Database schema¶

PyWacli persists everything through SQLAlchemy. SQLite is the default (WAL mode, foreign keys on), with PostgreSQL and MySQL also supported — see Configuration → database.

Tables¶

Table	Contents
`messages`	Text messages with sender metadata
`message_edits`	Edit history for messages
`statuses`	WhatsApp status / story updates
`reactions`	Emoji reactions
`media`	Media metadata (type, filename, path, sender)
`media_handshake`	Per-entry sync status (success / failure reason)
`conversations`	Unified conversation log

How it's used¶

messages powers the dashboard feed, contact list, and the conversation history loaded by AI Automate.
media + media_handshake back the media viewer: metadata in one, upload outcome in the other.
conversations is the unified per-chat log derived from the conversation.new event.

Choosing a backend¶

SQLitePostgreSQLMySQL

Zero setup, single file, great for one machine. Defaults to ~/.pywacli/pywacli.db when no path is configured. Runs with PRAGMA journal_mode=WAL and PRAGMA foreign_keys=ON.

Use for multi-user or server deployments. Connection is built as postgresql://user:password@host:port/name. Inserts use RETURNING to fetch new row ids.

Connection is built as mysql+pymysql://user:password@host:port/name.

Connection pooling

The engine is created with pool_pre_ping=True, so stale connections are detected and refreshed automatically — handy for long-running connect sessions against a remote database.