An atom feed with non-ASCII characters throws a UnicodeEncodeError.

As log messages for uploads of pdf and audio files are sometimes confusing (#5498), it would be great to see the actual celery output for uploads in the Media processing panel.

Looks like the pdf plugin can't handle PDF files.

1. Setup production environment:
2. Enable PDF support: ​https://mediagoblin.readthedocs.io/en/latest/siteadmin/media-types.html#pdf-and-document

sudo apt-get install -y poppler-utils echo "mediagoblin.media_types.pdf" >> mediagoblin_local.ini git submodule init && git submodule update ./bin/gmg dbupdate sudo service mediagoblin-paster restart

3. try to upload attached sample PDF

2017-02-10 00:49:20,798 INFO    [mediagoblin.init.plugins] Importing plugin module: mediagoblin.media_types.pdf
2017-02-10 01:20:43,799 INFO    [mediagoblin.media_types] No plugins handled extension .pdf
2017-02-10 01:20:43,800 INFO    [mediagoblin.media_types] No plugins using two-step checking found

From mediagoblin/media_types/init.py

def sniff_media(media_file, filename):
    '''
    Iterate through the enabled media types and find those suited
    for a certain file.
    '''
    # copy the contents to a .name-enabled temporary file for further checks
    # TODO: there are cases when copying is not required
    tmp_media_file = tempfile.NamedTemporaryFile()
    shutil.copyfileobj(media_file, tmp_media_file)
    media_file.seek(0)
    tmp_media_file.seek(0)
    try:
        return type_match_handler(tmp_media_file, filename)
    except TypeNotFound as e:
        _log.info('No plugins using two-step checking found')

Log messages look similar to #5348.