static gboolean
get_file_metadata (TrackerExtractTask *task,
TrackerExtractInfo **info_out)
{
TrackerExtractInfo *info;
GFile *file;
gchar *mime_used = NULL;
gint items = 0;
*info_out = NULL;
file = g_file_new_for_uri (task->file);
info = tracker_extract_info_new (file, task->mimetype, task->graph);
g_object_unref (file);
if (task->mimetype && *task->mimetype) {
/* We know the mime */
mime_used = g_strdup (task->mimetype);
} else {
tracker_extract_info_unref (info);
return FALSE;
}
/* Now we have sanity checked everything, actually get the
* data we need from the extractors.
*/
if (mime_used) {
if (task->cur_func) {
TrackerSparqlBuilder *statements;
g_debug ("Using %s...", g_module_name (task->cur_module));
(task->cur_func) (info);
statements = tracker_extract_info_get_metadata_builder (info);
items = tracker_sparql_builder_get_length (statements);
if (items > 0) {
tracker_sparql_builder_insert_close (statements);
task->success = TRUE;
}
}
g_free (mime_used);
}
g_debug ("Done (%d objects added)\n", items);
if (items == 0) {
tracker_extract_info_unref (info);
info = NULL;
}
*info_out = info;
return (items > 0);
}
void
tracker_extract_get_metadata_by_cmdline (TrackerExtract *object,
const gchar *uri,
const gchar *mime)
{
GError *error = NULL;
TrackerExtractPrivate *priv;
TrackerExtractTask *task;
TrackerExtractInfo *info;
gboolean no_modules = TRUE;
priv = TRACKER_EXTRACT_GET_PRIVATE (object);
priv->disable_summary_on_finalize = TRUE;
g_return_if_fail (uri != NULL);
task = extract_task_new (object, uri, mime, NULL, NULL, NULL, &error);
if (error) {
g_printerr ("Extraction failed, %s\n", error->message);
g_error_free (error);
return;
}
task->mimetype_handlers = tracker_extract_module_manager_get_mimetype_handlers (task->mimetype);
task->cur_module = tracker_mimetype_info_get_module (task->mimetype_handlers, &task->cur_func, NULL);
while (task->cur_module && task->cur_func) {
if (!filter_module (object, task->cur_module) &&
get_file_metadata (task, &info)) {
const gchar *preupdate_str, *postupdate_str, *statements_str, *where;
TrackerSparqlBuilder *builder;
no_modules = FALSE;
preupdate_str = statements_str = postupdate_str = NULL;
builder = tracker_extract_info_get_metadata_builder (info);
if (tracker_sparql_builder_get_length (builder) > 0) {
statements_str = tracker_sparql_builder_get_result (builder);
}
builder = tracker_extract_info_get_preupdate_builder (info);
if (tracker_sparql_builder_get_length (builder) > 0) {
preupdate_str = tracker_sparql_builder_get_result (builder);
}
builder = tracker_extract_info_get_postupdate_builder (info);
if (tracker_sparql_builder_get_length (builder) > 0) {
postupdate_str = tracker_sparql_builder_get_result (builder);
}
where = tracker_extract_info_get_where_clause (info);
g_print ("\n");
g_print ("SPARQL pre-update:\n--\n%s--\n\n",
preupdate_str ? preupdate_str : "");
g_print ("SPARQL item:\n--\n%s--\n\n",
statements_str ? statements_str : "");
g_print ("SPARQL where clause:\n--\n%s--\n\n",
where ? where : "");
g_print ("SPARQL post-update:\n--\n%s--\n\n",
postupdate_str ? postupdate_str : "");
tracker_extract_info_unref (info);
break;
} else {
if (!tracker_mimetype_info_iter_next (task->mimetype_handlers)) {
break;
}
task->cur_module = tracker_mimetype_info_get_module (task->mimetype_handlers,
&task->cur_func,
NULL);
}
}
if (no_modules) {
g_print ("No modules found to handle metadata extraction\n\n");
}
extract_task_free (task);
}
static gboolean
get_file_metadata (TrackerExtractTask *task,
TrackerExtractInfo **info_out)
{
TrackerExtractInfo *info;
GFile *file;
gchar *mime_used = NULL;
#ifdef HAVE_LIBSTREAMANALYZER
gchar *content_type = NULL;
#endif
gint items = 0;
g_debug ("Extracting...");
*info_out = NULL;
file = g_file_new_for_uri (task->file);
info = tracker_extract_info_new (file, task->mimetype, task->graph);
g_object_unref (file);
#ifdef HAVE_LIBSTREAMANALYZER
/* FIXME: This entire section is completely broken,
* it doesn't even build these days. It should be removed or fixed.
* -mr (05/09/11)
*/
if (!priv->force_internal_extractors) {
g_debug (" Using libstreamanalyzer...");
tracker_topanalyzer_extract (task->file, statements, &content_type);
if (tracker_sparql_builder_get_length (statements) > 0) {
g_free (content_type);
tracker_sparql_builder_insert_close (statements);
*info_out = info;
return TRUE;
}
} else {
g_debug (" Using internal extractors ONLY...");
}
#endif /* HAVE_LIBSTREAMANALYZER */
if (task->mimetype && *task->mimetype) {
/* We know the mime */
mime_used = g_strdup (task->mimetype);
}
#ifdef HAVE_LIBSTREAMANALYZER
else if (content_type && *content_type) {
/* We know the mime from LSA */
mime_used = content_type;
g_strstrip (mime_used);
}
#endif /* HAVE_LIBSTREAMANALYZER */
else {
tracker_extract_info_unref (info);
return FALSE;
}
/* Now we have sanity checked everything, actually get the
* data we need from the extractors.
*/
if (mime_used) {
if (task->cur_func) {
TrackerSparqlBuilder *statements;
g_debug (" Using %s...", g_module_name (task->cur_module));
(task->cur_func) (info);
statements = tracker_extract_info_get_metadata_builder (info);
items = tracker_sparql_builder_get_length (statements);
if (items > 0) {
tracker_sparql_builder_insert_close (statements);
g_debug ("Done (%d items)", items);
task->success = TRUE;
}
}
g_free (mime_used);
}
if (items == 0) {
g_debug ("No extractor or failed");
tracker_extract_info_unref (info);
info = NULL;
}
*info_out = info;
return (items > 0);
}
