Fixing dates when uploaded in a video to a mediagoblin instance.

There were some "serializing to json strings" issues.  They should be
fixed now... much more careful whitelist and cleaning of the video
"tags" metadata out of gstreamer.

This commit sponsored by Aimee Sullivan.  Thanks!

1 files changed, 32 insertions, 1 deletions

diff --git a/mediagoblin/media_types/video/processing.py b/mediagoblin/media_types/video/processing.py
index d6439b78..ff2c94a0 100644
--- a/mediagoblin/media_types/video/processing.py
+++ b/mediagoblin/media_types/video/processing.py
@@ -16,6 +16,7 @@
 
 from tempfile import NamedTemporaryFile
 import logging
+import datetime
 
 from mediagoblin import mg_globals as mgg
 from mediagoblin.processing import \
@@ -165,7 +166,7 @@ def store_metadata(media_entry, metadata):
          for key in [
                  "videoheight", "videolength", "videowidth",
                  "audiorate", "audiolength", "audiochannels", "audiowidth",
-                 "mimetype", "tags"]
+                 "mimetype"]
          if key in metadata])
 
     # We have to convert videorate into a sequence because it's a
@@ -175,6 +176,36 @@ def store_metadata(media_entry, metadata):
         videorate = metadata["videorate"]
         stored_metadata["videorate"] = [videorate.num, videorate.denom]
 
+    # Also make a whitelist conversion of the tags.
+    if "tags" in metadata:
+        tags_metadata = metadata['tags']
+
+        # we don't use *all* of these, but we know these ones are
+        # safe...
+        tags = dict(
+            [(key, tags_metadata[key])
+             for key in [
+                     "application-name", "artist", "audio-codec", "bitrate",
+                     "container-format", "copyright", "encoder", 
+                     "encoder-version", "license", "nominal-bitrate", "title",
+                     "video-codec"]
+             if key in tags_metadata])
+        if 'date' in tags_metadata:
+            date = tags_metadata['date']
+            tags['date'] = "%s-%s-%s" % (
+                date.year, date.month, date.day)
+
+        # TODO: handle timezone info; gst.get_time_zone_offset +
+        #   python's tzinfo should help
+        if 'datetime' in tags_metadata:
+            dt = tags_metadata['datetime']
+            tags['datetime'] = datetime.datetime(
+                dt.get_year(), dt.get_month(), dt.get_day(), dt.get_hour(),
+                dt.get_minute(), dt.get_second(),
+                dt.get_microsecond()).isoformat()
+    
+        metadata['tags'] = tags
+
     # Only save this field if there's something to save
     if len(stored_metadata):
         media_entry.media_data_init(