MCP/mcp-pdf-tools
1
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity

Compare commits

base: MCP:31b8b2e6d40325603f5ac142b0b40ab88e596a6f
Branches Tags
MCP:main
..
compare: MCP:b2d9073f04979f1fa45a6ed5eded46da4d7ab1ec
Branches Tags
MCP:main

No commits in common. "31b8b2e6d40325603f5ac142b0b40ab88e596a6f" and "b2d9073f04979f1fa45a6ed5eded46da4d7ab1ec" have entirely different histories.
31b8b2e6d4 ... b2d9073f04

2 changed files with 37 additions and 218 deletions
Show Stats Expand all files Collapse all files

149
QUICKSTART.md
View File

@ -2,38 +2,12 @@
## 1. Installation ## 1. Installation
### Option A: Run from PyPI with uvx (Recommended for end users) ### Option A: Using UV (Recommended for Development)
No clone required — `uvx` fetches and runs in an isolated cached venv:
```bash
# Bare install
uvx mcp-pdf
# With markdown_to_pdf support (requires pandoc on host)
uvx --from "mcp-pdf[markdown]" mcp-pdf
# Force a refresh after a new release
uvx --refresh --from "mcp-pdf[markdown]" mcp-pdf
```
### Option B: pip install from PyPI
```bash
pip install mcp-pdf
# Or with optional extras:
pip install "mcp-pdf[markdown]" # adds markdown_to_pdf
pip install "mcp-pdf[forms]" # adds form creation tools
pip install "mcp-pdf[tables]" # adds Camelot/Tabula table extraction
pip install "mcp-pdf[all]" # everything
```
### Option C: Local development with uv
```bash ```bash
# Clone the repository # Clone the repository
git clone https://github.com/rsp2k/mcp-pdf git clone https://github.com/rpm/mcp-pdf-tools
cd mcp-pdf cd mcp-pdf-tools
# Install with uv # Install with uv
uv sync uv sync
@ -42,66 +16,41 @@ uv sync
uv run python examples/verify_installation.py uv run python examples/verify_installation.py
``` ```
### Option D: Using Docker ### Option B: Using Docker
```bash ```bash
git clone https://github.com/rsp2k/mcp-pdf # Clone the repository
cd mcp-pdf git clone https://github.com/rpm/mcp-pdf-tools
cd mcp-pdf-tools
docker compose build # Build and run with Docker
docker compose run --rm mcp-pdf python examples/verify_installation.py docker-compose build
docker-compose run --rm mcp-pdf-tools python examples/verify_installation.py
```
### Option C: From PyPI
```bash
pip install mcp-pdf-tools
``` ```
## 2. System Dependencies ## 2. System Dependencies
`uvx` and `pip` only handle Python deps. Some tools call out to system binaries that you'll need to install separately:
| Binary | Required for |
|--------|-------------|
| `tesseract` | `ocr_pdf` |
| `ghostscript` | Camelot table extraction |
| `java` (JRE) | Tabula table extraction |
| `poppler` | PDF→image conversion |
| `pandoc` | `markdown_to_pdf` |
| `xelatex` / `pdflatex` / `tectonic` / `weasyprint` / `wkhtmltopdf` | `markdown_to_pdf` (need at least one) |
> **Note on the LaTeX engine:** `texlive-xetex` alone is often not enough for real markdown docs — pandoc's default template needs LaTeX packages (`lastpage`, `xcolor`, `framed`, `fancyhdr`, etc.) that live in `texlive-latex-extra` (Debian) / `texlive-latexextra` (Arch). If you don't already use TeX, **`tectonic` is a much better choice** — it's a ~30 MB static binary that downloads packages on demand. See the README's "Picking a PDF engine" table for details.
### Ubuntu/Debian ### Ubuntu/Debian
```bash ```bash
sudo apt-get update sudo apt-get update
sudo apt-get install -y \ sudo apt-get install -y \
tesseract-ocr tesseract-ocr-eng \ tesseract-ocr \
poppler-utils ghostscript \ tesseract-ocr-eng \
python3-tk default-jre-headless poppler-utils \
ghostscript \
# For markdown_to_pdf, pick one of: python3-tk \
sudo apt-get install -y pandoc # then install tectonic separately default-jre-headless
sudo apt-get install -y pandoc texlive-xetex texlive-latex-extra texlive-fonts-extra # full TeX
sudo apt-get install -y pandoc && pip install weasyprint # skip TeX
``` ```
### Arch Linux ### macOS
```bash
sudo pacman -S \
tesseract tesseract-data-eng \
poppler ghostscript \
jre-openjdk-headless tk
# For markdown_to_pdf, pick one of:
sudo pacman -S pandoc tectonic # recommended
sudo pacman -S pandoc texlive-xetex texlive-latexextra texlive-fontsextra # full TeX
sudo pacman -S pandoc && pip install weasyprint # skip TeX
```
### macOS (Homebrew)
```bash ```bash
brew install tesseract poppler ghostscript brew install tesseract poppler ghostscript
# For markdown_to_pdf, pick one of:
brew install pandoc tectonic # recommended
brew install pandoc && brew install --cask mactex-no-gui # full TeX
brew install pandoc weasyprint # skip TeX
``` ```
### Windows ### Windows
@ -109,33 +58,18 @@ brew install pandoc weasyprint # skip TeX
- Install Poppler: http://blog.alivate.com.au/poppler-windows/ - Install Poppler: http://blog.alivate.com.au/poppler-windows/
- Install Ghostscript: https://www.ghostscript.com/download/gsdnld.html - Install Ghostscript: https://www.ghostscript.com/download/gsdnld.html
- Install Java: https://www.java.com/download/ - Install Java: https://www.java.com/download/
- Install Pandoc (for `markdown_to_pdf`): https://pandoc.org/installing.html
- Install MiKTeX or wkhtmltopdf for the PDF engine
## 3. Adding to Claude Code / Claude Desktop ## 3. Claude Desktop Configuration
### Easiest — `claude mcp add` with uvx Add to `~/Library/Application Support/Claude/claude_desktop_config.json`:
```bash
# Bare
claude mcp add pdf-tools -- uvx mcp-pdf
# With markdown_to_pdf support
claude mcp add pdf-tools -- uvx --from "mcp-pdf[markdown]" mcp-pdf
```
The `--` separator is required so the Claude CLI doesn't try to parse `--from` as one of its own flags.
### Manual config (Claude Desktop)
Edit `~/Library/Application Support/Claude/claude_desktop_config.json` (macOS) or `~/.config/Claude/claude_desktop_config.json` (Linux):
```json ```json
{ {
"mcpServers": { "mcpServers": {
"pdf-tools": { "pdf-tools": {
"command": "uvx", "command": "uv",
"args": ["--from", "mcp-pdf[markdown]", "mcp-pdf"] "args": ["run", "mcp-pdf-tools"],
"cwd": "/home/rpm/claude/mcp-pdf-tools"
} }
} }
} }
@ -152,20 +86,14 @@ uv run python examples/test_pdf_tools.py /path/to/your/document.pdf
### OCR not working ### OCR not working
- Check Tesseract is installed: `tesseract --version` - Check Tesseract is installed: `tesseract --version`
- Install language packs: `sudo apt-get install tesseract-ocr-[lang]` (Debian) or `sudo pacman -S tesseract-data-[lang]` (Arch) - Install language packs: `sudo apt-get install tesseract-ocr-[lang]`
### Table extraction failing ### Table extraction failing
- Check Java is installed: `java -version` - Check Java is installed: `java -version`
- For Camelot issues, ensure Ghostscript is installed: `gs --version` - For Camelot issues, ensure Ghostscript is installed
### `markdown_to_pdf` errors
- "pandoc binary not found" → install pandoc (see System Dependencies)
- "No PDF engine found" → install at least one of `xelatex`, `pdflatex`, `tectonic`, `weasyprint`, `wkhtmltopdf`
- "Pandoc died with exitcode 43" + `mktexfmt` errors → your TeX install is missing format files; rebuild with `sudo fmtutil-sys --all` or use a different engine via `pdf_engine="weasyprint"`
- The tool reports `detected_engines` in its response — check that field to see what's actually available
### Large PDF issues ### Large PDF issues
- Process specific pages: `pages="1-10"` or `pages="1,3,5"` - Process specific pages: `pages=[0, 1, 2]`
- Increase memory: `export JAVA_OPTS="-Xmx2g"` - Increase memory: `export JAVA_OPTS="-Xmx2g"`
## 6. Example Usage in Claude ## 6. Example Usage in Claude
@ -177,21 +105,6 @@ Once configured, you can ask Claude:
- "Extract all tables from /path/to/report.pdf and format as markdown" - "Extract all tables from /path/to/report.pdf and format as markdown"
- "Convert /path/to/document.pdf to markdown format" - "Convert /path/to/document.pdf to markdown format"
- "Extract images from the first 5 pages of /path/to/presentation.pdf" - "Extract images from the first 5 pages of /path/to/presentation.pdf"
- "Build a PDF from /path/to/notes.md with a table of contents"
## 7. Verify the Built-in Test
Convert this README itself to PDF as a smoke test once everything is wired up:
```python
markdown_to_pdf(
markdown_path="QUICKSTART.md",
output_path="/tmp/quickstart.pdf",
toc=True,
)
```
The response includes `detected_engines` so you can see exactly what's installed on your host.
## Need Help? ## Need Help?

106
README.md
View File

@ -6,7 +6,7 @@
**A FastMCP server for PDF processing** **A FastMCP server for PDF processing**
*47 tools for text extraction, OCR, tables, forms, annotations, markdown↔PDF, and more* *46 tools for text extraction, OCR, tables, forms, annotations, and more*
[![Python 3.11+](https://img.shields.io/badge/python-3.11+-blue.svg?style=flat-square)](https://www.python.org/downloads/) [![Python 3.11+](https://img.shields.io/badge/python-3.11+-blue.svg?style=flat-square)](https://www.python.org/downloads/)
[![FastMCP](https://img.shields.io/badge/FastMCP-2.0+-green.svg?style=flat-square)](https://github.com/jlowin/fastmcp) [![FastMCP](https://img.shields.io/badge/FastMCP-2.0+-green.svg?style=flat-square)](https://github.com/jlowin/fastmcp)
@ -31,25 +31,19 @@ MCP PDF extracts content from PDFs using multiple libraries with automatic fallb
- **Document assembly** - merge, split, reorder pages - **Document assembly** - merge, split, reorder pages
- **Annotations** - sticky notes, highlights, stamps - **Annotations** - sticky notes, highlights, stamps
- **Vector graphics** - extract to SVG for schematics and technical drawings - **Vector graphics** - extract to SVG for schematics and technical drawings
- **Format conversion** - PDF ↔ Markdown (PDF→MD via PyMuPDF, MD→PDF via pandoc)
--- ---
## Quick Start ## Quick Start
```bash ```bash
# Run from PyPI (one-shot, no permanent install) # Install from PyPI
uvx mcp-pdf uvx mcp-pdf
# Add to Claude Code — note the `--` separator before uvx # Or add to Claude Code
claude mcp add pdf-tools -- uvx mcp-pdf claude mcp add pdf-tools uvx mcp-pdf
# Include the markdown_to_pdf tool (requires pandoc on host)
claude mcp add pdf-tools -- uvx --from "mcp-pdf[markdown]" mcp-pdf
``` ```
> `uvx` caches tool installs aggressively. After upgrading to a new release, force a refresh with `uvx --refresh mcp-pdf` (or `uvx --refresh --from "mcp-pdf[markdown]" mcp-pdf` if you're using extras).
<details> <details>
<summary><b>Development Installation</b></summary> <summary><b>Development Installation</b></summary>
@ -61,11 +55,6 @@ uv sync
# System dependencies (Ubuntu/Debian) # System dependencies (Ubuntu/Debian)
sudo apt-get install tesseract-ocr tesseract-ocr-eng poppler-utils ghostscript sudo apt-get install tesseract-ocr tesseract-ocr-eng poppler-utils ghostscript
# For markdown_to_pdf — pick one PDF-engine route:
sudo apt-get install pandoc tectonic # recommended (small)
# or: sudo apt-get install pandoc texlive-xetex texlive-latex-extra # full TeX
# or: sudo apt-get install pandoc && pip install weasyprint # skip TeX
# Verify # Verify
uv run python examples/verify_installation.py uv run python examples/verify_installation.py
``` ```
@ -84,18 +73,10 @@ uv run python examples/verify_installation.py
| `extract_tables` | Extract tables to JSON, CSV, or Markdown | | `extract_tables` | Extract tables to JSON, CSV, or Markdown |
| `extract_images` | Extract embedded images | | `extract_images` | Extract embedded images |
| `extract_links` | Get all hyperlinks with page filtering | | `extract_links` | Get all hyperlinks with page filtering |
| `pdf_to_markdown` | Convert PDF to markdown preserving structure |
| `ocr_pdf` | OCR scanned documents using Tesseract | | `ocr_pdf` | OCR scanned documents using Tesseract |
| `extract_vector_graphics` | Export vector graphics to SVG (schematics, charts, drawings) | | `extract_vector_graphics` | Export vector graphics to SVG (schematics, charts, drawings) |
### Format Conversion
| Tool | What it does |
|------|-------------|
| `pdf_to_markdown` | Convert PDF to markdown preserving structure; extracts images and SVG vectors to disk |
| `markdown_to_pdf` | Convert `.md` files (or inline text) to PDF via pandoc with auto-detected engine |
**`markdown_to_pdf` requires:** `pip install mcp-pdf[markdown]` plus the pandoc binary and at least one PDF engine (`xelatex`, `pdflatex`, `tectonic`, `weasyprint`, or `wkhtmltopdf`) on PATH. The tool auto-detects what's available and uses the highest-quality one. Pass `pdf_engine=` to override or `extra_args=` for raw pandoc options.
### Document Analysis ### Document Analysis
| Tool | What it does | | Tool | What it does |
@ -212,87 +193,12 @@ Some features require system packages:
| Camelot tables | `ghostscript` | | Camelot tables | `ghostscript` |
| Tabula tables | `default-jre-headless` | | Tabula tables | `default-jre-headless` |
| PDF to images | `poppler-utils` | | PDF to images | `poppler-utils` |
| `markdown_to_pdf` | `pandoc` + one of: `tectonic`, `texlive-xetex` (+ `texlive-latex-extra`), `weasyprint`, `wkhtmltopdf` |
### Picking a PDF engine for `markdown_to_pdf` Ubuntu/Debian:
Pandoc takes markdown → HTML or LaTeX → PDF. The LaTeX path produces the most polished output but needs a TeX install. Trade-offs:
| Engine | Disk size | Notes |
|--------|----------|-------|
| **`tectonic`** | ~30 MB | **Recommended for new installs.** Single static binary. Downloads LaTeX packages on demand — no upfront mass-install. |
| `xelatex` + `texlive-latex-extra` | ~500 MB | Best output once installed. Use if you already run TeX. The `-extra` package matters: pandoc's default template needs `lastpage`, `xcolor`, `framed`, `fancyhdr`, etc. — all of which live there, **not** in `texlive-xetex`. |
| `xelatex` alone (just `texlive-xetex`) | ~200 MB | **Often breaks.** Expect `! LaTeX Error: File 'X.sty' not found` on real docs. |
| `weasyprint` | ~40 MB | Pure-Python (`pip install weasyprint`) + cairo/pango system libs. HTML/CSS path — no LaTeX. Good for simple docs; weaker on math, footnotes, citations. |
| `wkhtmltopdf` | ~40 MB | Older HTML-to-PDF tool. Adequate but less actively maintained. |
**Ubuntu/Debian:**
```bash ```bash
sudo apt-get install tesseract-ocr tesseract-ocr-eng poppler-utils ghostscript default-jre-headless sudo apt-get install tesseract-ocr tesseract-ocr-eng poppler-utils ghostscript default-jre-headless
# For markdown_to_pdf — pick one engine route:
# Option A — tectonic (smallest, downloads packages on demand)
sudo apt-get install pandoc
# tectonic isn't in apt — install via cargo or download static binary:
# https://tectonic-typesetting.github.io/en-US/install.html
# Option B — full TeX (best quality, large download)
sudo apt-get install pandoc texlive-xetex texlive-latex-extra texlive-fonts-extra
# Option C — weasyprint (skip TeX entirely)
sudo apt-get install pandoc
pip install weasyprint
``` ```
**Arch Linux:**
```bash
sudo pacman -S tesseract tesseract-data-eng poppler ghostscript jre-openjdk-headless
# For markdown_to_pdf — pick one engine route:
# Option A — tectonic (recommended for new installs, in official repo)
sudo pacman -S pandoc tectonic
# Option B — full TeX (best output, ~500 MB)
sudo pacman -S pandoc texlive-xetex texlive-latexextra texlive-fontsextra
# Option C — weasyprint (skip TeX)
sudo pacman -S pandoc
pip install weasyprint # or: uv pip install weasyprint
# Option D — wkhtmltopdf (from AUR)
yay -S wkhtmltopdf-static
```
**macOS (Homebrew):**
```bash
brew install tesseract poppler ghostscript
# For markdown_to_pdf — pick one engine route:
# Option A — tectonic (recommended)
brew install pandoc tectonic
# Option B — full TeX (mactex-no-gui includes the latex-extra equivalent)
brew install pandoc
brew install --cask mactex-no-gui
# Option C — weasyprint
brew install pandoc weasyprint
```
## Optional Extras
The base install stays lean. Heavy or niche dependencies are gated behind extras:
| Extra | Adds | When to install |
|-------|------|----------------|
| `mcp-pdf[forms]` | `reportlab` | Form creation tools (`create_form_pdf`, permit forms) |
| `mcp-pdf[tables]` | `camelot-py`, `tabula-py` | Higher-accuracy table extraction (also needs Java + Ghostscript) |
| `mcp-pdf[markdown]` | `pypandoc` | `markdown_to_pdf` tool (also needs pandoc binary) |
| `mcp-pdf[all]` | All of the above | Want everything |
--- ---
## Configuration ## Configuration