When submitting a PDF file with unicode characters in the document info, mediagoblin fails with UnicodeDecodeError. E.g. a file containing

Creator: Microsoft\xc2\xae Word 2013

fails with

UnicodeDecodeError('ascii', 'Creator: Microsoft\xc2\xae Word 2013\n', 25, 26, 'ordinal not in range(128)')

at media_types/pdf/processing.py:210, where the output of pdfinfo is decoded with:

lines = [l.decode() for l in lines]

The problem is that decode uses the default string encoding, which is ascii in my case. It can be fixed by specifying UTF-8 encoding explicitly:

lines = [l.decode('utf-8', 'ignore') for l in lines]

However, I am not sure if pdfinfo always outputs the information in UTF-8, or it is document dependent.

The line was added by ​cda3055

The output of ./lazyserver.sh:

2014-10-13 09:27:09,900 DEBUG   [mediagoblin.processing.task] Processing <MediaEntry 10: doc>
2014-10-13 09:27:10,025 ERROR   [mediagoblin.processing.task] An unhandled exception was raised while processing <MediaEntry 10: doc>
2014-10-13 09:27:10,025 WARNING [mediagoblin.processing] No idea what happened here, but it failed: UnicodeDecodeError('ascii', 'Creator:        Microsoft\xc2\xae Word 2013\n', 25, 26, 'ordinal not in range(128)')
2014-10-13 09:27:10,153 WARNING [mediagoblin.processing] No idea what happened here, but it failed: UnicodeDecodeError('ascii', 'Creator:        Microsoft\xc2\xae Word 2013\n', 25, 26, 'ordinal not in range(128)')
[2014-10-13 09:27:10 +0000] [4679] [ERROR] Error handling request
Traceback (most recent call last):
  File "/mnt/data/mediagoblin/env/local/lib/python2.7/site-packages/gunicorn-19.1.1-py2.7.egg/gunicorn/workers/sync.py", line 93, in handle
    self.handle_request(listener, req, client, addr)
  File "/mnt/data/mediagoblin/env/local/lib/python2.7/site-packages/gunicorn-19.1.1-py2.7.egg/gunicorn/workers/sync.py", line 134, in handle_request
    respiter = self.wsgi(environ, resp.start_response)
  File "/mnt/data/mediagoblin/mediagoblin/app.py", line 268, in __call__
    return self.call_backend(environ, start_response)
  File "/mnt/data/mediagoblin/env/local/lib/python2.7/site-packages/Werkzeug-0.9.6-py2.7.egg/werkzeug/wsgi.py", line 588, in __call__
    return self.app(environ, start_response)
  File "/mnt/data/mediagoblin/mediagoblin/app.py", line 245, in call_backend
    response = controller(request)
  File "/mnt/data/mediagoblin/mediagoblin/decorators.py", line 46, in wrapper
    return controller(request, *args, **kwargs)
  File "/mnt/data/mediagoblin/mediagoblin/decorators.py", line 73, in new_controller_func
    return controller(request, *args, **kwargs)
  File "/mnt/data/mediagoblin/mediagoblin/decorators.py", line 46, in wrapper
    return controller(request, *args, **kwargs)
  File "/mnt/data/mediagoblin/mediagoblin/decorators.py", line 73, in new_controller_func
    return controller(request, *args, **kwargs)
  File "/mnt/data/mediagoblin/mediagoblin/decorators.py", line 102, in wrapper
    return controller(request, *args, **kwargs)
  File "/mnt/data/mediagoblin/mediagoblin/submit/views.py", line 69, in submit_start
    urlgen=request.urlgen)
  File "/mnt/data/mediagoblin/mediagoblin/submit/lib.py", line 201, in submit_media
    run_process_media(entry, feed_url)
  File "/mnt/data/mediagoblin/mediagoblin/submit/lib.py", line 252, in run_process_media
    task_id=entry.queued_task_id)
  File "/mnt/data/mediagoblin/env/local/lib/python2.7/site-packages/celery-3.1.16-py2.7.egg/celery/app/task.py", line 547, in apply_async
    link=link, link_error=link_error, **options)
  File "/mnt/data/mediagoblin/env/local/lib/python2.7/site-packages/celery-3.1.16-py2.7.egg/celery/app/task.py", line 739, in apply
    request=request, propagate=throw)
  File "/mnt/data/mediagoblin/env/local/lib/python2.7/site-packages/celery-3.1.16-py2.7.egg/celery/app/trace.py", line 354, in eager_trace_task
    uuid, args, kwargs, request)
  File "/mnt/data/mediagoblin/env/local/lib/python2.7/site-packages/celery-3.1.16-py2.7.egg/celery/app/trace.py", line 253, in trace_task
    I, R, state, retval = on_error(task_request, exc, uuid)
  File "/mnt/data/mediagoblin/env/local/lib/python2.7/site-packages/celery-3.1.16-py2.7.egg/celery/app/trace.py", line 240, in trace_task
    R = retval = fun(*args, **kwargs)
  File "/mnt/data/mediagoblin/mediagoblin/processing/task.py", line 99, in run
    processor.process(**reprocess_info)
  File "/mnt/data/mediagoblin/mediagoblin/media_types/pdf/processing.py", line 413, in process
    self.extract_pdf_info()
  File "/mnt/data/mediagoblin/mediagoblin/media_types/pdf/processing.py", line 339, in extract_pdf_info
    pdf_info_dict = pdf_info(self.pdf_filename)
  File "/mnt/data/mediagoblin/mediagoblin/media_types/pdf/processing.py", line 210, in pdf_info
    lines = [l.decode() for l in lines]
UnicodeDecodeError: 'ascii' codec can't decode byte 0xc2 in position 25: ordinal not in range(128)

As discussed on IRC, some Web browsers (Apple Safari and Microsoft Internet Explorer) don't support WebM VP8+Vorbis media. Apparently Xiph.Org offers ​Apple QuickTime Compenents and ​Microsoft DirectShow Filters that should enable these browsers to play WebM VP8+Vorbis media.

MediaGoblin could recommend users of these browsers to install these packages to play media on MediaGoblin sites.

Another solution may be to support encoding in Ogg Theora+Vorbis (as requested in #567) and use ​Cortado to play media in these browsers (as well as older browsers that don't support HTML5 <video>).

With this change, visitors to the MediaGoblin site can try a live instance.

For context, the demo is hosted on a demo server run by Sandstorm Development Group, which runs the FLOSS Sandstorm software. The path component in the URL is the app ID, so when new versions are uploaded to that server, this link will automatically update to install those new versions.

The Sandstorm "port" of MediaGoblin has the source code available for its fork here: ​https://github.com/jparyani/mediagoblin/ and links to it in the footer, as a general good practice and also as a way of complying with the AGPLv3.

It uses target=_blank because, in my manual testing, after clicking this link and playing with the MediaGoblin demo, the default thing to do is to close the tab/window. It would be nice if the MediaGoblin website were still waiting for me so I could be reminded that I tried the demo to convince myself that I want to actually install MediaGoblin.