FGP Pandoc Daemon

Ultra-fast document conversion with persistent Pandoc instance. 10-30x faster than spawning pandoc per command.

Why FGP?

Operation	FGP Daemon	Direct pandoc	Speedup
Markdown → HTML	2-5ms	~50ms	10-25x
Markdown → PDF	20-50ms	~300ms	6-15x
DOCX → Markdown	10-25ms	~150ms	6-15x
HTML → EPUB	15-40ms	~200ms	5-13x

Persistent daemon eliminates Haskell runtime startup overhead.

Installation

brew install fast-gateway-protocol/tap/fgp-pandoc

# Or
bash ~/.claude/skills/fgp-pandoc/scripts/install.sh

Prerequisites

Pandoc must be installed:

# macOS
brew install pandoc

# For PDF output, also install:
brew install basictex  # or mactex for full install

# Ubuntu/Debian
sudo apt install pandoc texlive-xetex

# Verify
pandoc --version

Usage

Basic Conversion

# Markdown to HTML
fgp pandoc convert doc.md --to html

# Markdown to PDF
fgp pandoc convert doc.md --to pdf

# Markdown to DOCX
fgp pandoc convert doc.md --to docx

# DOCX to Markdown
fgp pandoc convert doc.docx --to markdown

# HTML to Markdown
fgp pandoc convert page.html --to markdown

# With output file
fgp pandoc convert doc.md --to pdf --output document.pdf

Supported Formats

Input formats:

• Markdown (md, markdown, gfm, commonmark)
• HTML (html, html5)
• Word (docx)
• OpenDocument (odt)
• LaTeX (tex, latex)
• reStructuredText (rst)
• Org-mode (org)
• Textile, MediaWiki, Jira, and more

Output formats:

• HTML (html, html5)
• PDF (via LaTeX or wkhtmltopdf)
• Word (docx)
• EPUB (epub, epub3)
• LaTeX (tex, latex)
• Presentation (pptx, revealjs, beamer)
• Plain text (plain)
• And 40+ more

Markdown Flavors

# GitHub Flavored Markdown
fgp pandoc convert doc.md --from gfm --to html

# CommonMark
fgp pandoc convert doc.md --from commonmark --to html

# Pandoc Markdown (default)
fgp pandoc convert doc.md --from markdown --to html

# With extensions
fgp pandoc convert doc.md --from markdown+emoji+footnotes --to html

PDF Generation

# Basic PDF
fgp pandoc convert doc.md --to pdf

# With custom template
fgp pandoc convert doc.md --to pdf --template report.tex

# With table of contents
fgp pandoc convert doc.md --to pdf --toc

# With numbered sections
fgp pandoc convert doc.md --to pdf --number-sections

# Custom margins
fgp pandoc convert doc.md --to pdf --margin 1in

# Custom font
fgp pandoc convert doc.md --to pdf --font "Helvetica"

# A4 paper
fgp pandoc convert doc.md --to pdf --paper a4

# Via HTML (faster, different styling)
fgp pandoc convert doc.md --to pdf --engine wkhtmltopdf

DOCX Generation

# Basic DOCX
fgp pandoc convert doc.md --to docx

# With reference doc (for styling)
fgp pandoc convert doc.md --to docx --reference-doc template.docx

# With table of contents
fgp pandoc convert doc.md --to docx --toc

# Preserve formatting from source
fgp pandoc convert source.html --to docx --track-changes

EPUB Generation

# Basic EPUB
fgp pandoc convert book.md --to epub

# EPUB3
fgp pandoc convert book.md --to epub3

# With cover image
fgp pandoc convert book.md --to epub --cover cover.jpg

# With metadata
fgp pandoc convert book.md --to epub --metadata title="My Book" --metadata author="Name"

# With CSS styling
fgp pandoc convert book.md --to epub --css style.css

# With table of contents
fgp pandoc convert book.md --to epub --toc --toc-depth 2

Presentations

# PowerPoint
fgp pandoc convert slides.md --to pptx

# With template
fgp pandoc convert slides.md --to pptx --reference-doc template.pptx

# Reveal.js (HTML slides)
fgp pandoc convert slides.md --to revealjs --output slides.html

# Reveal.js with theme
fgp pandoc convert slides.md --to revealjs --variable theme=moon

# Beamer (LaTeX PDF slides)
fgp pandoc convert slides.md --to beamer --output slides.pdf

LaTeX

# Markdown to LaTeX
fgp pandoc convert doc.md --to latex

# With document class
fgp pandoc convert doc.md --to latex --documentclass article

# With custom preamble
fgp pandoc convert doc.md --to latex --include-in-header preamble.tex

# LaTeX to PDF
fgp pandoc convert doc.tex --to pdf

HTML Options

# Standalone HTML (with head/body)
fgp pandoc convert doc.md --to html --standalone

# With custom CSS
fgp pandoc convert doc.md --to html --css style.css

# With syntax highlighting
fgp pandoc convert doc.md --to html --highlight-style pygments

# Self-contained (embedded images/CSS)
fgp pandoc convert doc.md --to html --self-contained

# With MathJax for equations
fgp pandoc convert doc.md --to html --mathjax

# HTML5 semantic elements
fgp pandoc convert doc.md --to html5

Metadata & Variables

# Set title
fgp pandoc convert doc.md --to pdf --metadata title="Report"

# Set multiple metadata
fgp pandoc convert doc.md --to pdf \
  --metadata title="Report" \
  --metadata author="John Doe" \
  --metadata date="2024-01-15"

# From YAML file
fgp pandoc convert doc.md --to pdf --metadata-file metadata.yaml

# Template variables
fgp pandoc convert doc.md --to html --variable fontsize=12pt

Filters & Processing

# With Lua filter
fgp pandoc convert doc.md --to html --lua-filter filter.lua

# With multiple filters
fgp pandoc convert doc.md --to pdf --filter pandoc-crossref --filter pandoc-citeproc

# Bibliography/citations
fgp pandoc convert doc.md --to pdf --citeproc --bibliography refs.bib

# Citation style
fgp pandoc convert doc.md --to pdf --citeproc --csl ieee.csl

Templates

# List available templates
fgp pandoc templates --format html

# Use custom template
fgp pandoc convert doc.md --to html --template custom.html

# Get default template (for customization)
fgp pandoc template html > my-template.html

Batch Conversion

# Convert all markdown files
fgp pandoc batch convert ./docs --to html

# Convert to PDF
fgp pandoc batch convert ./chapters --to pdf --output ./pdfs

# With options
fgp pandoc batch convert ./posts --to html --css blog.css --standalone

# Pattern matching
fgp pandoc batch convert ./docs/*.md --to docx

Extract & Transform

# Extract text from DOCX
fgp pandoc convert doc.docx --to plain

# Extract to Markdown (clean)
fgp pandoc convert doc.docx --to gfm

# Strip formatting
fgp pandoc convert doc.html --to plain --wrap none

# JSON AST (for processing)
fgp pandoc convert doc.md --to json

# From JSON AST
fgp pandoc convert doc.json --from json --to html

Combine Documents

# Merge multiple files
fgp pandoc merge chapter1.md chapter2.md chapter3.md --to pdf --output book.pdf

# With metadata file
fgp pandoc merge --metadata-file book.yaml *.md --to epub --output book.epub

# From file list
fgp pandoc merge --files chapters.txt --to pdf

Available Commands

Command	Description	Example
convert	Convert document	fgp pandoc convert doc.md --to pdf
batch	Batch convert	fgp pandoc batch convert ./docs --to html
merge	Combine documents	fgp pandoc merge *.md --to pdf
template	Get template	fgp pandoc template html
templates	List templates	fgp pandoc templates --format pdf
formats	List formats	fgp pandoc formats
extensions	List extensions	fgp pandoc extensions markdown

Common Workflows

Blog post publishing

# Markdown to SEO-friendly HTML
fgp pandoc convert post.md --to html5 \
  --standalone \
  --metadata-file post.yaml \
  --css blog.css \
  --output post.html

Academic paper

# With citations and cross-references
fgp pandoc convert paper.md --to pdf \
  --citeproc \
  --bibliography refs.bib \
  --csl ieee.csl \
  --filter pandoc-crossref \
  --number-sections \
  --toc

Technical documentation

# Markdown to DOCX for clients
fgp pandoc convert docs/*.md --to docx \
  --reference-doc company-template.docx \
  --toc \
  --output documentation.docx

Ebook from chapters

# Combine chapters into EPUB
fgp pandoc merge \
  --metadata title="My Book" \
  --metadata author="Author Name" \
  chapters/*.md \
  --to epub3 \
  --cover cover.jpg \
  --toc \
  --output book.epub

Presentation from notes

# Markdown to reveal.js slides
fgp pandoc convert notes.md --to revealjs \
  --variable theme=black \
  --variable transition=slide \
  --standalone \
  --output presentation.html

Convert legacy docs

# DOCX to clean Markdown
fgp pandoc batch convert ./legacy/*.docx \
  --to gfm \
  --extract-media ./images \
  --output ./markdown

Troubleshooting

Pandoc not found

Error: pandoc not found in PATH

Install Pandoc: brew install pandoc or apt install pandoc

PDF generation failed

Error: pdflatex not found

Install LaTeX: brew install basictex or apt install texlive-xetex

Missing fonts

Error: Font not found

Install required fonts or use: --pdf-engine xelatex --variable mainfont="DejaVu Sans"

Unicode issues

Error: Unicode character not supported

Use XeLaTeX engine: --pdf-engine xelatex

Citation errors

Error: bibliography not found

Check path to .bib file and install pandoc-citeproc if needed.

Architecture

• Pandoc subprocess with persistent connection
• UNIX socket at ~/.fgp/services/pandoc/daemon.sock
• Template caching for faster repeated conversions
• Parallel batch processing for multiple files
• AST caching for incremental builds