Archivebox

This is a bugfix release, many parts of the parsing process have been improved or fixed.

- Shaarli RSS export support
- Fix issues with plain text link parsing including quotes, whitespace, and closing tags in URLs
- add USER_AGENT to archive.org submissions so they can track archivebox usage
- remove all icons similar to archive.org branding from archive UI
- hide some of the noisier youtubedl and wget errors
- set permissions on youtubedl media folder
- fix chrome data dir incorrect path and quoting
- better chrome binary finding
- show which parser is used when importing links, show progress when fetching titles

This is a feature-packed release, so it's likely to be a little buggier than usual!

New features:

- ability to load any plain text list of links (also the new fallback for all parses)
- WARC file saving via wget: `FETCH_WARC=True`
- Git repository downloading with git clone: `FETCH_GIT=True` `GIT_DOMAINS=github.com,gitlab.com,bitbucket.org`
- Media downloading with youtube-dl: `FETCH_MEDIA=True` `MEDIA_TIMEOUT=36000`

Bugfixes:
- autodetect the correct chromium binary in almost all cases
- create browser history export folder automatically
- higher allowed timestamp precision

New logo:

<img src="https://i.imgur.com/JffzvjI.png" width="300px"/>

No major changes other than the name.

Warning: Running this version will move the old `html/` output folder to the new location: `output/`.

Changes:

- entirely new folder structure & code layout
- moved scripts into `bin/` folder, symlinked `setup` and `archive` for backwards-compatibility
- removed TEMPLATE_INDEX* config options, just symlink the files in `templates/` to your custom versions
- added support for `./bin/export-browser-history` JSON imports of browsing history from Chrome and Firefox

New Features:
- Support for parsing links from RSS feeds
- Support for specifying a URL as well as local file paths: `./archive.py https://example.com/path/to/rss/feed.xml`
- Support for `--user-data-dir` for archiving restricted sites with chrome headless
- Simple & Fancy HTML & JSON indexes for each individual link
- Archive attempt history stored in link `index.json`

Improvements:
- Append to existing archive instead of overwriting the index each time
- Reduced unnecessary config options, it should "just work"
- Smartly dedupe and cleanup messy archive folders
- Massively cleaned up codebase

- refactor codebase into separate files
- check for minimum python version before running
- fix utf-8 encoding errors when writing index.html
- make index easier to customize with `templates/` folder
- WIP audio & video downloading with youtube-dl