Merge pull request #1 from splieth/feature/drawio

Add draw.io diagram export as PNG

Author: splieth

pyproject.toml

@@ -2,6 +2,7 @@
 name = "confluence2md"
 version = "0.2.0"
 description = "Export Confluence pages to Markdown files"
+readme = "README.md"
 requires-python = ">=3.12"
 dependencies = [
     "atlassian-python-api>=3.41",

src/confluence2md/cli.py

@@ -80,14 +80,14 @@ def _handle_export(args: argparse.Namespace, config: Config) -> None:
 
     if args.page_id:
         page = fetch_page(confluence, args.page_id)
-        path = export_page(page, config.output)
+        path = export_page(page, config.output, confluence)
         print(f"Exported: {path}")
 
         include_children = args.include_children or config.output.include_children
         if include_children:
             children = fetch_child_pages(confluence, args.page_id)
             if children:
-                child_paths = export_pages(children, config.output)
+                child_paths = export_pages(children, config.output, confluence)
                 for p in child_paths:
                     print(f"Exported: {p}")
                 total = 1 + len(child_paths)
@@ -99,7 +99,7 @@ def _handle_export(args: argparse.Namespace, config: Config) -> None:
         if not pages:
             print("No pages found.")
             return
-        paths = export_pages(pages, config.output)
+        paths = export_pages(pages, config.output, confluence)
         for p in paths:
             print(f"Exported: {p}")
         print(f"\n{len(paths)} page(s) exported to {config.output.directory}")
@@ -108,7 +108,7 @@ def _handle_export(args: argparse.Namespace, config: Config) -> None:
         if not pages:
             print("No pages found.")
             return
-        paths = export_pages(pages, config.output)
+        paths = export_pages(pages, config.output, confluence)
         for p in paths:
             print(f"Exported: {p}")
         print(f"\n{len(paths)} page(s) exported to {config.output.directory}")

src/confluence2md/client.py

@@ -1,4 +1,5 @@
 from dataclasses import dataclass, field
+from pathlib import Path
 from typing import Any
 
 from atlassian import Confluence
@@ -18,6 +19,14 @@ class Page:
     parent_title: str = ""
 
 
+@dataclass
+class Attachment:
+    id: str
+    title: str
+    media_type: str
+    download_url: str
+
+
 def connect(config: ConfluenceConfig) -> Confluence:
     """Create an authenticated Confluence client."""
     if config.username:
@@ -151,3 +160,34 @@ def _build_page_url(raw: dict[str, Any], confluence: Confluence) -> str:
             base = links.get("base", confluence.url.rstrip("/"))
             return f"{base}{webui}"
     return ""
+
+
+def fetch_attachments(confluence: Confluence, page_id: str) -> list[Attachment]:
+    """Fetch all attachments for a page."""
+    results = confluence.get_attachments_from_content(page_id)  # type: ignore[no-untyped-call]
+    attachments = []
+    for item in results.get("results", []):
+        download_url = item.get("_links", {}).get("download", "")
+        attachments.append(
+            Attachment(
+                id=str(item.get("id", "")),
+                title=item.get("title", ""),
+                media_type=item.get("metadata", {}).get("mediaType", ""),
+                download_url=download_url,
+            )
+        )
+    return attachments
+
+
+def download_attachment(
+    confluence: Confluence, attachment: Attachment, dest: Path
+) -> Path:
+    """Download an attachment to the given directory. Returns the file path."""
+    dest.mkdir(parents=True, exist_ok=True)
+    filepath = dest / attachment.title
+    response = confluence.request(
+        method="GET",
+        path=attachment.download_url,
+    )
+    filepath.write_bytes(response.content)
+    return filepath

src/confluence2md/renderer.py

@@ -1,9 +1,16 @@
 import re
 from pathlib import Path
+from typing import Optional
 
+from atlassian import Confluence
 from markdownify import markdownify
 
-from .client import Page
+from .client import (
+    Attachment,
+    Page,
+    download_attachment,
+    fetch_attachments,
+)
 from .config import OutputConfig
 
 
@@ -42,27 +49,132 @@ def render_page(page: Page, config: OutputConfig) -> str:
     return "\n".join(lines)
 
 
-def export_page(page: Page, config: OutputConfig) -> Path:
+def export_page(
+    page: Page,
+    config: OutputConfig,
+    confluence: Optional[Confluence] = None,
+) -> Path:
     """Export a single page to a Markdown file. Returns the file path."""
-    content = render_page(page, config)
-    filename = _safe_filename(config.filename_pattern.format(title=page.title)) + ".md"
     output_dir = Path(config.directory)
     output_dir.mkdir(parents=True, exist_ok=True)
+
+    body = page.body
+    if confluence:
+        body = _process_drawio_macros(body, page, output_dir, confluence)
+
+    content = render_page(
+        Page(
+            id=page.id,
+            title=page.title,
+            space_key=page.space_key,
+            body=body,
+            labels=page.labels,
+            url=page.url,
+            version=page.version,
+            parent_title=page.parent_title,
+        ),
+        config,
+    )
+
+    filename = _safe_filename(config.filename_pattern.format(title=page.title)) + ".md"
     filepath = output_dir / filename
     filepath.write_text(content, encoding="utf-8")
     return filepath
 
 
-def export_pages(pages: list[Page], config: OutputConfig) -> list[Path]:
+def export_pages(
+    pages: list[Page],
+    config: OutputConfig,
+    confluence: Optional[Confluence] = None,
+) -> list[Path]:
     """Export multiple pages to Markdown files."""
-    return [export_page(page, config) for page in pages]
+    return [export_page(page, config, confluence) for page in pages]
 
 
 def _convert_body(html: str) -> str:
     """Convert Confluence storage format (HTML) to Markdown."""
     return markdownify(html, heading_style="ATX", strip=["style"])
 
 
+def _extract_drawio_diagram_names(html: str) -> list[str]:
+    """Extract draw.io diagram names from Confluence storage format HTML."""
+    names: list[str] = []
+    # Confluence macros use ac: namespace prefixes which aren't valid XML
+    # without namespace declarations, so we use regex to extract them.
+    macro_pattern = re.compile(
+        r'<ac:structured-macro[^>]*ac:name=["\']drawio["\'][^>]*>'
+        r"(.*?)</ac:structured-macro>",
+        re.DOTALL,
+    )
+    param_pattern = re.compile(
+        r'<ac:parameter[^>]*ac:name=["\']diagramName["\'][^>]*>'
+        r"(.*?)</ac:parameter>",
+        re.DOTALL,
+    )
+    for macro_match in macro_pattern.finditer(html):
+        macro_body = macro_match.group(1)
+        param_match = param_pattern.search(macro_body)
+        if param_match:
+            names.append(param_match.group(1).strip())
+    return names
+
+
+def _find_drawio_png(
+    diagram_name: str, attachments: list[Attachment]
+) -> Optional[Attachment]:
+    """Find the PNG attachment for a draw.io diagram."""
+    # draw.io stores previews with various naming conventions
+    candidates = [
+        f"{diagram_name}.png",
+        f"{diagram_name}.drawio.png",
+    ]
+    for att in attachments:
+        if att.title in candidates:
+            return att
+    return None
+
+
+def _process_drawio_macros(
+    html: str,
+    page: Page,
+    output_dir: Path,
+    confluence: Confluence,
+) -> str:
+    """Replace draw.io macros with image references and download PNGs."""
+    diagram_names = _extract_drawio_diagram_names(html)
+    if not diagram_names:
+        return html
+
+    attachments = fetch_attachments(confluence, page.id)
+
+    macro_pattern = re.compile(
+        r'<ac:structured-macro[^>]*ac:name=["\']drawio["\'][^>]*>'
+        r"(.*?)</ac:structured-macro>",
+        re.DOTALL,
+    )
+    param_pattern = re.compile(
+        r'<ac:parameter[^>]*ac:name=["\']diagramName["\'][^>]*>'
+        r"(.*?)</ac:parameter>",
+        re.DOTALL,
+    )
+
+    def _replace_macro(match: re.Match[str]) -> str:
+        macro_body = match.group(1)
+        param_match = param_pattern.search(macro_body)
+        if not param_match:
+            return match.group(0)
+
+        diagram_name = param_match.group(1).strip()
+        png_attachment = _find_drawio_png(diagram_name, attachments)
+        if not png_attachment:
+            return match.group(0)
+
+        download_attachment(confluence, png_attachment, output_dir)
+        return f'<img src="{png_attachment.title}" alt="{diagram_name}" />'
+
+    return macro_pattern.sub(_replace_macro, html)
+
+
 def _safe_filename(name: str) -> str:
     """Sanitize a string for use as a filename."""
     # Replace characters that are problematic in filenames