[tracker] Added more information about module use to tracker-extract

[Jürg Billeter <juergbi src gnome org>]

commit d8d8686df9a165398bee595e9310e296c6cfd9d5
Author: Martyn Russell <martyn imendio com>
Date:   Tue Apr 21 17:32:04 2009 +0100

    Added more information about module use to tracker-extract
    
    * src/tracker-extract/tracker-extract.c: Added debugging so we know
      which modules are loaded on start up and to know which modules are
      used to extract content on a per file basis to help when we have
      multiple extractors handling the same mime types.
    
    * src/tracker-extract/tracker-main.c: Change logging format when
      running stand alone and not writing to a log file and don't log
      everything, only log based on the verbosity setting.
---
 src/tracker-extract/tracker-extract.c |  136 +++++++++++++++++++++++----------
 src/tracker-extract/tracker-main.c    |   60 ++++++++++++++-
 2 files changed, 154 insertions(+), 42 deletions(-)

diff --git a/src/tracker-extract/tracker-extract.c b/src/tracker-extract/tracker-extract.c
index 4c2fd5b..e8c6746 100644
--- a/src/tracker-extract/tracker-extract.c
+++ b/src/tracker-extract/tracker-extract.c
@@ -40,9 +40,15 @@
 #define TRACKER_EXTRACT_GET_PRIVATE(obj) (G_TYPE_INSTANCE_GET_PRIVATE ((obj), TRACKER_TYPE_EXTRACT, TrackerExtractPrivate))
 
 typedef struct {
-	GArray *extractors;
+	GArray *specific_extractors;
+	GArray *generic_extractors;
 } TrackerExtractPrivate;
 
+typedef struct { 
+	const GModule *module;
+	const TrackerExtractData *edata; 
+}  ModuleData;
+
 static void tracker_extract_finalize (GObject *object);
 
 G_DEFINE_TYPE(TrackerExtract, tracker_extract, G_TYPE_OBJECT)
@@ -78,7 +84,8 @@ tracker_extract_finalize (GObject *object)
 	tracker_topanalyzer_shutdown ();
 #endif
 
-	g_array_free (priv->extractors, TRUE);
+	g_array_free (priv->specific_extractors, TRUE);
+	g_array_free (priv->generic_extractors, TRUE);
 
 	G_OBJECT_CLASS (tracker_extract_parent_class)->finalize (object);
 }
@@ -102,7 +109,7 @@ tracker_extract_new (void)
 	GDir *dir;
 	GError *error;
 	const gchar *name;
-	GArray *extractors;
+	GArray *specific_extractors;
 	GArray *generic_extractors;
 
 	if (!g_module_supported ()) {
@@ -110,33 +117,25 @@ tracker_extract_new (void)
 		return NULL;
 	}
 
-	extractors = g_array_sized_new (FALSE,
-					TRUE,
-					sizeof (TrackerExtractData),
-					10);
-
-	/* This array is going to be used to store
-	 * temporarily extractors with mimetypes such as "audio / *"
-	 */
-	generic_extractors = g_array_sized_new (FALSE,
-						TRUE,
-						sizeof (TrackerExtractData),
-						10);
-
 	error = NULL;
 	dir = g_dir_open (MODULESDIR, 0, &error);
 
 	if (!dir) {
 		g_error ("Error opening modules directory: %s", error->message);
 		g_error_free (error);
-		g_array_free (generic_extractors, TRUE);
-		g_array_free (extractors, TRUE);
 		return NULL;
 	}
 
+	specific_extractors = g_array_new (FALSE,
+					   TRUE,
+					   sizeof (ModuleData));
+
+	generic_extractors = g_array_new (FALSE,
+					  TRUE,
+					  sizeof (ModuleData));
+
 	while ((name = g_dir_read_name (dir)) != NULL) {
 		TrackerExtractDataFunc func;
-		TrackerExtractData *data;
 		GModule *module;
 		gchar *module_path;
 
@@ -149,7 +148,9 @@ tracker_extract_new (void)
 		module = g_module_open (module_path, G_MODULE_BIND_LOCAL);
 
 		if (!module) {
-			g_warning ("Could not load module '%s': %s", name, g_module_error ());
+			g_warning ("Could not load module '%s': %s", 
+				   name, 
+				   g_module_error ());
 			g_free (module_path);
 			continue;
 		}
@@ -157,13 +158,23 @@ tracker_extract_new (void)
 		g_module_make_resident (module);
 
 		if (g_module_symbol (module, "tracker_get_extract_data", (gpointer *) &func)) {
-			data = (func) ();
+			ModuleData mdata;
+		
+			mdata.module = module;
+			mdata.edata = (func) ();
 
-			for (; data->mime; data++) {
-				if (strchr (data->mime, '*') != NULL) {
-					g_array_append_val (generic_extractors, *data);
+			g_message ("Adding extractor:'%s' with:",
+				   g_module_name ((GModule*) mdata.module));
+
+			for (; mdata.edata->mime; mdata.edata++) {
+				if (G_UNLIKELY (strchr (mdata.edata->mime, '*') != NULL)) {
+					g_message ("  Generic  match for mime:'%s'",
+						   mdata.edata->mime);
+					g_array_append_val (generic_extractors, mdata);
 				} else {
-					g_array_append_val (extractors, *data);
+					g_message ("  Specific match for mime:'%s'",
+						   mdata.edata->mime);
+					g_array_append_val (specific_extractors, mdata);
 				}
 			}
 		}
@@ -173,20 +184,13 @@ tracker_extract_new (void)
 
 	g_dir_close (dir);
 
-	/* Append the generic extractors at the end of
-	 * the list, so the specific ones are used first
-	 */
-	g_array_append_vals (extractors,
-			     generic_extractors->data,
-			     generic_extractors->len);
-	g_array_free (generic_extractors, TRUE);
-
 	/* Set extractors */
 	object = g_object_new (TRACKER_TYPE_EXTRACT, NULL);
 
 	priv = TRACKER_EXTRACT_GET_PRIVATE (object);
 
-	priv->extractors = extractors;
+	priv->specific_extractors = specific_extractors;
+	priv->generic_extractors = generic_extractors;
 
 	return object;
 }
@@ -303,24 +307,66 @@ get_file_metadata (TrackerExtract *extract,
 	 */
 	if (mime_used) {
 		TrackerExtractPrivate *priv;
-		TrackerExtractData *data;
 		guint i;
 
 		priv = TRACKER_EXTRACT_GET_PRIVATE (extract);
 
-		for (i = 0; i < priv->extractors->len; i++) {
-			data = &g_array_index (priv->extractors, TrackerExtractData, i);
+		for (i = 0; i < priv->specific_extractors->len; i++) {
+			const TrackerExtractData *edata;
+			ModuleData mdata;
+
+			mdata = g_array_index (priv->specific_extractors, ModuleData, i);
+			edata = mdata.edata;
+
+			if (g_pattern_match_simple (edata->mime, mime_used)) {
+				gint items;
+
+				tracker_dbus_request_comment (request_id,
+							      "  Extracting with module:'%s'",
+							      g_module_name ((GModule*) mdata.module));
+
+				(*edata->extract) (uri, statements);
 
-			if (g_pattern_match_simple (data->mime, mime_used)) {
-				(*data->extract) (uri, statements);
+				items = statements->len;
 
-				if (statements->len == 0) {
+				tracker_dbus_request_comment (request_id,
+							      "  Found %d metadata items",
+							      items);
+				if (items == 0) {
 					continue;
 				}
 
+				g_free (mime_used);
+				g_free (content_type);
+
+				return statements;
+			}
+		}
+
+		for (i = 0; i < priv->generic_extractors->len; i++) {
+			const TrackerExtractData *edata;
+			ModuleData mdata;
+
+			mdata = g_array_index (priv->generic_extractors, ModuleData, i);
+			edata = mdata.edata;
+
+			if (g_pattern_match_simple (edata->mime, mime_used)) {
+				gint items;
+
+				tracker_dbus_request_comment (request_id,
+							      "  Extracting with module:'%s'",
+							      g_module_name ((GModule*) mdata.module));
+				
+				(*edata->extract) (uri, statements);
+
+				items = statements->len;
+
 				tracker_dbus_request_comment (request_id,
 							      "  Found %d metadata items",
-							      statements->len);
+							      items);
+				if (items == 0) {
+					continue;
+				}
 
 				g_free (mime_used);
 				g_free (content_type);
@@ -355,12 +401,20 @@ tracker_extract_get_metadata_by_cmdline (TrackerExtract *object,
 
 	g_return_if_fail (uri != NULL);
 
+	tracker_dbus_request_new (request_id,
+				  "Command line request to extract metadata, "
+				  "uri:'%s', mime:%s",
+				  uri,
+				  mime);
+
 	/* NOTE: Don't reset the timeout to shutdown here */
 	statements = get_file_metadata (object, request_id, uri, mime);
 
 	if (statements) {
 		statements_free (statements);
 	}
+
+	tracker_dbus_request_success (request_id);
 }
 
 void
diff --git a/src/tracker-extract/tracker-main.c b/src/tracker-extract/tracker-main.c
index e8e4477..c140ee6 100644
--- a/src/tracker-extract/tracker-main.c
+++ b/src/tracker-extract/tracker-main.c
@@ -222,6 +222,37 @@ initialize_signal_handler (void)
 #endif /* G_OS_WIN32 */
 }
 
+static void
+log_handler (const gchar    *domain,
+	     GLogLevelFlags  log_level,
+	     const gchar    *message,
+	     gpointer	     user_data)
+{
+	if (((log_level & G_LOG_LEVEL_DEBUG) && verbosity < 3) ||
+	    ((log_level & G_LOG_LEVEL_INFO) && verbosity < 2) ||
+	    ((log_level & G_LOG_LEVEL_MESSAGE) && verbosity < 1)) {
+		return;
+	}
+
+	switch (log_level) {
+	case G_LOG_LEVEL_WARNING:
+	case G_LOG_LEVEL_CRITICAL:
+	case G_LOG_LEVEL_ERROR:
+	case G_LOG_FLAG_RECURSION:
+	case G_LOG_FLAG_FATAL:
+		g_fprintf (stderr, "%s\n", message);
+		fflush (stderr);
+		break;
+	case G_LOG_LEVEL_MESSAGE:
+	case G_LOG_LEVEL_INFO:
+	case G_LOG_LEVEL_DEBUG:
+	case G_LOG_LEVEL_MASK:
+		g_fprintf (stdout, "%s\n", message);
+		fflush (stdout);
+		break;
+	}	
+}
+
 int
 main (int argc, char *argv[])
 {
@@ -229,6 +260,8 @@ main (int argc, char *argv[])
 	GError         *error = NULL;
 	TrackerConfig  *config;
 	gchar          *log_filename;
+	gboolean        stand_alone = FALSE;
+	guint           log_handler_id = 0;
 
 	bindtextdomain (GETTEXT_PACKAGE, LOCALEDIR);
 	bind_textdomain_codeset (GETTEXT_PACKAGE, "UTF-8");
@@ -283,6 +316,24 @@ main (int argc, char *argv[])
 
 	setlocale (LC_ALL, "");
 
+	/* Set conditions when we use stand alone settings */
+	stand_alone |= filename != NULL;
+
+	if (stand_alone) {
+		/* Set log handler for library messages */
+		log_handler_id = g_log_set_handler (NULL,
+						    G_LOG_LEVEL_MASK | G_LOG_FLAG_FATAL,
+						    log_handler,
+						    NULL);
+		
+		g_log_set_default_handler (log_handler, NULL);
+
+		/* Set the default verbosity if unset */
+		if (verbosity == -1) {
+			verbosity = 3;
+		}
+	}
+
 	if (filename) {
 		TrackerExtract *object;
 		GFile *file;
@@ -302,7 +353,7 @@ main (int argc, char *argv[])
 		g_object_unref (file);
 		g_free (uri);
 
-		return EXIT_SUCCESS;
+		goto done;
 	}
 
 	config = tracker_config_new ();
@@ -361,5 +412,12 @@ main (int argc, char *argv[])
 	g_free (log_filename);
 	g_object_unref (config);
 
+done:
+
+	if (log_handler_id != 0) {
+		/* Unset log handler */
+		g_log_remove_handler (NULL, log_handler_id);
+	}
+
 	return EXIT_SUCCESS;
 }