Configuration¶
chordgen reads its config from ~/.config/chordgen/config.yaml,
created by chordgen setup. The full schema (with every option's
default) is generated from the
pydantic config model,
see Schema.
For non-trivial setups (custom layouts, key codes, priority tiers), my dotfiles are a useful reference.
Top-level keys¶
gen— chord generation: vocabulary file, keyboard model, alt-generator categories, and assignment options. See gen.output— which output formats to emit and per-format options (key codes, key positions, chord timeouts, ...). See Output formats.learn— learn-mode TUI knobs (daily quotas, leech / mastery thresholds, slow grading). See learn.drill— drill-mode TUI knobs (mode, count, timer). See drill.theme— Textual theme used by the learn and drill TUIs. Updated automatically when you change the theme via the in-app command palette (Ctrl+P).
Ignoring words¶
The gen.ignore_words list tells chordgen gen to leave certain rows
alone. Ignored words remain in chords.csv, but they are skipped when
scoring and assigning chords, so they won't consume chord keys or
trigger "unable to find options" warnings. Their chord and alt slots
are cleared each run. This is useful for low-value short words
(fillers, interjections, dialect forms) that appear in frequency lists
but aren't worth a dedicated chord.
The default list is: