- switch mediatitles.sh to Python instead of bash
- handle redirected images
- mediatitles - figure out what 'logo' is in each language
- build.sh - take list of languages from env
Lots of logos don't have 'logo' in the name. This is the outline of how to get them from the infobox section of a wikipedia page.
- Download .bz
- Unpack .bz
- Process each file
- Scan for Infobox
- Get logo svg
- Get website
- Get Article name
- Save (article title, website, logo.svg)
- Figure out how to use the multi-stream versions