PDF Engine Fallback

This skill provides a diagnostic-retry workflow for PDF generation when Pandoc's default PDF engine fails. It systematically checks available engines and retries with alternatives optimized for different content types (e.g., xelatex for Unicode).

When to Use

• Pandoc PDF generation fails with errors
• Document contains non-ASCII characters (Unicode, CJK, etc.)
• Need robust PDF generation across different system configurations
• Working in CI/CD or automated document pipelines

Steps

1. Attempt Initial PDF Generation

First, try generating with pandoc's default engine:

pandoc input.md -o output.pdf

2. Diagnose on Failure

If the initial attempt fails, check which PDF engines are available:

# Check for pdflatex
which pdflatex && echo "pdflatex: available" || echo "pdflatex: not found"

# Check for xelatex (better Unicode support)
which xelatex && echo "xelatex: available" || echo "xelatex: not found"

# Check for wkhtmltopdf (HTML-based rendering)
which wkhtmltopdf && echo "wkhtmltopdf: available" || echo "wkhtmltopdf: not found"

3. Retry with Alternative Engine

Based on available engines and document content, retry with an appropriate engine:

# For documents with Unicode/special characters - prefer xelatex
pandoc input.md -o output.pdf --pdf-engine=xelatex

# Alternative: use pdflatex for simple Latin documents
pandoc input.md -o output.pdf --pdf-engine=pdflatex

# For HTML-heavy content or complex layouts
pandoc input.md -o output.pdf --pdf-engine=wkhtmltopdf

4. Handle Missing Engines

If no PDF engines are available, install one:

# Ubuntu/Debian - install texlive with xelatex
sudo apt-get install texlive-xetex

# Or install minimal teTeX
sudo apt-get install texlive-latex-base texlive-latex-extra

# macOS with Homebrew
brew install basictex

Complete Fallback Script

#!/bin/bash
# pdf-fallback.sh - Robust PDF generation with engine fallback

INPUT_FILE="${1:-input.md}"
OUTPUT_FILE="${2:-output.pdf}"

echo "Attempting PDF generation: $INPUT_FILE -> $OUTPUT_FILE"

# Try default engine first
if pandoc "$INPUT_FILE" -o "$OUTPUT_FILE" 2>/dev/null; then
    echo "SUCCESS: PDF generated with default engine"
    exit 0
fi

echo "Default engine failed. Checking available engines..."

# Check and try engines in priority order
if command -v xelatex &>/dev/null; then
    echo "Retrying with xelatex (Unicode support)..."
    if pandoc "$INPUT_FILE" -o "$OUTPUT_FILE" --pdf-engine=xelatex; then
        echo "SUCCESS: PDF generated with xelatex"
        exit 0
    fi
fi

if command -v pdflatex &>/dev/null; then
    echo "Retrying with pdflatex..."
    if pandoc "$INPUT_FILE" -o "$OUTPUT_FILE" --pdf-engine=pdflatex; then
        echo "SUCCESS: PDF generated with pdflatex"
        exit 0
    fi
fi

if command -v wkhtmltopdf &>/dev/null; then
    echo "Retrying with wkhtmltopdf..."
    if pandoc "$INPUT_FILE" -o "$OUTPUT_FILE" --pdf-engine=wkhtmltopdf; then
        echo "SUCCESS: PDF generated with wkhtmltopdf"
        exit 0
    fi
fi

echo "FAILURE: No available PDF engine succeeded"
echo "Install a PDF engine (texlive-xetex recommended) and retry"
exit 1

Usage

# Make executable
chmod +x pdf-fallback.sh

# Run with defaults (input.md -> output.pdf)
./pdf-fallback.sh

# Run with custom files
./pdf-fallback.sh report.md report.pdf

Engine Selection Guide

Engine	Best For	Unicode	Speed
xelatex	Non-Latin scripts, modern fonts	Excellent	Moderate
pdflatex	Simple Latin documents	Limited	Fast
wkhtmltopdf	HTML/CSS styling	Good	Fast

Troubleshooting

• Font errors: Install required fonts or use xelatex with system fonts
• Package errors: Install missing LaTeX packages via tlmgr
• Large documents: Try pdflatex for speed, xelatex for compatibility
• Citation issues: Ensure pandoc-citeproc or citeproc is installed