# cal-notif

Calendar notifications for the desktop. Reads the `.ics` files vdirsyncer
keeps in sync, shows a popup at each alarm (with a Snooze button) and says
it out loud with a local Kokoro voice. Nothing leaves the machine.

- Each event's own alarms (VALARM) set when it notifies.
- Events without alarms can get per-calendar defaults, or per-event
  alarms chosen from a rofi menu. A per-event choice replaces the event's
  own alarms, and `none` silences it.

## Requirements

- Python 3.11+ with `icalendar`, `python-dateutil`, `numpy`, `kokoro-onnx`
- Kokoro model files (`kokoro-v1.0.onnx`, `voices-v1.0.bin`)
- `notify-send` and a notification server that supports actions
- `aplay`, `rofi`
- vdirsyncer with a `filesystem` storage for the calendars

## Usage

    cal-notif daemon     # start from your compositor's autostart
    cal-notif pick       # bind to a key: set alarms for one event
    cal-notif say TEXT   # test the voice

Config: `~/.config/cal-notif/config.toml`, see `config.example.toml`.
Per-event alarms from `pick` go to `~/.config/cal-notif/overrides.toml`
(`"<event UID>" = ["1h", "10m"]`), which can also be edited by hand. The
daemon picks up changes to both files, and to the calendars, within 30 s.

In the picker, type durations like `1d 2h` (one alarm) or `1d, 1h` (two
alarms), `none` to silence the event, `reset` to go back to its own alarms.

With Do Not Disturb on, the notification server may close the popup into
its drawer without reporting an action, so that alarm can no longer be
snoozed. The voice still plays.

## Tests

    /usr/bin/python3 test_cal_notif.py

## License

GPLv2 only, see `LICENSE`.

## Development Approach

This project is developed using AI-assisted tools. Code is generated with the help of AI based on human-provided specifications, design decisions, and iterative feedback.

All contributions are reviewed, tested, and curated by the maintainer before being included in the codebase. AI is used as a productivity and exploration tool, while human oversight remains central to all decisions.

The goal is to combine the flexibility of AI-assisted development with standard open-source practices such as transparency, review, and accountability.