Optimizations

Reference: Optimizations & Improvements

Ideas for making DailyDriver faster, cleaner, and more maintainable. Moved here from the former root OPTIMIZATIONS.md.

---

🔴 High Priority (immediate impact, low risk)

Enable SQLite WAL mode

• One‑line change: PRAGMA journal_mode=WAL; after opening the DB.
• Write‑Ahead Logging makes writes faster and reduces locking, even for a single‑user app.
• Effort: One line in database.py. Benefit: Slightly faster writes, safer file handling.

---

🟡 Medium Priority (add when convenient)

Smarter _update_complete_until

• After logging a qada far in the past, avoid scanning all intermediate days.
• Check only if the logged date is now complete, then advance only as far as contiguous completed dates go.
• Effort: Small. Benefit: Faster after large backlog catches up.

Cache weather condition translations in memory

• weather_conditions.json is loaded from disk every time a condition is translated.
• Load it once at module level and reuse the dictionary.
• Effort: Tiny. Benefit: Slightly faster header, cleaner code.

Reduce database queries in header

• Several queries can be combined or cached (e.g., prayer slot lookups, birthday checks).
• Consider single‑query fetches where possible to reduce round‑trips.
• Effort: Medium. Benefit: Snappier header, especially on Termux.

Add pagination to p q backlog list

• Currently shows all unlogged slots at once.
• Add n/p/q navigation (like view and search).
• Effort: Small. Benefit: Cleaner UI for long backlogs.

Expand test coverage

• Continue adding focused coverage for prayer status/nudges and calendar boundary corrections; weather fallback remains intentionally deferred.
• Keep the feature and integration suites aligned with behavior changes rather than duplicating command implementation details.
• Effort: Medium. Benefit: Prevent regressions.

---

🟢 Low Priority (nice to have, no urgency)

Add dispatch comments in commander.py

• Add # prayer_log.py, # sleep.py, etc. next to each handler in make_dispatch().
• Effort: Tiny. Benefit: Easier navigation.

Pre-compile regex patterns

• Move frequently used regexes (time parsing, @ tagging) to module‑level re.compile().
• Effort: Tiny. Benefit: Micro‑optimization, cleaner code.

Lazy‑load calendar event JSON files

• Currently all three event files are loaded when the module is first imported.
• Defer loading until get_events() is actually called.
• Effort: Small refactor in calendar_events.py. Benefit: Faster startup for quick commands that don’t touch the calendar.

Threaded weather fetch

• Fetch weather in background thread (with timeout) to avoid blocking startup.
• Current _fetch_failed_this_session flag is a good stopgap.
• Effort: Medium. Benefit: Startup never hangs on slow network.

Periodic VACUUM / DB health check

• Over many months, the DB can become fragmented.
• Show DB size in stats and optionally run VACUUM when size exceeds a threshold.
• Effort: Small. Benefit: Long‑term health, keeps backup sizes small.

Prune old weather_log rows (optional)

• The table grows by one row per successful fetch.
• Could add a weather prune command to keep only the most recent N days of history.
• User preference: may want to keep full history for future analysis; feature can be added later.
• Effort: Tiny. Benefit: Keeps DB slim on devices with limited storage.

Extract calendar event loaders

• calendar_events.py has a single _convert_all_events handling three calendars.
• Split into _load_jalali_events, _load_gregorian_events, _load_hijri_events for readability.
• Effort: Small. Benefit: Easier to maintain.

Future city support for prayer times

• prayer_times.py has a baked Tehran coordinate/calculation convention.
• When adding cities, move the location and calculation parameters into a keyed configuration.
• Effort: Deferred. Benefit: Clean multi‑city support.

Review EXPLAIN QUERY PLAN on slow queries

• Run EXPLAIN QUERY PLAN on header queries.
• Add indexes if needed (e.g., prayer_logs(jalali_date, prayer_slot) for backlog scanning).
• Effort: Tiny. Benefit: Prevents future slowness.

---

🟣 Very Low Priority (maybe someday)

Full FTS stemming of the index

• Currently we stem the search query but not the indexed description text. Stemming both sides could improve recall.
• Effort: Medium. Benefit: Slightly better search matches.

Synonym expansion for search

• Add a synonyms.json map to expand search queries (e.g., “coding” → “programming”).
• Effort: Small. Benefit: More forgiving search.

Toggle to disable ANSI colors

• Allow a plain‑text mode for terminals that don’t support ANSI escapes.
• Effort: Small. Benefit: Accessibility / compatibility.

Review Hijri year‑boundary edge cases

• Some Hijri events may not map correctly to the correct Jalali year due to the year‑boundary logic. Needs testing with real dates.
• Effort: Small investigation. Benefit: Correct calendar display.

Refactor to a central DailyDriverApp class

• Move the REPL loop, database connection, dispatcher, and state into a single app class for cleaner architecture and easier testing.
• Effort: Large refactor. Benefit: Better structure for future TUI integration.