DeetsOTD

Status: working and in regular use · Platform: Python — Discord bot, CLI, and a local web harness

Reads every hyperlink out of a Discord channel and writes them to Markdown + CSV.

Three ways to trigger the same core (linkscan.py):

Entry point	Use it for
bot.py	/links slash command inside Discord; DMs you the files
scan.py	one-off run from the terminal
harness.py	device-side local web/API trigger (browser form or curl)

Output lands in exports/ as links_<channel>_<timestamp>.md (deduped, grouped, with author/date/jump links) and .csv (every record, for spreadsheets).

Setup

pip install -r requirements.txt
cp config.example.py config.py        # set DEFAULT_CHANNEL_ID / ALLOWED_CHANNEL_IDS
cp .env.example .env                  # paste your DISCORD_TOKEN

In the Discord developer portal: your app → Bot → enable Message Content Intent. Invite the bot to your server with the Read Message History permission.

Channel IDs: enable Developer Mode in Discord (Settings → Advanced), then right-click a channel → Copy ID.

Run

python scan.py 123456789012345678          # one-off, specific channel
python scan.py 123456789012345678 --limit 500

python bot.py                              # then use /links in Discord

python harness.py                          # open http://127.0.0.1:8077

Per-user breakdown

To see what each user has sent in a channel — every user's links grouped together, with repeats per user collapsed — add --by-user:

python scan.py 123456789012345678 --by-user                 # CSV (default)
python scan.py 123456789012345678 --by-user --format tsv    # TSV
python scan.py 123456789012345678 --by-user --format xml    # XML

This writes exports/by_user_<channel>_<timestamp>.<ext> and prints a links-per-user tally. CSV/TSV are flat tables sorted so each user's links sit together (author, author_id, link_count, url, …); XML nests <link> elements under one <user> element each. Users are ordered by most recent activity.

The journal + SongOTD archive

Treat the channel as a Song-of-the-Day journal that accumulates over time in a local SQLite DB (storage.db, gitignored):

python scan.py <channel_id> --archive   # fold posts into the journal
python scan.py <channel_id> --enrich    # + resolve each song via iTunes

--archive records one post per song link a user typed (display name, Discord user ID, the message's date/time, URL, and detected service — apple_music / spotify / youtube / …). It's idempotent on (message_id, url), so re-scanning never double-counts.

--enrich (implies --archive) resolves each post against the free iTunes APIs into a songs catalog — title, artist, album, genre, release date, artwork, 30s preview, store link. Apple Music links resolve by their track id (direct); Spotify/YouTube/etc. resolve via oEmbed title → iTunes search (search). Posts that don't resolve (e.g. a non-music link) stay in the journal unenriched.

Schema lives in storage.py; enrichment in enrich.py.

Website export (songs.json)

Feed the Deets.Solutions Song-of-the-Day page a static, display-only JSON straight from the journal:

python scan.py <channel_id> --enrich --web \
    --web-out ../deetssolutions/sotd/songs.json

--web (implies --archive) writes {generated_at, channel, count, songs[]}. Each song carries only display fields — track/artist/album, genre, artwork, 30s preview, Apple Music link, duration_sec, dates, poster name — and drops the private Discord ids (message id, author id, jump link) so the file is safe on a public site. Numbers come out as real numbers (or null) so the page can sort/filter client-side. --web-out sets the path (default exports/songs.json); point it at the site repo and commit the result.

Regenerating from an existing storage.db needs no Discord/iTunes:

python -c "import storage; storage.export_web(<channel_id>, out_path='../deetssolutions/sotd/songs.json')"

Personal, single-user. The harness binds to localhost with no auth — don't expose it.

License

MIT — see LICENSE.