textproc/pymupdf4llm-mcp: New port: MCP server for pymupdf4llm: convert PDF to Markdown for LLMs

2026-06-02 11:08:52 +00:00 · 2026-04-07 16:38:38 -07:00
parent 12bf1393bf
commit c9ecc8e87b
5 changed files with 71 additions and 0 deletions
@@ -1775,6 +1775,7 @@
    SUBDIR += py-zensical
    SUBDIR += py-zope.i18nmessageid
    SUBDIR += py-zpt
+    SUBDIR += pymupdf4llm-mcp
    SUBDIR += qo
    SUBDIR += qprint
    SUBDIR += qr
@@ -0,0 +1,31 @@
+PORTNAME=	pymupdf4llm-mcp
+DISTVERSION=	0.0.4
+CATEGORIES=	textproc python
+MASTER_SITES=	PYPI
+DISTNAME=	${PORTNAME:S/-/_/}-${DISTVERSION}
+
+MAINTAINER=	yuri@FreeBSD.org
+COMMENT=	MCP server for pymupdf4llm: convert PDF to Markdown for LLMs
+WWW=		https://github.com/pymupdf/pymupdf4llm-mcp
+
+LICENSE=	AGPLv3
+LICENSE_FILE=	${WRKSRC}/LICENSE
+
+BUILD_DEPENDS=	${PYTHON_PKGNAMEPREFIX}hatchling>0:devel/py-hatchling@${PY_FLAVOR}
+RUN_DEPENDS=	${PYTHON_PKGNAMEPREFIX}mcp>=1.6.0:misc/py-mcp@${PY_FLAVOR} \
+		${PYTHON_PKGNAMEPREFIX}pydantic2>=2.11.3:devel/py-pydantic2@${PY_FLAVOR} \
+		${PYTHON_PKGNAMEPREFIX}pymupdf4llm>=0.0.21:textproc/py-pymupdf4llm@${PY_FLAVOR} \
+		${PYTHON_PKGNAMEPREFIX}typer>=0.15.2:devel/py-typer@${PY_FLAVOR}
+TEST_DEPENDS=	${PYTHON_PKGNAMEPREFIX}inline-snapshot>0:devel/py-inline-snapshot@${PY_FLAVOR} \
+		${PYTHON_PKGNAMEPREFIX}pytest-asyncio>0.25.3:devel/py-pytest-asyncio@${PY_FLAVOR}
+
+USES=		python
+USE_PYTHON=	autoplist pep517 pytest
+
+TEST_ENV=	${MAKE_ENV} PYTHONPATH=${STAGEDIR}${PYTHONPREFIX_SITELIBDIR}
+
+NO_ARCH=	yes
+
+# tests as of 0.0.4: 1 passed in 0.22s
+
+.include <bsd.port.mk>
@@ -0,0 +1,3 @@
+TIMESTAMP = 1775602664
+SHA256 (pymupdf4llm_mcp-0.0.4.tar.gz) = 90bdf45b6cc1f867a84048d4804fc781aeacab90d3af79e94795c91448f65348
+SIZE (pymupdf4llm_mcp-0.0.4.tar.gz) = 71180
@@ -0,0 +1,26 @@
+--- tests/test_tool.py.orig	2026-04-07 22:59:54 UTC
+++ tests/test_tool.py
+@@ -14,11 +14,9 @@ def test_convert_pdf_to_markdown():
+     assert result["success"] is True
+     assert "markdown_content" in result
+     assert result["markdown_content"] == snapshot("""\
+-# **Dummy PDF file**
+**Dummy PDF file**
+ 
+ 
+------
+-
+ """)
+ 
+     # temporary file to write the markdown content
+@@ -31,9 +29,7 @@ -----
+         with open(temp_file_path, encoding="utf-8") as f:
+             content = f.read()
+             assert content == snapshot("""\
+-# **Dummy PDF file**
+**Dummy PDF file**
+ 
+-
+------
+ 
+ """)
@@ -0,0 +1,10 @@
+pymupdf4llm-mcp is an MCP (Model Context Protocol) server that exposes
+pymupdf4llm functionality as tools for LLM agents. It allows AI assistants
+to convert PDF documents to Markdown format, making them suitable for
+processing by Large Language Models (LLMs) and Retrieval-Augmented Generation
+(RAG) systems.
+
+The server provides tools for:
+- Converting PDF files to Markdown with preserved structure
+- Handling tables, images, and text formatting
+- Saving converted Markdown to files