Mevzuat MCP

Instructions

Implementation Reference

• mevzuat_mcp_server.py:39-128 (handler)

  Main handler function for 'search_kanun' tool. Decorated with @app.tool() for registration in FastMCP server. Validates input params, creates MevzuatSearchRequestNew with type 'Kanun', calls MevzuatApiClientNew.search_documents(), handles errors, and returns MevzuatSearchResultNew.
  @app.tool() async def search_kanun( aranacak_ifade: str = Field( ..., description='Search query with optional Boolean operators: simple word (yatırımcı), AND (yatırımcı AND tazmin), OR (vergi OR ücret), NOT (yatırımcı NOT kurum), + for required (+term), grouping with (), exact phrase with quotes ("mali sıkıntı")' ), tam_cumle: bool = Field( False, description="Exact phrase match (true) or any word match (false, default). Set to true when searching for exact phrases like 'mali sıkıntı'." ), baslangic_tarihi: Optional[str] = Field( None, description="Start date for filtering results (format: DD.MM.YYYY, e.g., '01.01.2020')" ), bitis_tarihi: Optional[str] = Field( None, description="End date for filtering results (format: DD.MM.YYYY, e.g., '31.12.2024')" ), page_number: int = Field( 1, ge=1, description="Page number for pagination (starts at 1)" ), page_size: int = Field( 25, ge=1, le=100, description="Number of results per page (1-100)" ) ) -> MevzuatSearchResultNew: """ Search for Turkish laws (Kanun) in both titles and content. This tool searches in law titles and full text content. Use 'search_within_kanun' to search within a specific law's articles and get article-level results. Query Syntax: - Simple keyword: yatırımcı - Boolean AND: yatırımcı AND tazmin (both terms) - Boolean OR: yatırımcı OR müşteri (at least one term) - Boolean NOT: yatırımcı NOT kurum (first yes, second no) - Required term: +yatırımcı +tazmin (similar to AND) - Grouping: (yatırımcı OR müşteri) AND tazmin - Exact phrase: "mali sıkıntı" (or use tam_cumle=true) Returns: - Law number, title, and acceptance date - Official Gazette publication date and issue number - URLs for viewing online and downloading PDF Example queries: - "sermaye piyasası" - Find Capital Markets Law - "vergi OR gelir" - Laws with tax or income in title/content - "ceza muhakemesi" - Find Criminal Procedure Law """ search_req = MevzuatSearchRequestNew( mevzuat_tur="Kanun", aranacak_ifade=aranacak_ifade, aranacak_yer=1, # 1=Title only tam_cumle=tam_cumle, mevzuat_no=None, baslangic_tarihi=baslangic_tarihi, bitis_tarihi=bitis_tarihi, page_number=page_number, page_size=page_size ) log_params = search_req.model_dump(exclude_defaults=True) logger.info(f"Tool 'search_kanun' called with parameters: {log_params}") try: result = await mevzuat_client.search_documents(search_req) if not result.documents and not result.error_message: result.error_message = "No legislation found matching the specified criteria." return result except Exception as e: logger.exception("Error in tool 'search_mevzuat'") return MevzuatSearchResultNew( documents=[], total_results=0, current_page=page_number, page_size=page_size, total_pages=0, query_used=log_params, error_message=f"An unexpected error occurred: {str(e)}" )
• mevzuat_models.py:107-117 (schema)

  Output schema for the tool: MevzuatSearchResultNew pydantic model defining the structure of search results including list of documents, pagination info, used query, and optional error.
  class MevzuatSearchResultNew(BaseModel): """Model for search results from mevzuat.gov.tr""" documents: List[MevzuatDocumentNew] total_results: int current_page: int page_size: int total_pages: int query_used: Dict[str, Any] error_message: Optional[str] = None
• mevzuat_models.py:32-84 (schema)

  Input schema model used internally: MevzuatSearchRequestNew defines parameters for API search requests, including legislation type (set to 'Kanun'), search expression, search location, exact match flag, dates, and pagination.
  class MevzuatSearchRequestNew(BaseModel): """Request model for searching legislation on mevzuat.gov.tr""" mevzuat_tur: MevzuatTurLiteral = Field( "Kanun", description="Type of legislation. Currently only 'Kanun' (laws) are fully supported for content extraction." ) aranacak_ifade: Optional[str] = Field( None, description="Search term or phrase to look for in legislation" ) aranacak_yer: int = Field( 3, ge=1, le=3, description="Where to search: 1=Title only, 2=Article titles, 3=Full text (default)" ) tam_cumle: bool = Field( False, description="Exact phrase match (true) or any word match (false, default)" ) mevzuat_no: Optional[str] = Field( None, description="Specific legislation number to search for" ) baslangic_tarihi: Optional[str] = Field( None, description="Start date for filtering (format: DD.MM.YYYY)" ) bitis_tarihi: Optional[str] = Field( None, description="End date for filtering (format: DD.MM.YYYY)" ) page_number: int = Field( 1, ge=1, description="Page number of results" ) page_size: int = Field( 10, ge=1, le=100, description="Number of results per page" )
• mevzuat_client.py:451-556 (helper)

  Core helper method in MevzuatApiClientNew that performs the actual API search on mevzuat.gov.tr using DataTables endpoint. Handles session, payload construction (including 'YonetmelikMevzuatTur': 'OsmanliKanunu'), response parsing, and returns structured results.
  async def search_documents(self, request: MevzuatSearchRequestNew) -> MevzuatSearchResultNew: """Search for legislation documents using httpx with Playwright cookies.""" # Get session/cookies with Playwright first await self._ensure_session() # If session establishment failed, fallback to full Playwright method if not self._antiforgery_token and not self._cookies: logger.warning("Session establishment failed, using full Playwright search method as fallback") return await self.search_documents_with_playwright(request) # Build DataTables compatible payload # Normalize mevzuat type for API mevzuat_tur_api = self._normalize_mevzuat_tur_for_api(request.mevzuat_tur) payload = { "draw": 1, "columns": [ {"data": None, "name": "", "searchable": True, "orderable": False, "search": {"value": "", "regex": False}}, {"data": None, "name": "", "searchable": True, "orderable": False, "search": {"value": "", "regex": False}}, {"data": None, "name": "", "searchable": True, "orderable": False, "search": {"value": "", "regex": False}} ], "order": [], "start": (request.page_number - 1) * request.page_size, "length": request.page_size, "search": {"value": "", "regex": False}, "parameters": { "MevzuatTur": mevzuat_tur_api, "YonetmelikMevzuatTur": "OsmanliKanunu", # Required for all searches "AranacakIfade": request.aranacak_ifade or "", "TamCumle": "true" if request.tam_cumle else "false", "AranacakYer": str(request.aranacak_yer), "MevzuatNo": request.mevzuat_no or "", "KurumId": "0", "AltKurumId": "0", "BaslangicTarihi": request.baslangic_tarihi or "", "BitisTarihi": request.bitis_tarihi or "", "antiforgerytoken": self._antiforgery_token or "" } } try: # Log payload for debugging logger.debug(f"Search payload: {payload}") # Make request with cookies properly response = await self._http_client.post( self.SEARCH_ENDPOINT, json=payload, cookies=self._cookies if self._cookies else None ) # Log response for debugging if response.status_code != 200: logger.error(f"Search API error {response.status_code}: {response.text[:500]}") response.raise_for_status() data = response.json() total_results = data.get("recordsTotal", 0) documents = [] for item in data.get("data", []): doc = MevzuatDocumentNew( mevzuat_no=item.get("mevzuatNo", ""), mev_adi=item.get("mevAdi", ""), kabul_tarih=item.get("kabulTarih", ""), resmi_gazete_tarihi=item.get("resmiGazeteTarihi", ""), resmi_gazete_sayisi=item.get("resmiGazeteSayisi", ""), mevzuat_tertip=item.get("mevzuatTertip", ""), mevzuat_tur=item.get("tur", 1), url=item.get("url", "") ) documents.append(doc) return MevzuatSearchResultNew( documents=documents, total_results=total_results, current_page=request.page_number, page_size=request.page_size, total_pages=(total_results + request.page_size - 1) // request.page_size if request.page_size > 0 else 0, query_used=request.model_dump() ) except httpx.HTTPStatusError as e: logger.error(f"Search request failed: {e.response.status_code}") return MevzuatSearchResultNew( documents=[], total_results=0, current_page=request.page_number, page_size=request.page_size, total_pages=0, query_used=request.model_dump(), error_message=f"API request failed: {e.response.status_code}" ) except Exception as e: logger.exception("Unexpected error during search") return MevzuatSearchResultNew( documents=[], total_results=0, current_page=request.page_number, page_size=request.page_size, total_pages=0, query_used=request.model_dump(), error_message=f"An unexpected error occurred: {e}" )
• mevzuat_models.py:86-106 (schema)

  Schema for individual search result documents: MevzuatDocumentNew includes metadata like number, title, dates, URLs, and utility methods for PDF/web links.
  class MevzuatDocumentNew(BaseModel): """Model for a single legislation document from mevzuat.gov.tr""" mevzuat_no: str = Field(..., description="Legislation number") mev_adi: str = Field(..., description="Legislation title/name") kabul_tarih: Optional[str] = Field(None, description="Acceptance date") resmi_gazete_tarihi: Optional[str] = Field(None, description="Official Gazette publication date") resmi_gazete_sayisi: Optional[str] = Field(None, description="Official Gazette issue number") mevzuat_tertip: str = Field(..., description="Legislation series/order") mevzuat_tur: int = Field(..., description="Legislation type code") url: str = Field(..., description="Relative URL to view the legislation") def get_pdf_url(self) -> str: """Generate PDF download URL for this legislation (only works for Kanun).""" return f"https://www.mevzuat.gov.tr/MevzuatMetin/{self.mevzuat_tur}.{self.mevzuat_tertip}.{self.mevzuat_no}.pdf" def get_web_url(self) -> str: """Generate web page URL for this legislation.""" return f"https://www.mevzuat.gov.tr/{self.url}"