not-matthias/hashcards

<!-- Source: https://github.com/eudoxia0/hashcards -->

Hashcards is a plain-text, Rust-based flashcard system using the FSRS algorithm. Cards are Markdown files identified by content hash — editing a card resets its review progress.

Card Syntax

Question-Answer

Q: What is the capital of France?
A: Paris

Cloze Deletion

C: The capital of France is [Paris].

Multiple deletions in one card generate multiple review items:

C: [Paris] is the capital of [France].

Escape literal brackets with backslash: \[not a deletion\].

Separators

Use --- between cards for visual organization (optional, ignored by parser).

Rich Content

• LaTeX math: Inline $E = mc^2$, display $$\int_0^1 x\,dx$$
• Custom macros: Define in frontmatter or shared macro files
• Images: ![alt](path/to/image.png) (relative paths)
• Audio: Standard Markdown link/embed syntax
• Code blocks: Fenced with syntax highlighting
• Tables: Standard Markdown tables

Deck Frontmatter (optional TOML)

+++
deck = "physics"
+++

Q: What is Newton's second law?
A: F = ma

CLI Commands

hashcards drill [DIRECTORY]   # Start review session (web UI on localhost:8000)
hashcards drill --host 0.0.0.0  # Bind to all interfaces
hashcards stats               # Collection statistics as JSON
hashcards check               # Verify collection integrity
hashcards orphans             # Manage deleted cards still in database
hashcards export              # Export collections as JSON

Drill Options

• --host <addr> — Bind address (default: 127.0.0.1)
• Port selection for custom listening port
• Card limit to cap review session size
• Deck filtering to review specific decks only

Review Grading

Four levels: Forgot / Hard / Good / Easy (keyboard shortcuts available in web UI).

Sibling burial: when a cloze card generates multiple items, reviewing one buries the others to prevent spoilers.

Workflow Best Practices

• Version control: Store card files in Git — content-addressed design works naturally with diffs and history
• Scripted generation: Generate cards programmatically from CSV or structured data, writing Q:/A: or C: format to .md files
• Deck organization: Use directories and/or TOML frontmatter deck field to group cards by topic
• Editing awareness: Changing card text resets its FSRS progress (the hash changes). Typo fixes or rewords will lose review history
• Integrity checks: Run hashcards check periodically to catch corrupt or malformed cards
• Stats monitoring: Use hashcards stats to track review load and identify struggling decks