diff options
Diffstat (limited to 'README')
| -rw-r--r-- | README | 48 |
1 files changed, 0 insertions, 48 deletions
@@ -1,48 +0,0 @@ -roka: stream directory of audiobooks to podcasting apps via RSS - -demo (no audio): https://demo.jordan.im/roka/ -iOS podcast app: https://demo.jordan.im/roka/apple-podcasts.png - -installation ------------- - -a) copy and populate app.cfg and uwsgi.ini from examples - -b) install python dependencies flask and uwsgi - - $ pip install --user flask uwsgi - -c) run roka.py with --scan to populate audiobook JSON cache (can be re-run to - update cache upon download of new books) - - $ ./roka.py --scan - -d) execute uwsgi.sh to start the server - - $ ./uwsgi.sh - -design decisions ----------------- - -1. directories contained within config:ROOT_PATH are marked as audiobooks if and - only if they contain at least one MP3 file - -2. audiobooks are uniquely identifiable by the collective hash of each MP3 file - contained in the audiobook directory - - pro: if the directory structure is changed or files are moved, RSS/download - link integrity is maintained, preserving app-side listening progress and - history - - con: each MP3 file is hashed, which can be slow on spinning rust w/ large - collections - -3. XML pubDate and list order is derived from MP3 track attributes; if not - present or duplicates exist, tracks are sorted alphanumerically - - if a book's track numbers are unique but incorrect, a preference for filename - sort can be established by creating an 'ignore_tracknum' file in the - audiobook's path - -4. no rebuild endpoint exists; cache-affecting routines are run externally via - rebuild.py |