A Web scraper for InfoQ.

InfoQ hosts a lot of great presentations, unfortunately it is not possible to watch them outside of the browser or if you do not have Flash installed. The video cannot simply be downloaded because the audio stream and the slide stream are not in the same media. By downloading the video you only get the audio track and a video of the presenter but you don't get the slide.

infoqscraper allows you to:

• list and search for presentations
• download the resources (video, audio track, slides)
• build a movie including the slides and the audio track from the resources

Only Python 2.6 or later is currently supported since PIL is not yet available for Python 3.

Install

Infoqscraper releases can be installed using pip and PyPI.

    pip install [--user] infoqscraper

The infoqscraper executable should be available in your path.

If the command cannot be found, you have to add the installation directory (usually $HOME/.local/bin with --user) to the PATH environment variable or specify the full path of the command.

CLI usage

Overview:

    infoqscraper [global options] module [module options] command [command options]

presentation module

This module allows to list, search and download presentation from the website

presentation list

The following command displays the 20 latest presentations.

    infoqscraper presentation list -n 20

You can also search for a specific topic:

    infoqscraper presentation list -p agile

By default at most 10 hits are returned and only the latest 80 entries are fetched from the website. The -n and -m options can be used to display more hits or to fetch more entries. Please be a good web citizen and avoid to set -m to a very large value. Using a search engine or the website could be a better idea.

presentation download

The following command automatically downloads the presentation and creates a movie file named Distributed-Systems-with-ZeroMQ-and-gevent.avi in the current directory.

    infoqscraper presentation download Distributed-Systems-with-ZeroMQ-and-gevent

cache module

This module allows to manage the disk cache.

Disk cache is disabled by default and is only useful if you plan to hack on infoqscraper. To avoid fetching the same resource several time from the server, all fetched resources can be put in cache. This is done by setting the -c global option. The resources are stored in the XDG_CACHE_DIR/infoqscraper directory.

The following command show the size of the cache:

    infoqscraper cache size

And the clear command allows to wipe the cache

    infoqscraper cache clear

Dependencies

infoqscraper relies on theses 3rd party tools:

1. ffmpeg
2. swftools
3. rtmpdump

If theses tools are not in the PATH, their location can be specified. Use infoqscraper presentation download -h to learn more.

The following python packages are required:

1. BeautifulSoup4
2. PIL

For normal users, they are automatically installed by pip. However, if you plan to hack infoqscraper you will most likely install them using the requirements.txt file:

    pip install -r requirements.txt

Compatibility

infoqscraper is known to work on:

• Arch Linux
• Fedora 17, 18, 19
• Ubuntu 12.04.2 LTS

ffmpeg versions from 0.7 to 2+ are supported. Users of Debian Squeeze or Ubuntu 10.04 LTS must use a newer ffmpeg release than the one provided by their distro. ffmpeg 0.5 is too old to be supported (help wanted).

Arch linux

Packages python2, swftools, rtmpdump and ffmpeg must be installed.

Ubuntu 12.04 LTS

Packages ffmpeg, rtmpdump and libavcodec-extra-53 must be installed.

swftools is not available and must installed manually.

Ubuntu 12.10 and later

Packages ffmpeg, rtmpdump, swftools and libavcodec-extra-53 must be installed.

Fedora

Packages ffmpeg, swftools and rtmpdump from rpmfusion must be installed.

rtmpdump from Fedora 19 is currently broken. You have to download rtmpdump source code and compile it.

Help

Feel free to contact me if you have any question or feature request.

If you find this project useful, any contribution or feedback is welcome. If you are not a developer, improving the packaging, the documentation or fixing my broken English could be a good start.