MediaWiki API MCP Server

"""MediaWiki page retrieval handlers for MCP server.""" import logging from collections.abc import Sequence from typing import Any import mcp.types as types from ..client import MediaWikiClient logger = logging.getLogger(__name__) async def handle_get_page( client: MediaWikiClient, arguments: dict[str, Any] ) -> Sequence[types.TextContent]: """Handle wiki_page_get tool calls with support for multiple retrieval methods.""" title = arguments.get("title") pageid = arguments.get("pageid") method = arguments.get("method", "revisions") content_format = arguments.get("format", "wikitext") sentences = arguments.get("sentences") chars = arguments.get("chars") if not title and not pageid: return [types.TextContent( type="text", text="Error: Either 'title' or 'pageid' must be provided" )] try: if method == "raw": return await _handle_raw_method(client, title, pageid) elif method == "parse": return await _handle_parse_method(client, title, pageid, content_format) elif method == "extracts": return await _handle_extracts_method(client, title, pageid, content_format, sentences, chars) else: # Default to "revisions" return await _handle_revisions_method(client, title, pageid) except Exception as e: return [types.TextContent( type="text", text=f"Error retrieving page: {str(e)}" )] async def _handle_raw_method( client: MediaWikiClient, title: str | None, pageid: int | None ) -> Sequence[types.TextContent]: """Handle raw action method for fastest wikitext retrieval.""" try: content = await client.get_page_raw(title=title, pageid=pageid) page_identifier = title or f"ID: {pageid}" return [types.TextContent( type="text", text=f"Page: {page_identifier} (Raw Wikitext)\n\n{content}" )] except Exception as e: return [types.TextContent( type="text", text=f"Error with raw method: {str(e)}" )] async def _handle_revisions_method( client: MediaWikiClient, title: str | None, pageid: int | None ) -> Sequence[types.TextContent]: """Handle revisions API method with support for both formatversion 1 and 2 response parsing.""" result = await client.get_page_info(title=title, pageid=pageid) if "query" not in result or "pages" not in result["query"]: return [types.TextContent( type="text", text="Error: Unexpected response format from MediaWiki API" )] pages = result["query"]["pages"] if not pages: return [types.TextContent( type="text", text="No pages found" )] # Handle both formatversion=1 (dict with page IDs as keys) and formatversion=2 (array) if isinstance(pages, dict): # formatversion=1: pages is a dict with page IDs as keys page_data = next(iter(pages.values())) else: # formatversion=2: pages is an array page_data = pages[0] if "missing" in page_data: page_identifier = title or f"ID: {pageid}" return [types.TextContent( type="text", text=f"Page not found: {page_identifier}" )] page_title = page_data.get("title", "Unknown") page_id = page_data.get("pageid", "Unknown") # Extract content from both formatversion structures content = "No content available" if "revisions" in page_data and page_data["revisions"]: revision = page_data["revisions"][0] # formatversion=2 structure if "slots" in revision and "main" in revision["slots"]: main_slot = revision["slots"]["main"] if "content" in main_slot: content = main_slot["content"] # formatversion=1 structure elif "*" in revision: content = revision["*"] return [types.TextContent( type="text", text=f"Page: {page_title} (ID: {page_id})\nMethod: Revisions API\nFormat: Wikitext\n\nContent:\n{content}" )] async def _handle_parse_method( client: MediaWikiClient, title: str | None, pageid: int | None, content_format: str ) -> Sequence[types.TextContent]: """Handle parse API method for HTML or wikitext content.""" try: result = await client.get_page_parse(title=title, pageid=pageid, format_type=content_format) if "parse" not in result: return [types.TextContent( type="text", text="Error: Unexpected response format from Parse API" )] parse_data = result["parse"] page_title = parse_data.get("title", "Unknown") page_id = parse_data.get("pageid", "Unknown") if content_format == "html" and "text" in parse_data: content = parse_data["text"] format_label = "HTML" elif "wikitext" in parse_data: content = parse_data["wikitext"] format_label = "Wikitext" else: content = "No content available" format_label = content_format.capitalize() return [types.TextContent( type="text", text=f"Page: {page_title} (ID: {page_id})\nMethod: Parse API\nFormat: {format_label}\n\nContent:\n{content}" )] except Exception as e: return [types.TextContent( type="text", text=f"Error with parse method: {str(e)}" )] async def _handle_extracts_method( client: MediaWikiClient, title: str | None, pageid: int | None, content_format: str, sentences: int | None, chars: int | None ) -> Sequence[types.TextContent]: """Handle TextExtracts API method for plain text extracts.""" try: plain_text = content_format == "text" result = await client.get_page_extracts( title=title, pageid=pageid, sentences=sentences, chars=chars, plain_text=plain_text ) if "query" not in result or "pages" not in result["query"]: return [types.TextContent( type="text", text="Error: Unexpected response format from TextExtracts API" )] pages = result["query"]["pages"] if not pages: return [types.TextContent( type="text", text="No pages found" )] page_data = pages[0] # formatversion=2 returns pages as array if "missing" in page_data: page_identifier = title or f"ID: {pageid}" return [types.TextContent( type="text", text=f"Page not found: {page_identifier}" )] page_title = page_data.get("title", "Unknown") page_id = page_data.get("pageid", "Unknown") extract = page_data.get("extract", "No extract available") limit_info = "" if sentences: limit_info = f" (Limited to {sentences} sentences)" elif chars: limit_info = f" (Limited to {chars} characters)" format_label = "Plain Text" if plain_text else "Limited HTML" return [types.TextContent( type="text", text=f"Page: {page_title} (ID: {page_id})\nMethod: TextExtracts API\nFormat: {format_label}{limit_info}\n\nExtract:\n{extract}" )] except Exception as e: return [types.TextContent( type="text", text=f"Error with extracts method: {str(e)}" )]