PySplit

PySplit is a LiveSplit-style speedrun timer, run via the command line.

Usage

To start a run on PySplit:

$ python pysplit/runner.py run <speedrun_name>

PySplit consists of a client and server. By default, the command above will run instances of the client and server together.

The client is a curses app with the following control defaults:

• space: start/split
• backspace: stop/reset
• q: quit

These controls can be configured. Values should be ASCII key IDs as used by curses.window.getch:

controls:
  advance: 32  # space
  stop_reset: 127  # backspace/delete
  quit: 113  # q

Configuration

Configuration is done via the file ~/.pysplit.yaml. An alternative config file can be specified with the argument --config. There is a section each for the client and server.

Client config

The first required config is your name:

client:
  runner_name: Some speedrunner

Then to add level names for a run category, for example 'Halo 1 Legendary', add the category name under split_names with a list of split names:

client:
  split_names:
    'Halo 1 Legendary':
      - Pillar of Autumn
      - Halo
      - Truth and Reconciliation
      - Silent Cartographer
      - Assault on the Control Room
      - 343 Guilty Spark
      - The Library
      - Two Betrayals
      - Keyes
      - The Maw

Because PySplit saves the index of each split (i.e, whether it is split 1, 2, etc.) and not the name, level names can be changed at any time - just make sure the list of levels is still the same length. You can also automatically assign a run category's level names to another category by specifying a category name instead of a list of level names.

client:
  split_names:
    'Halo 1 Legendary':
      - Reset here a lot
      - Betraying teammates
      - TnR
      - Choke door launch here
      - AotCR
      - 343GS
      - Brown hallways and Flood
      - Lose run here
      - Keyes bump
      - Warthog run

    'Halo 1 Easy': 'Halo 1 Legendary'

The following optional configs are available to change the control keys and to play sounds when gold splits and PBs are achieved:

gold_sound: path/to/gold_sound.mp3
pb_sound: path/to/pb_sound.mp3
controls:
    advance: 32  # default = 32 (space)
    stop_reset: 127  # default = 127 (backspace)
    quit: 113  # default = 113 (q)

Server config

There are no required server configs, but the following options are available:

server:
  record_db: path/to/pysplit.sqlite  # default = ~/.pysplit.sqlite
  port: <server_port>  # default = 5000

Running the client and server separately

The server can be run on its own with the command:

$ python pysplit/runner.py server

Then to point the client at this, add the option server_url to the config file:

client:
    server_url: http://localhost:5000

Then the client will connect to the running server:

$ python pysplit/runner.py run <speedrun_name> ...

About saving split data

By default, PySplit uses a splits file in ~/.pysplit.sqlite. A PySplit splits file is a SQLite database with the following schema:

CREATE TABLE runs (id numeric UNIQUE, name text, runner text, start_time datetime, end_time datetime, total_time numeric);
CREATE TABLE splits (id numeric UNIQUE, run_id numeric REFERENCES runs(id), idx numeric, start_time datetime, end_time datetime, total_time numeric);

PySplit uses the sqlite3 Python library to push and pull data from this file. It can also be interacted with directly in a sqlite3 shell:

sqlite> SELECT id, runner, start_time, total_time FROM runs WHERE name="Halo 1 Legendary" ORDER BY total_time ASC;

Signal handling

The server handles SIGINT to stop cleanly. The client handles SIGUSR1 to call self.advance. This means that the PySplit client can run in the background under the control of an external process.

Roadmap

• More command line tools for looking at past runs
  • best run, avg run, best possible run
• Web interface?
• Cross-platform sound