mirror of
https://github.com/greenart7c3/Amber.git
synced 2026-10-30 06:47:33 +00:00
greenart7c3
d1ce1dd1ed
Logs were capped at 1000 rows / 3 days and history at 2000 rows / 7 days per account. Relay lifecycle events (connecting, connected, sent, OK...) fill the log cap within hours, so logs disappeared almost immediately. The cap also didn't bound disk use under heavy signing: cleanup only ran once a day. - Merge repeated log entries: an entry identical to one logged in the last hour (same url, type, message) bumps a repeatCount on that row instead of inserting. Log DB v3 adds the column and a (url, time) index. - Replace the row caps with LogRetentionManager: delete rows past the retention period, then the oldest rows in batches while the database is over its budget (20 MB logs, 50 MB history). It runs from ClearLogsWorker and, throttled to once a minute, after writes via a Room InvalidationTracker observer, so bursts are trimmed right away. - Databases without auto-vacuum are switched to incremental auto-vacuum so deleted rows give space back to the filesystem. - New "Keep logs for" (default 7 days) and "Keep activity for" (default 30 days) settings in Security. Retention is read synchronously from SharedPreferences so a prune before the async settings load never applies a shorter default. - Logs screen, relay dialog and export show the repeat count.