * as not an IRI, but an unsafe CURIE */
   if(rval == NULL && ((ctype == CURIE_TYPE_SAFE) ||
         ((ctype == CURIE_TYPE_IRI_OR_UNSAFE) &&
          ((mode == CURIE_PARSE_INSTANCEOF_DATATYPE) ||
           (mode == CURIE_PARSE_PROPERTY) ||
           (mode == CURIE_PARSE_RELREV) ||
           (context->rdfa_version == RDFA_VERSION_1_1 &&
              mode == CURIE_PARSE_ABOUT_RESOURCE)))))
   {
      char* working_copy = NULL;
      char* wcptr = NULL;
      char* prefix = NULL;
      char* curie_reference = NULL;
      const char* expanded_prefix = NULL;
      size_t uri_len = strlen(uri);
      working_copy = (char*)malloc(uri_len + 1);
      memcpy(working_copy, uri, uri_len + 1);/*rdfa_replace_string(working_copy, uri);*/

      /* if this is a safe CURIE, chop off the beginning and the end */
      if(ctype == CURIE_TYPE_SAFE)
      {
         prefix = strtok_r(working_copy, "[:]", &wcptr);
         if(wcptr)
            curie_reference = strtok_r(NULL, "[]", &wcptr);
      }
      else if(ctype == CURIE_TYPE_IRI_OR_UNSAFE)
      {
         prefix = strtok_r(working_copy, ":", &wcptr);
         if(wcptr)
            curie_reference = strtok_r(NULL, "", &wcptr);
      }

      /* fully resolve the prefix and get its length */

      /* if a colon was found, but no prefix, use the XHTML vocabulary URI
       * as the expanded prefix */
      if((uri[0] == ':') || (strcmp(uri, "[:]") == 0))
      {
         expanded_prefix = XHTML_VOCAB_URI;
         curie_reference = prefix;
         prefix = NULL;
      }
      else if(uri[0] == ':')
      {
         /* FIXME: This looks like a bug - don't know why this code is
          * in here. I think it's for the case where ":next" is
          * specified, but the code's not checking that -- manu */
         expanded_prefix = context->base;
         curie_reference = prefix;
         prefix = NULL;
      }
      else if(prefix != NULL)
      {
         if((mode != CURIE_PARSE_PROPERTY) &&
            (mode != CURIE_PARSE_RELREV) &&
            strcmp(prefix, "_") == 0)
         {
            /* if the prefix specifies this as a blank node, then we
             * use the blank node prefix */
            expanded_prefix = "_";
         }
         else
         {
            /* if the prefix was defined, get it from the set of URI mappings. */
#ifdef LIBRDFA_IN_RAPTOR
            if(!strcmp(prefix, "xml"))
            {
               expanded_prefix = RAPTOR_GOOD_CAST(const char*, raptor_xml_namespace_uri);
            }
            else
            {
               raptor_namespace *nspace;
               raptor_uri* ns_uri;
               nspace = raptor_namespaces_find_namespace(&context->sax2->namespaces,
                                                         (const unsigned char*)prefix,
                                                         (int)strlen(prefix));
               if(nspace) {
                  ns_uri = raptor_namespace_get_uri(nspace);
                  if(ns_uri)
                     expanded_prefix = (const char*)raptor_uri_as_string(ns_uri);
               }
            }
#else
            expanded_prefix =
               rdfa_get_mapping(context->uri_mappings, prefix);

            /* Generate the processor warning if the prefix was not found */
            if(expanded_prefix == NULL && strstr(uri, ":") != NULL &&
               strstr(uri, "://") == NULL)
            {
#define FORMAT_2 "The '%s' prefix was not found. You may want to check that it is declared before it is used, or that it is a valid prefix string."
#ifdef LIBRDFA_IN_RAPTOR
              raptor_parser_warning((raptor_parser*)context->callback_data, 
                                    FORMAT_2, prefix);
#else
              char msg[1024];
              snprintf(msg, 1024, FORMAT_2, prefix);

               rdfa_processor_triples(context, RDFA_PROCESSOR_WARNING, msg);
#endif
            }
#endif
         }