s-sigdel/vimhjkl

Vimhjkl is a practice tool for learning Vim, the text editor, through hands-on drills run inside a real Vim session rather than a simulated environment. It covers 61 lessons and 222 challenges focusing on techniques that beginners often do not encounter in the standard Vim tutorial: things like text objects, macros, operator-motion combinations, registers, and command-line operations. The goal shown during a drill sits in a pane next to your actual editing buffer, so you are always practicing a real move. When you complete a challenge, the tool grades your keystrokes on both correctness and efficiency. Getting the right result is not enough on its own: your keystroke count is compared against a verified minimum, and using more than twice that minimum counts as inefficient. This grading discourages brute-force approaches and pushes you toward the idiomatic way of doing each task. Command-based drills that use Vim's colon commands require you to type an actual command rather than editing the file by hand. The tool tracks your progress using spaced repetition, a learning method where material you struggle with comes back sooner than material you know well. Each skill has a box level from 1 to 5, and a passing attempt moves it up while a wrong answer knocks it down. A skill is fully grooved after 25 clean, efficient repetitions, after which it appears only occasionally for maintenance. Harder material unlocks only once the easier tier below it is sufficiently mastered, so the curriculum expands gradually. There are several modes depending on what kind of practice you want. Learn mode shows you the technique before you attempt it. Blind mode shows only the before and after states and expects you to recall the method yourself. Practice mode queues up the skills you have been missing most. Grind mode lets you repeat a single skill many times in a row. A no-editor review mode works as plain flashcards without opening Vim. Installation is available through Homebrew on macOS and Linux, through the Arch Linux package manager, or directly from source. The only dependencies are Python and a working Vim or Neovim installation.