cheatsheet

name	cheatsheet
description	Create dense, printable A4 landscape cheatsheets from Markdown using a custom syntax with multi-column layout. Use when the user asks to create a cheatsheet, revision sheet, crib sheet, or condensed reference document for exams, courses, or technical topics. Supports math formulas, tables, highlights, notes, and multi-page output. Converts Markdown to styled HTML and exports to PDF for printing.

Cheatsheet Creator

Create dense, printable A4 landscape cheatsheets (297mm x 210mm, 5 columns, 6pt font) from a custom Markdown dialect.

Pipeline

Content → Write .md (custom syntax) → Run md2html.py → .html → Export PDF → Print

Quick Start

1. Create a .md file with front matter and content (see Syntax below)
2. Convert: python3 scripts/md2html.py input.md -o output.html
3. Export PDF: open HTML in browser → Ctrl+P → A4 Landscape, No margins, Background graphics ON → Save as PDF

Markdown Syntax

For complete syntax details, see references/cheatsheet_spec.md.

Block Elements

Syntax	Result
# Title	Blue section header banner
## Title	Red inline subsection header (first paragraph joins inline)
---PAGE---	Page break (new A4 page)
> text	Green note box
| a | b |	Table (with |---| separator = blue header)
Blank line	Paragraph separator

Inline Elements

Syntax	Result
**bold**	Bold
$formula$	Blue italic math
$=key formula=$	Yellow-highlighted math
==key text==	Yellow highlight
@@inline title@@	Red inline header (no break-avoid)
^{text}	Superscript
_{text}	Subscript
~{text}~	Grey small text

Document Structure

---
title: My Cheat Sheet
subtitle:
---

# Section Title

## Subsection
First paragraph joins inline with title. $formula$ and **bold**.

Second paragraph is separate. ==highlight==.

## Another Subsection
| Col1 | Col2 |
|------|------|
| data | data |

> Important note

---PAGE---

# Page 2 Section
...

Layout Rules

• Page: A4 landscape (297mm x 210mm), 5 columns, 3mm gaps, 3mm padding
• Font: 6pt base, line-height 1.22
• Column fill: auto — left-to-right, top-to-bottom
• Overflow: hidden — content beyond page boundary is silently lost
• Break-avoid: Every ## block stays intact (won't split across columns), causing small gaps at column bottoms

Capacity Planning

• Each page holds ~370-400 source lines of rendered content
• Target: fill all 5 columns per page (last column can be slightly short)
• Use wc -l to track line count during editing

Content Density Tips

• Use Unicode symbols: → ≈ ∈ ⊙ Σ Π ∂ ≤ ≥ ² ³
• Use abbreviations and shorthand
• Keep ## blocks small (1-3 paragraphs) to minimize column-bottom gaps
• Split large blocks into multiple ## sections
• Tables: keep under 8 rows
• Use @@sub-title@@ for sub-headers within a ## block (avoids creating new break-avoid)

Multi-Page Workflow — Page-by-Page

Work one page at a time. Do NOT write all pages at once.

1. Page 1: Write ~380 lines of content (no ---PAGE--- yet)
2. Run md2html.py → open HTML → verify all 5 columns are filled
3. If under-filled: add more content. If overflowing: trim or move content to next page
4. Once Page 1 is good, append ---PAGE--- and start Page 2 content
5. Repeat: write → build → verify → adjust for each page
6. Continue until all content is placed

Why page-by-page? Writing everything at once makes it hard to judge density. Tables and ## break-avoid blocks consume unpredictable vertical space. Filling one page at a time ensures every page uses all 5 columns with no overflow.

HTML Generation

The converter script (scripts/md2html.py) is zero-dependency Python:

python3 scripts/md2html.py input.md -o output.html

Open output.html in a browser to preview. The CSS renders the exact print layout on screen.

PDF Export

Method 1: Browser Print (Recommended)

Open HTML in Chrome/Safari → Ctrl+P (Cmd+P on Mac):

• Layout: Landscape
• Paper: A4
• Margins: None
• Background graphics: ON
• Save as PDF

Method 2: Playwright Screenshot-to-PDF

When browser print is unavailable, use Playwright to capture high-resolution screenshots and convert to PDF with Pillow. This bypasses CSS print pagination issues.

# 1. Start HTTP server in the directory with the HTML file
# python3 -m http.server 8765

# 2. Playwright: capture each .page element at 4x scale
from playwright.sync_api import sync_playwright

with sync_playwright() as p:
    browser = p.chromium.launch()
    ctx = browser.new_context(
        viewport={"width": 1122, "height": 794},
        device_scale_factor=4
    )
    page = ctx.new_page()
    page.goto("http://localhost:8765/output.html")
    page.wait_for_load_state("networkidle")

    pages_el = page.locator(".page")
    for i in range(pages_el.count()):
        pages_el.nth(i).screenshot(path=f"page{i+1}.png")

    browser.close()

# 3. Convert to PDF with Pillow
from PIL import Image

A4_W, A4_H = 3508, 2480  # 300 DPI A4 landscape
images = []
for i in range(num_pages):
    img = Image.open(f"page{i+1}.png")
    img = img.resize((A4_W, A4_H), Image.LANCZOS).convert("RGB")
    images.append(img)

images[0].save(
    "output.pdf", "PDF", resolution=300.0,
    save_all=True, append_images=images[1:]
)

Do NOT use Chrome headless --print-to-pdf or Playwright page.pdf() — these fail to render multi-page .page divs correctly (second page renders blank).

Gotchas

Issue	Solution
Content invisible / cut off	overflow: hidden — reduce content or add ---PAGE---
Large column-bottom gaps	Break ## blocks into smaller ones (1-3 paragraphs)
Table overflows column	Shorten cell text, remove backtick formatting, abbreviate
| in formulas parsed as table	Use ‖ or · instead of | inside table cells
Second PDF page blank	Use screenshot-to-PDF method, NOT Chrome print-to-pdf
Content not enough / right columns empty	Add more content to fill ~380 lines per page

Verification Checklist

After generating HTML, verify:

1. No overflow: scrollHeight ≈ clientHeight for each .page div (check via browser DevTools)
2. All columns filled: visually confirm 5 columns on each page have content
3. No table overflow: no text bleeding into adjacent columns
4. Highlights render: yellow == and $= =$ backgrounds visible
5. Page count correct: exact number of intended pages