feat(extradocx): experimental DOCX → GFM Markdown AST converter

extradocx/pyproject.toml

@@ -0,0 +1,28 @@
+[project]
+name = "extradocx"
+version = "0.1.0"
+description = "Experimental DOCX → GFM Markdown AST converter"
+requires-python = ">=3.11"
+dependencies = []
+
+[project.scripts]
+extradocx = "extradocx.cli:main"
+
+[build-system]
+requires = ["hatchling"]
+build-backend = "hatchling.build"
+
+[tool.hatch.build.targets.wheel]
+packages = ["src/extradocx"]
+
+[tool.uv]
+dev-dependencies = [
+    "pytest>=8.0",
+    "ruff>=0.4",
+]
+
+[tool.ruff]
+line-length = 100
+
+[tool.ruff.lint]
+select = ["E", "F", "I"]

extradocx/src/extradocx/__init__.py

@@ -0,0 +1,24 @@
+"""
+extradocx — experimental DOCX → GFM Markdown AST converter.
+
+Proof-of-concept for bidirectional DOCX ↔ Markdown transformation via an
+intermediate AST that:
+  - Represents GFM markdown structure (headings, lists, tables, …)
+  - Preserves text at run granularity (bold, italic, … per TextRun node)
+  - Points every node back to the source DOCX XML via XPath
+
+Usage::
+
+    from extradocx import DocxParser, to_json, to_markdown
+
+    parser = DocxParser("report.docx")
+    doc = parser.parse()
+
+    json_str = to_json(doc)        # full-fidelity JSON with XPath pointers
+    md_str = to_markdown(doc)      # GFM markdown
+"""
+
+from extradocx.parser import DocxParser
+from extradocx.serializers import to_json, to_markdown
+
+__all__ = ["DocxParser", "to_json", "to_markdown"]