Media Types¶
In the future, there will be all sorts of media types you can enable, but in the meanwhile there are six additional media types: video, audio, raw image, ASCII art, STL/3D models, PDF and Document.
First, you should probably read “Configuring MediaGoblin” to make sure you know how to modify the MediaGoblin config file.
Enabling Media Types¶
Note
Media types are now plugins
Media types are enabled in your MediaGoblin configuration file.
Most media types require additional dependencies that you will have to install. You will find descriptions on how to satisfy the requirements of each media type below.
To enable a media type, add the the media type under the [plugins] section
in you mediagoblin.ini. For example, if your system supported image
and video media types, then it would look like this:
[plugins]
[[mediagoblin.media_types.image]]
[[mediagoblin.media_types.video]]
Note that after enabling new media types, you must run dbupdate. If you have deployed MediaGoblin as an unprivileged user as described in “Considerations for Production Deployments”, you’ll first need to switch to this account:
sudo su mediagoblin --shell=/bin/bash
$ cd /srv/mediagoblin.example.org/mediagoblin
Now run dbupdate:
$ ./bin/gmg dbupdate
If you are running an active site, depending on your server configuration, you may need to stop it first (and it’s certainly a good idea to restart it after the update).
How does MediaGoblin decide which media type to use for a file?¶
MediaGoblin has two methods for finding the right media type for an uploaded file. One is based on the file extension of the uploaded file; every media type maintains a list of supported file extensions. The second is based on a sniffing handler, where every media type may inspect the uploaded file and tell if it will accept it.
The file-extension-based approach is used before the sniffing-based approach, if the file-extension-based approach finds a match, the sniffing-based approach will be skipped as it uses far more processing power.
Configuring Media Types¶
Each media type has a config_spec.ini file with configurable
options and comments explaining their intended side effect. For
instance the video media type configuration can be found in
mediagoblin/media_types/video/config_spec.ini.
Audio¶
To enable audio, install the GStreamer and python-gstreamer bindings (as well as whatever GStreamer plugins you want, good/bad/ugly):
# Debian and co.
sudo apt install python3-gst-1.0 gstreamer1.0-plugins-{base,bad,good,ugly} \
gstreamer1.0-libav
# Fedora and co.
sudo dnf install gstreamer1-plugins-{base,bad-free,good,ugly-free}
Note
MediaGoblin previously generated spectrograms for uploaded audio. This feature has been removed due to incompatibility with Python 3. We may consider re-adding this feature in the future.
Add [[mediagoblin.media_types.audio]] under the [plugins] section in your
mediagoblin.ini and update MediaGoblin:
$ ./bin/gmg dbupdate
Restart MediaGoblin (and Celery if applicable). You should now be able to upload and listen to audio files!
Video¶
To enable video, first install GStreamer and the python-gstreamer bindings (as well as whatever GStreamer extensions you want, good/bad/ugly):
# Debian and co.
sudo apt install python3-gi gstreamer1.0-tools gir1.2-gstreamer-1.0 \
gir1.2-gst-plugins-base-1.0 gstreamer1.0-plugins-{good,bad,ugly} \
gstreamer1.0-libav python3-gst-1.0
Note
We unfortunately do not have working installation instructions for Fedora and co. Some incomplete information is available on the Hacking Howto wiki page
Add [[mediagoblin.media_types.video]] under the [plugins] section in
your mediagoblin.ini and restart MediaGoblin.
Run:
$ ./bin/gmg dbupdate
Restart MediaGoblin (and Celery if applicable). Now you should be able to submit videos, and MediaGoblin should transcode them.
Note
You will likely need to increase the client_max_body_size setting in
Nginx to upload larger videos.
You almost certainly want to separate Celery from the normal paste process or your users will probably find that their connections time out as the video transcodes. To set that up, check out the “Considerations for Production Deployments” section of this manual.
Raw image¶
To enable raw image you need to install pyexiv2:
# Debian and co.
sudo apt install python3-pyexiv2
Add [[mediagoblin.media_types.raw_image]] under the [plugins]
section in your mediagoblin.ini and restart MediaGoblin.
Run:
$ ./bin/gmg dbupdate
Restart MediaGoblin (and Celery if applicable). Now you should be able to submit raw images, and MediaGoblin should extract the JPEG preview from them.
ASCII art¶
To enable ASCII art support, first install the chardet library, which is necessary for creating thumbnails of ASCII art:
$ ./bin/easy_install chardet
Next, modify your mediagoblin.ini. In the [plugins] section, add
[[mediagoblin.media_types.ascii]].
Run:
$ ./bin/gmg dbupdate
Restart MediaGoblin (and Celery if applicable). Now any .txt file you uploaded will be processed as ASCII art!
STL / 3D model support¶
To enable the “STL” 3D model support plugin, first make sure you have a recent Blender installed and available on your execution path. This feature has been tested with Blender 2.63. It may work on some earlier versions, but that is not guaranteed (and is surely not to work prior to Blender 2.5X).
Add [[mediagoblin.media_types.stl]] under the [plugins] section in your
mediagoblin.ini and restart MediaGoblin.
Run:
$ ./bin/gmg dbupdate
Restart MediaGoblin (and Celery if applicable). You should now be able to upload .obj and .stl files and MediaGoblin will be able to present them to your wide audience of admirers!
PDF and Document¶
To enable the “PDF and Document” support plugin, you need:
pdftocairo and pdfinfo for PDF only support.
unoconv with headless support to support converting LibreOffice supported documents as well, such as doc/ppt/xls/odf/odg/odp and more. For the full list see mediagoblin/media_types/pdf/processing.py, unoconv_supported.
All executables must be on your execution path.
To install this on Fedora:
sudo dnf install poppler-utils unoconv libreoffice-headless
Note: You can leave out unoconv and libreoffice-headless if you want only PDF support. This will result in a much smaller list of dependencies.
pdf.js relies on git submodules, so be sure you have fetched them:
$ git submodule init
$ git submodule update
- This feature has been tested on Fedora with:
poppler-utils-0.20.2-9.fc18.x86_64 unoconv-0.5-2.fc18.noarch libreoffice-headless-3.6.5.2-8.fc18.x86_64
It may work on some earlier versions, but that is not guaranteed.
Add [[mediagoblin.media_types.pdf]] under the [plugins] section in your
mediagoblin.ini and restart MediaGoblin.
Run:
$ ./bin/gmg dbupdate
Blog (HIGHLY EXPERIMENTAL)¶
MediaGoblin has a blog media type, which you might notice by looking through the docs! However, it is highly experimental. We have not security reviewed this, and it acts in a way that is not like normal blogs (the blog posts are themselves media types!).
So you can play with this, but it is not necessarily recommended yet for production use! :)
