fix(aym,uyusmazlik): adapt to rebuilt AYM and Uyuşmazlık sites
Both sites were rebuilt and their old endpoints now 404:
- AYM moved to a single-page app backed by a JSON API
(POST /api/core/public/search, kararTipi NormDenetimi/BireyselBasvuru;
full text via {id, size:1} -> "icerik" HTML). Old /Ara, /ND/, /BB/ gone.
- Uyuşmazlık moved to ASP.NET WebForms (viewstate postback to /, GridView
results, decisions served as /Uploads/{EsasNo}.pdf). Old /Arama/Search gone.
Changes:
- New anayasa_mcp_module/api_client.py: shared KBB JSON client, base64url
document-id codec, HTML->Markdown + HTML text stripping helpers.
- Rewrite anayasa client/bireysel_client/unified_client over the new API,
preserving public method names and response models.
- Rewrite uyusmazlik client for the postback flow + GridView parse + PDF
document conversion; simplify request model to text + scope + paging.
- Trim search_anayasa_unified and search_uyusmazlik_decisions tool signatures
to parameters the new APIs actually support; drop dead enums.
Verified live via FastMCP client: search + document retrieval work for AYM
norm/bireysel and Uyuşmazlık.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
saidsurucu
co-authored by
Claude Opus 4.8
Claude Opus 4.8
parent
6bbc656dc6
commit
fadc3b0bc0
@@ -0,0 +1,201 @@
|
||||
# anayasa_mcp_module/api_client.py
|
||||
# Low-level client for the new Anayasa Mahkemesi "Kararlar Bilgi Bankası" (KBB) JSON API.
|
||||
#
|
||||
# Both the Norm Denetimi host (normkararlarbilgibankasi.anayasa.gov.tr) and the
|
||||
# Bireysel Başvuru host (kararlarbilgibankasi.anayasa.gov.tr) share the SAME
|
||||
# backend, exposed at POST /api/core/public/search. The request differs only by
|
||||
# the "kararTipi" discriminator:
|
||||
#
|
||||
# {"kararTipi": "NormDenetimi", "query": "mülkiyet", "page": 1, "size": 10}
|
||||
# -> {"total": N, "page": 1, "data": [...summary records...], "page_size": 10}
|
||||
#
|
||||
# {"kararTipi": "NormDenetimi", "id": "<uuid>", "page": 1, "size": 1}
|
||||
# -> data[0] additionally includes "icerik" = full decision HTML
|
||||
#
|
||||
# The previous HTML-scraping endpoints (/Ara, /ND/.., /BB/..) were retired when
|
||||
# the sites were rebuilt as a single-page app; they now return HTTP 404.
|
||||
|
||||
import base64
|
||||
import html as html_module
|
||||
import io
|
||||
import logging
|
||||
import re
|
||||
from typing import Any, Dict, Optional, Tuple
|
||||
from urllib.parse import urlparse, parse_qs, quote
|
||||
|
||||
import httpx
|
||||
from bs4 import BeautifulSoup
|
||||
from markitdown import MarkItDown
|
||||
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
# Markdown pagination chunk size (characters), shared across AYM document tools.
|
||||
DOCUMENT_MARKDOWN_CHUNK_SIZE = 5000
|
||||
|
||||
|
||||
def strip_html_text(value: Optional[str]) -> str:
|
||||
"""Return plain text from a possibly-HTML field (e.g. kararKonusu)."""
|
||||
if not value:
|
||||
return ""
|
||||
text = BeautifulSoup(html_module.unescape(value), "html.parser").get_text(" ", strip=True)
|
||||
return re.sub(r"\s+", " ", text).strip()
|
||||
|
||||
|
||||
def convert_icerik_to_markdown(icerik_html: Optional[str]) -> Optional[str]:
|
||||
"""Convert the "icerik" decision HTML returned by the KBB API to Markdown.
|
||||
|
||||
The icerik field is a self-contained HTML fragment (the rendered decision
|
||||
body). Scripts/styles are stripped before handing it to MarkItDown.
|
||||
"""
|
||||
if not icerik_html:
|
||||
return None
|
||||
|
||||
processed_html = html_module.unescape(icerik_html)
|
||||
soup = BeautifulSoup(processed_html, "html.parser")
|
||||
for tag in soup.find_all(["script", "style"]):
|
||||
tag.decompose()
|
||||
|
||||
body = soup.find("body")
|
||||
html_fragment = str(body) if body else str(soup)
|
||||
if not html_fragment.strip().lower().startswith(("<html", "<!doctype")):
|
||||
html_fragment = f'<html><head><meta charset="UTF-8"></head><body>{html_fragment}</body></html>'
|
||||
|
||||
try:
|
||||
html_stream = io.BytesIO(html_fragment.encode("utf-8"))
|
||||
conversion_result = MarkItDown().convert(html_stream)
|
||||
return conversion_result.text_content
|
||||
except Exception as e: # pragma: no cover - defensive
|
||||
logger.error("AnayasaApiClient: MarkItDown conversion error: %s", e)
|
||||
return None
|
||||
|
||||
# kararTipi discriminator values accepted by the API.
|
||||
KARAR_TIPI_NORM = "NormDenetimi"
|
||||
KARAR_TIPI_BIREYSEL = "BireyselBasvuru"
|
||||
|
||||
NORM_HOST = "https://normkararlarbilgibankasi.anayasa.gov.tr"
|
||||
BIREYSEL_HOST = "https://kararlarbilgibankasi.anayasa.gov.tr"
|
||||
SEARCH_PATH = "/api/core/public/search"
|
||||
|
||||
# Map kararTipi -> the host whose SPA can display the decision (cosmetic only;
|
||||
# either host's API answers for any kararTipi).
|
||||
_HOST_FOR_TIPI = {
|
||||
KARAR_TIPI_NORM: NORM_HOST,
|
||||
KARAR_TIPI_BIREYSEL: BIREYSEL_HOST,
|
||||
}
|
||||
|
||||
|
||||
def encode_document_token(uuid: str) -> str:
|
||||
"""Encode a raw decision UUID into the base64url token the SPA uses in its URLs.
|
||||
|
||||
The SPA addresses decisions as base64url("kbb:" + uuid) (no padding).
|
||||
"""
|
||||
raw = f"kbb:{uuid}".encode("utf-8")
|
||||
return base64.urlsafe_b64encode(raw).decode("ascii").rstrip("=")
|
||||
|
||||
|
||||
def decode_document_token(token: str) -> Optional[str]:
|
||||
"""Decode a base64url SPA token back into the raw decision UUID.
|
||||
|
||||
Returns None if the token is not a valid "kbb:<uuid>" token.
|
||||
"""
|
||||
try:
|
||||
padded = token + "=" * (-len(token) % 4)
|
||||
decoded = base64.urlsafe_b64decode(padded.encode("ascii")).decode("utf-8")
|
||||
except Exception:
|
||||
return None
|
||||
if decoded.startswith("kbb:"):
|
||||
return decoded[len("kbb:"):]
|
||||
return None
|
||||
|
||||
|
||||
def build_document_url(karar_tipi: str, uuid: str) -> str:
|
||||
"""Build a clickable SPA URL for a decision, used as its document_url."""
|
||||
host = _HOST_FOR_TIPI.get(karar_tipi, BIREYSEL_HOST)
|
||||
token = encode_document_token(uuid)
|
||||
return f"{host}/kbb/pages/search/{karar_tipi}?id={quote(token)}&type={karar_tipi}"
|
||||
|
||||
|
||||
def parse_document_url(document_url: str) -> Tuple[Optional[str], Optional[str]]:
|
||||
"""Extract (karar_tipi, uuid) from a document URL.
|
||||
|
||||
Handles the new SPA URLs (?id=<token>&type=<kararTipi>) and is lenient about
|
||||
older /ND/ and /BB/ style paths so historical references still resolve.
|
||||
Returns (None, None) if neither the type nor id can be determined.
|
||||
"""
|
||||
parsed = urlparse(document_url)
|
||||
qs = parse_qs(parsed.query)
|
||||
|
||||
karar_tipi = None
|
||||
type_param = qs.get("type", [None])[0]
|
||||
path = parsed.path or ""
|
||||
if type_param in (KARAR_TIPI_NORM, KARAR_TIPI_BIREYSEL):
|
||||
karar_tipi = type_param
|
||||
elif "/ND/" in path or "NormDenetimi" in path:
|
||||
karar_tipi = KARAR_TIPI_NORM
|
||||
elif "/BB/" in path or "BireyselBasvuru" in path:
|
||||
karar_tipi = KARAR_TIPI_BIREYSEL
|
||||
|
||||
uuid = None
|
||||
id_param = qs.get("id", [None])[0]
|
||||
if id_param:
|
||||
# The id may be the raw uuid or the base64url SPA token.
|
||||
uuid = decode_document_token(id_param) or id_param
|
||||
|
||||
return karar_tipi, uuid
|
||||
|
||||
|
||||
class AnayasaApiClient:
|
||||
"""Thin async wrapper around the KBB /api/core/public/search endpoint."""
|
||||
|
||||
def __init__(self, request_timeout: float = 60.0):
|
||||
self.http_client = httpx.AsyncClient(
|
||||
headers={
|
||||
"Accept": "application/json",
|
||||
"Content-Type": "application/json",
|
||||
"Accept-Language": "tr-TR,tr;q=0.9,en-US;q=0.8,en;q=0.7",
|
||||
"User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) "
|
||||
"AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36",
|
||||
},
|
||||
timeout=request_timeout,
|
||||
verify=True,
|
||||
follow_redirects=True,
|
||||
)
|
||||
|
||||
def _search_url(self, karar_tipi: str) -> str:
|
||||
host = _HOST_FOR_TIPI.get(karar_tipi, BIREYSEL_HOST)
|
||||
return f"{host}{SEARCH_PATH}"
|
||||
|
||||
async def search(
|
||||
self,
|
||||
karar_tipi: str,
|
||||
query: str = "",
|
||||
page: int = 1,
|
||||
size: int = 10,
|
||||
) -> Dict[str, Any]:
|
||||
"""Run a list search and return the parsed JSON envelope.
|
||||
|
||||
Envelope shape: {"total": int, "page": int, "data": [..], "page_size": int}.
|
||||
"""
|
||||
body: Dict[str, Any] = {"kararTipi": karar_tipi, "page": page, "size": size}
|
||||
if query:
|
||||
body["query"] = query
|
||||
logger.info("AnayasaApiClient: search kararTipi=%s query=%r page=%s size=%s",
|
||||
karar_tipi, query, page, size)
|
||||
response = await self.http_client.post(self._search_url(karar_tipi), json=body)
|
||||
response.raise_for_status()
|
||||
return response.json()
|
||||
|
||||
async def get_decision(self, karar_tipi: str, uuid: str) -> Optional[Dict[str, Any]]:
|
||||
"""Fetch a single decision record (including the "icerik" HTML) by UUID."""
|
||||
body = {"kararTipi": karar_tipi, "id": uuid, "page": 1, "size": 1}
|
||||
logger.info("AnayasaApiClient: get_decision kararTipi=%s id=%s", karar_tipi, uuid)
|
||||
response = await self.http_client.post(self._search_url(karar_tipi), json=body)
|
||||
response.raise_for_status()
|
||||
payload = response.json()
|
||||
data = payload.get("data") or []
|
||||
return data[0] if data else None
|
||||
|
||||
async def close(self):
|
||||
if self.http_client and not self.http_client.is_closed:
|
||||
await self.http_client.aclose()
|
||||
logger.info("AnayasaApiClient: HTTP client session closed.")
|
||||
@@ -1,24 +1,27 @@
|
||||
# anayasa_mcp_module/bireysel_client.py
|
||||
# This client is for Bireysel Başvuru: https://kararlarbilgibankasi.anayasa.gov.tr
|
||||
# Bireysel Başvuru client backed by the new KBB JSON API (see api_client.py).
|
||||
#
|
||||
# Same backend as Norm Denetimi, distinguished by kararTipi="BireyselBasvuru".
|
||||
# The legacy /Ara report-scraping endpoint was retired and now returns HTTP 404.
|
||||
|
||||
import asyncio
|
||||
import httpx
|
||||
from bs4 import BeautifulSoup, Tag
|
||||
from typing import Dict, Any, List, Optional, Tuple
|
||||
import logging
|
||||
import html
|
||||
import re
|
||||
import io
|
||||
from urllib.parse import urlencode, urljoin, quote
|
||||
from markitdown import MarkItDown
|
||||
import math # For math.ceil for pagination
|
||||
import math
|
||||
from typing import List, Optional
|
||||
|
||||
from .api_client import (
|
||||
AnayasaApiClient,
|
||||
KARAR_TIPI_BIREYSEL,
|
||||
DOCUMENT_MARKDOWN_CHUNK_SIZE,
|
||||
build_document_url,
|
||||
parse_document_url,
|
||||
convert_icerik_to_markdown,
|
||||
strip_html_text,
|
||||
)
|
||||
from .models import (
|
||||
AnayasaBireyselReportSearchRequest,
|
||||
AnayasaBireyselReportDecisionDetail,
|
||||
AnayasaBireyselReportDecisionSummary,
|
||||
AnayasaBireyselReportSearchResult,
|
||||
AnayasaBireyselBasvuruDocumentMarkdown, # Model for Bireysel Başvuru document
|
||||
AnayasaBireyselBasvuruDocumentMarkdown,
|
||||
)
|
||||
|
||||
logger = logging.getLogger(__name__)
|
||||
@@ -27,330 +30,93 @@ if not logger.hasHandlers():
|
||||
|
||||
|
||||
class AnayasaBireyselBasvuruApiClient:
|
||||
BASE_URL = "https://kararlarbilgibankasi.anayasa.gov.tr"
|
||||
SEARCH_PATH = "/Ara"
|
||||
DOCUMENT_MARKDOWN_CHUNK_SIZE = 5000 # Character limit per page
|
||||
"""Bireysel Başvuru search/document client over the KBB JSON API."""
|
||||
|
||||
def __init__(self, request_timeout: float = 60.0):
|
||||
self.http_client = httpx.AsyncClient(
|
||||
base_url=self.BASE_URL,
|
||||
headers={
|
||||
"Accept": "text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8",
|
||||
"Accept-Language": "tr-TR,tr;q=0.9,en-US;q=0.8,en;q=0.7",
|
||||
"User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36"
|
||||
},
|
||||
timeout=request_timeout,
|
||||
verify=True,
|
||||
follow_redirects=True
|
||||
)
|
||||
|
||||
def _build_query_params_for_bireysel_report(self, params: AnayasaBireyselReportSearchRequest) -> List[Tuple[str, str]]:
|
||||
query_params: List[Tuple[str, str]] = []
|
||||
query_params.append(("KararBulteni", "1")) # Specific to this report type
|
||||
|
||||
if params.keywords:
|
||||
for kw in params.keywords:
|
||||
query_params.append(("KelimeAra[]", kw))
|
||||
|
||||
if params.page_to_fetch and params.page_to_fetch > 1:
|
||||
query_params.append(("page", str(params.page_to_fetch)))
|
||||
|
||||
return query_params
|
||||
self.api = AnayasaApiClient(request_timeout)
|
||||
|
||||
async def search_bireysel_basvuru_report(
|
||||
self,
|
||||
params: AnayasaBireyselReportSearchRequest
|
||||
params: AnayasaBireyselReportSearchRequest,
|
||||
) -> AnayasaBireyselReportSearchResult:
|
||||
final_query_params = self._build_query_params_for_bireysel_report(params)
|
||||
request_url = self.SEARCH_PATH
|
||||
|
||||
logger.info(f"AnayasaBireyselBasvuruApiClient: Performing Bireysel Başvuru Report search. Path: {request_url}, Params: {final_query_params}")
|
||||
query = " ".join(t for t in (params.keywords or []) if t).strip()
|
||||
payload = await self.api.search(
|
||||
karar_tipi=KARAR_TIPI_BIREYSEL,
|
||||
query=query,
|
||||
page=params.page_to_fetch,
|
||||
size=getattr(params, "results_per_page", 10),
|
||||
)
|
||||
|
||||
try:
|
||||
response = await self.http_client.get(request_url, params=final_query_params)
|
||||
response.raise_for_status()
|
||||
html_content = response.text
|
||||
except httpx.RequestError as e:
|
||||
logger.error(f"AnayasaBireyselBasvuruApiClient: HTTP request error during Bireysel Başvuru Report search: {e}")
|
||||
raise
|
||||
except Exception as e:
|
||||
logger.error(f"AnayasaBireyselBasvuruApiClient: Error processing Bireysel Başvuru Report search request: {e}")
|
||||
raise
|
||||
|
||||
soup = BeautifulSoup(html_content, 'html.parser')
|
||||
|
||||
total_records = None
|
||||
bulunan_karar_div = soup.find("div", class_="bulunankararsayisi")
|
||||
if bulunan_karar_div:
|
||||
match_records = re.search(r'(\d+)\s*Karar Bulundu', bulunan_karar_div.get_text(strip=True))
|
||||
if match_records:
|
||||
total_records = int(match_records.group(1))
|
||||
|
||||
processed_decisions: List[AnayasaBireyselReportDecisionSummary] = []
|
||||
|
||||
report_content_area = soup.find("div", class_="HaberBulteni")
|
||||
if not report_content_area:
|
||||
logger.warning("HaberBulteni div not found, attempting to parse decision divs from the whole page.")
|
||||
report_content_area = soup
|
||||
|
||||
decision_divs = report_content_area.find_all("div", class_="KararBulteniBirKarar")
|
||||
if not decision_divs:
|
||||
logger.warning("No KararBulteniBirKarar divs found.")
|
||||
|
||||
|
||||
for decision_div in decision_divs:
|
||||
title_tag = decision_div.find("h4")
|
||||
title_text = title_tag.get_text(strip=True) if title_tag and title_tag.strong else (title_tag.get_text(strip=True) if title_tag else "")
|
||||
|
||||
|
||||
alti_cizili_div = decision_div.find("div", class_="AltiCizili")
|
||||
ref_no, dec_type, body, app_date, dec_date, url_path = "", "", "", "", "", ""
|
||||
if alti_cizili_div:
|
||||
link_tag = alti_cizili_div.find("a", href=True)
|
||||
if link_tag:
|
||||
ref_no = link_tag.get_text(strip=True)
|
||||
url_path = link_tag['href']
|
||||
|
||||
parts_text = alti_cizili_div.get_text(separator="|", strip=True)
|
||||
parts = [part.strip() for part in parts_text.split("|")]
|
||||
|
||||
# Clean ref_no from the first part if it was extracted from link
|
||||
if ref_no and parts and parts[0].strip().startswith(ref_no):
|
||||
parts[0] = parts[0].replace(ref_no, "").strip()
|
||||
if not parts[0]: parts.pop(0) # Remove empty string if ref_no was the only content
|
||||
|
||||
# Assign parts based on typical order, adjusting for missing ref_no at start
|
||||
current_idx = 0
|
||||
if not ref_no and len(parts) > current_idx and re.match(r"\d+/\d+", parts[current_idx]): # Check if first part is ref_no
|
||||
ref_no = parts[current_idx]
|
||||
current_idx += 1
|
||||
|
||||
dec_type = parts[current_idx] if len(parts) > current_idx else ""
|
||||
current_idx += 1
|
||||
body = parts[current_idx] if len(parts) > current_idx else ""
|
||||
current_idx += 1
|
||||
|
||||
app_date_raw = parts[current_idx] if len(parts) > current_idx else ""
|
||||
current_idx += 1
|
||||
dec_date_raw = parts[current_idx] if len(parts) > current_idx else ""
|
||||
|
||||
if app_date_raw and "Başvuru Tarihi :" in app_date_raw:
|
||||
app_date = app_date_raw.replace("Başvuru Tarihi :", "").strip()
|
||||
elif app_date_raw: # If label is missing but format matches
|
||||
app_date_match = re.search(r'(\d{1,2}/\d{1,2}/\d{4})', app_date_raw)
|
||||
if app_date_match: app_date = app_date_match.group(1)
|
||||
|
||||
|
||||
if dec_date_raw and "Karar Tarihi :" in dec_date_raw:
|
||||
dec_date = dec_date_raw.replace("Karar Tarihi :", "").strip()
|
||||
elif dec_date_raw: # If label is missing but format matches
|
||||
dec_date_match = re.search(r'(\d{1,2}/\d{1,2}/\d{4})', dec_date_raw)
|
||||
if dec_date_match: dec_date = dec_date_match.group(1)
|
||||
|
||||
|
||||
subject_div = decision_div.find(lambda tag: tag.name == 'div' and not tag.has_attr('class') and tag.get_text(strip=True).startswith("BAŞVURU KONUSU :"))
|
||||
subject_text = subject_div.get_text(strip=True).replace("BAŞVURU KONUSU :", "").strip() if subject_div else ""
|
||||
|
||||
details_list: List[AnayasaBireyselReportDecisionDetail] = []
|
||||
karar_detaylari_div = decision_div.find_next_sibling("div", id="KararDetaylari") # Corrected: was KararDetaylari
|
||||
if karar_detaylari_div:
|
||||
table = karar_detaylari_div.find("table", class_="table")
|
||||
if table and table.find("tbody"):
|
||||
for row in table.find("tbody").find_all("tr"):
|
||||
cells = row.find_all("td")
|
||||
if len(cells) == 4: # Hak, Müdahale İddiası, Sonuç, Giderim
|
||||
details_list.append(AnayasaBireyselReportDecisionDetail(
|
||||
hak=cells[0].get_text(strip=True) or "",
|
||||
mudahale_iddiasi=cells[1].get_text(strip=True) or "",
|
||||
sonuc=cells[2].get_text(strip=True) or "",
|
||||
giderim=cells[3].get_text(strip=True) or "",
|
||||
))
|
||||
|
||||
full_decision_page_url = urljoin(self.BASE_URL, url_path) if url_path else ""
|
||||
|
||||
processed_decisions.append(AnayasaBireyselReportDecisionSummary(
|
||||
title=title_text,
|
||||
decision_reference_no=ref_no,
|
||||
decision_page_url=full_decision_page_url,
|
||||
decision_type_summary=dec_type,
|
||||
decision_making_body=body,
|
||||
application_date_summary=app_date,
|
||||
decision_date_summary=dec_date,
|
||||
application_subject_summary=subject_text,
|
||||
details=details_list
|
||||
total_records = int(payload.get("total") or 0)
|
||||
decisions: List[AnayasaBireyselReportDecisionSummary] = []
|
||||
for item in payload.get("data") or []:
|
||||
decisions.append(AnayasaBireyselReportDecisionSummary(
|
||||
title=item.get("basvuruAdi") or "",
|
||||
decision_reference_no=item.get("basvuruNo") or "",
|
||||
decision_page_url=build_document_url(KARAR_TIPI_BIREYSEL, item.get("id", "")),
|
||||
decision_type_summary=item.get("kararTuruBasvuruSonucuLabel") or "",
|
||||
decision_making_body=item.get("kararVerenBirimLabel") or "",
|
||||
application_date_summary=item.get("basvuruTarihi") or "",
|
||||
decision_date_summary=item.get("kararTarihi") or "",
|
||||
application_subject_summary=strip_html_text(item.get("kararKonusu")),
|
||||
details=[],
|
||||
))
|
||||
|
||||
return AnayasaBireyselReportSearchResult(
|
||||
decisions=processed_decisions,
|
||||
decisions=decisions,
|
||||
total_records_found=total_records,
|
||||
retrieved_page_number=params.page_to_fetch
|
||||
retrieved_page_number=params.page_to_fetch,
|
||||
)
|
||||
|
||||
def _convert_html_to_markdown_bireysel(self, full_decision_html_content: str) -> Optional[str]:
|
||||
if not full_decision_html_content:
|
||||
return None
|
||||
|
||||
processed_html = html.unescape(full_decision_html_content)
|
||||
soup = BeautifulSoup(processed_html, "html.parser")
|
||||
html_input_for_markdown = ""
|
||||
|
||||
karar_tab_content = soup.find("div", id="Karar")
|
||||
if karar_tab_content:
|
||||
karar_html_span = karar_tab_content.find("span", class_="kararHtml")
|
||||
if karar_html_span:
|
||||
word_section = karar_html_span.find("div", class_="WordSection1")
|
||||
if word_section:
|
||||
for s in word_section.select('script, style, .item.col-xs-12.col-sm-12, center:has(b)'):
|
||||
s.decompose()
|
||||
html_input_for_markdown = str(word_section)
|
||||
else:
|
||||
logger.warning("AnayasaBireyselBasvuruApiClient: WordSection1 not found in span.kararHtml. Using span.kararHtml content.")
|
||||
for s in karar_html_span.select('script, style, .item.col-xs-12.col-sm-12, center:has(b)'):
|
||||
s.decompose()
|
||||
html_input_for_markdown = str(karar_html_span)
|
||||
else:
|
||||
logger.warning("AnayasaBireyselBasvuruApiClient: span.kararHtml not found in div#Karar. Using div#Karar content.")
|
||||
for s in karar_tab_content.select('script, style, .item.col-xs-12.col-sm-12, center:has(b)'):
|
||||
s.decompose()
|
||||
html_input_for_markdown = str(karar_tab_content)
|
||||
else:
|
||||
logger.warning("AnayasaBireyselBasvuruApiClient: div#Karar (KARAR tab) not found. Trying WordSection1 fallback.")
|
||||
word_section_fallback = soup.find("div", class_="WordSection1")
|
||||
if word_section_fallback:
|
||||
for s in word_section_fallback.select('script, style, .item.col-xs-12.col-sm-12, center:has(b)'):
|
||||
s.decompose()
|
||||
html_input_for_markdown = str(word_section_fallback)
|
||||
else:
|
||||
body_tag = soup.find("body")
|
||||
if body_tag:
|
||||
for s in body_tag.select('script, style, .item.col-xs-12.col-sm-12, center:has(b), .banner, .footer, .yazdirmaalani, .filtreler, .menu, .altmenu, .geri, .arabuton, .temizlebutonu, form#KararGetir, .TabBaslik, #KararDetaylari, .share-button-container'):
|
||||
s.decompose()
|
||||
html_input_for_markdown = str(body_tag)
|
||||
else:
|
||||
html_input_for_markdown = processed_html
|
||||
|
||||
markdown_text = None
|
||||
try:
|
||||
# Ensure the content is wrapped in basic HTML structure if it's not already
|
||||
if not html_input_for_markdown.strip().lower().startswith(("<html", "<!doctype")):
|
||||
html_content = f"<html><head><meta charset=\"UTF-8\"></head><body>{html_input_for_markdown}</body></html>"
|
||||
else:
|
||||
html_content = html_input_for_markdown
|
||||
|
||||
# Convert HTML string to bytes and create BytesIO stream
|
||||
html_bytes = html_content.encode('utf-8')
|
||||
html_stream = io.BytesIO(html_bytes)
|
||||
|
||||
# Pass BytesIO stream to MarkItDown to avoid temp file creation
|
||||
md_converter = MarkItDown()
|
||||
conversion_result = md_converter.convert(html_stream)
|
||||
markdown_text = conversion_result.text_content
|
||||
except Exception as e:
|
||||
logger.error(f"AnayasaBireyselBasvuruApiClient: MarkItDown conversion error: {e}")
|
||||
return markdown_text
|
||||
|
||||
async def get_decision_document_as_markdown(
|
||||
self,
|
||||
document_url_path: str, # e.g. /BB/2021/20295
|
||||
page_number: int = 1
|
||||
document_url_path: str,
|
||||
page_number: int = 1,
|
||||
) -> AnayasaBireyselBasvuruDocumentMarkdown:
|
||||
full_url = urljoin(self.BASE_URL, document_url_path)
|
||||
logger.info(f"AnayasaBireyselBasvuruApiClient: Fetching Bireysel Başvuru document for Markdown (page {page_number}) from URL: {full_url}")
|
||||
karar_tipi, uuid = parse_document_url(document_url_path)
|
||||
if karar_tipi is None:
|
||||
karar_tipi = KARAR_TIPI_BIREYSEL
|
||||
|
||||
basvuru_no_from_page = None
|
||||
karar_tarihi_from_page = None
|
||||
basvuru_tarihi_from_page = None
|
||||
karari_veren_birim_from_page = None
|
||||
karar_turu_from_page = None
|
||||
resmi_gazete_info_from_page = None
|
||||
|
||||
try:
|
||||
response = await self.http_client.get(full_url)
|
||||
response.raise_for_status()
|
||||
html_content_from_api = response.text
|
||||
|
||||
if not isinstance(html_content_from_api, str) or not html_content_from_api.strip():
|
||||
logger.warning(f"AnayasaBireyselBasvuruApiClient: Received empty HTML from {full_url}.")
|
||||
return AnayasaBireyselBasvuruDocumentMarkdown(
|
||||
source_url=full_url, markdown_chunk=None, current_page=page_number, total_pages=0, is_paginated=False
|
||||
)
|
||||
|
||||
soup = BeautifulSoup(html_content_from_api, 'html.parser')
|
||||
|
||||
meta_desc_tag = soup.find("meta", attrs={"name": "description"})
|
||||
if meta_desc_tag and meta_desc_tag.get("content"):
|
||||
content = meta_desc_tag["content"]
|
||||
bn_match = re.search(r"B\.\s*No:\s*([\d\/]+)", content)
|
||||
if bn_match: basvuru_no_from_page = bn_match.group(1).strip()
|
||||
|
||||
date_match = re.search(r"(\d{1,2}\/\d{1,2}\/\d{4}),\s*§", content)
|
||||
if date_match: karar_tarihi_from_page = date_match.group(1).strip()
|
||||
|
||||
karar_detaylari_tab = soup.find("div", id="KararDetaylari")
|
||||
if karar_detaylari_tab:
|
||||
table = karar_detaylari_tab.find("table", class_="table")
|
||||
if table:
|
||||
rows = table.find_all("tr")
|
||||
for row in rows:
|
||||
cells = row.find_all("td")
|
||||
if len(cells) == 2:
|
||||
key = cells[0].get_text(strip=True)
|
||||
value = cells[1].get_text(strip=True)
|
||||
if "Kararı Veren Birim" in key: karari_veren_birim_from_page = value
|
||||
elif "Karar Türü (Başvuru Sonucu)" in key: karar_turu_from_page = value
|
||||
elif "Başvuru No" in key and not basvuru_no_from_page: basvuru_no_from_page = value
|
||||
elif "Başvuru Tarihi" in key: basvuru_tarihi_from_page = value
|
||||
elif "Karar Tarihi" in key and not karar_tarihi_from_page: karar_tarihi_from_page = value
|
||||
elif "Resmi Gazete Tarih / Sayı" in key: resmi_gazete_info_from_page = value
|
||||
|
||||
full_markdown_content = await asyncio.to_thread(self._convert_html_to_markdown_bireysel, html_content_from_api)
|
||||
|
||||
if not full_markdown_content:
|
||||
return AnayasaBireyselBasvuruDocumentMarkdown(
|
||||
source_url=full_url,
|
||||
basvuru_no_from_page=basvuru_no_from_page,
|
||||
karar_tarihi_from_page=karar_tarihi_from_page,
|
||||
basvuru_tarihi_from_page=basvuru_tarihi_from_page,
|
||||
karari_veren_birim_from_page=karari_veren_birim_from_page,
|
||||
karar_turu_from_page=karar_turu_from_page,
|
||||
resmi_gazete_info_from_page=resmi_gazete_info_from_page,
|
||||
markdown_chunk=None,
|
||||
current_page=page_number,
|
||||
total_pages=0,
|
||||
is_paginated=False
|
||||
)
|
||||
|
||||
content_length = len(full_markdown_content)
|
||||
total_pages = math.ceil(content_length / self.DOCUMENT_MARKDOWN_CHUNK_SIZE)
|
||||
if total_pages == 0: total_pages = 1
|
||||
|
||||
current_page_clamped = max(1, min(page_number, total_pages))
|
||||
start_index = (current_page_clamped - 1) * self.DOCUMENT_MARKDOWN_CHUNK_SIZE
|
||||
end_index = start_index + self.DOCUMENT_MARKDOWN_CHUNK_SIZE
|
||||
markdown_chunk = full_markdown_content[start_index:end_index]
|
||||
record = await self.api.get_decision(karar_tipi, uuid) if uuid else None
|
||||
|
||||
if not record:
|
||||
logger.warning("AnayasaBireyselBasvuruApiClient: No record for %s", document_url_path)
|
||||
return AnayasaBireyselBasvuruDocumentMarkdown(
|
||||
source_url=full_url,
|
||||
basvuru_no_from_page=basvuru_no_from_page,
|
||||
karar_tarihi_from_page=karar_tarihi_from_page,
|
||||
basvuru_tarihi_from_page=basvuru_tarihi_from_page,
|
||||
karari_veren_birim_from_page=karari_veren_birim_from_page,
|
||||
karar_turu_from_page=karar_turu_from_page,
|
||||
resmi_gazete_info_from_page=resmi_gazete_info_from_page,
|
||||
markdown_chunk=markdown_chunk,
|
||||
current_page=current_page_clamped,
|
||||
total_pages=total_pages,
|
||||
is_paginated=(total_pages > 1)
|
||||
source_url=document_url_path, markdown_chunk=None,
|
||||
current_page=page_number, total_pages=0, is_paginated=False,
|
||||
)
|
||||
|
||||
except httpx.RequestError as e:
|
||||
logger.error(f"AnayasaBireyselBasvuruApiClient: HTTP error fetching Bireysel Başvuru document from {full_url}: {e}")
|
||||
raise
|
||||
except Exception as e:
|
||||
logger.error(f"AnayasaBireyselBasvuruApiClient: General error processing Bireysel Başvuru document from {full_url}: {e}")
|
||||
raise
|
||||
rg_tarihi = record.get("resmiGazeteTarihi") or ""
|
||||
rg_sayisi = record.get("resmiGazeteSayisi")
|
||||
official_gazette = f"{rg_tarihi} / {rg_sayisi}".strip(" /") if (rg_tarihi or rg_sayisi) else None
|
||||
|
||||
full_markdown = convert_icerik_to_markdown(record.get("icerik"))
|
||||
common = dict(
|
||||
source_url=document_url_path,
|
||||
basvuru_no_from_page=record.get("basvuruNo"),
|
||||
karar_tarihi_from_page=record.get("kararTarihi"),
|
||||
basvuru_tarihi_from_page=record.get("basvuruTarihi"),
|
||||
karari_veren_birim_from_page=record.get("kararVerenBirimLabel"),
|
||||
karar_turu_from_page=record.get("kararTuruBasvuruSonucuLabel"),
|
||||
resmi_gazete_info_from_page=official_gazette,
|
||||
)
|
||||
|
||||
if not full_markdown:
|
||||
return AnayasaBireyselBasvuruDocumentMarkdown(
|
||||
**common, markdown_chunk=None, current_page=page_number,
|
||||
total_pages=0, is_paginated=False,
|
||||
)
|
||||
|
||||
total_pages = max(1, math.ceil(len(full_markdown) / DOCUMENT_MARKDOWN_CHUNK_SIZE))
|
||||
current_page = max(1, min(page_number, total_pages))
|
||||
start = (current_page - 1) * DOCUMENT_MARKDOWN_CHUNK_SIZE
|
||||
chunk = full_markdown[start:start + DOCUMENT_MARKDOWN_CHUNK_SIZE]
|
||||
|
||||
return AnayasaBireyselBasvuruDocumentMarkdown(
|
||||
**common, markdown_chunk=chunk, current_page=current_page,
|
||||
total_pages=total_pages, is_paginated=(total_pages > 1),
|
||||
)
|
||||
|
||||
async def close_client_session(self):
|
||||
if hasattr(self, 'http_client') and self.http_client and not self.http_client.is_closed:
|
||||
await self.http_client.aclose()
|
||||
logger.info("AnayasaBireyselBasvuruApiClient: HTTP client session closed.")
|
||||
await self.api.close()
|
||||
logger.info("AnayasaBireyselBasvuruApiClient: HTTP client session closed.")
|
||||
|
||||
+111
-318
@@ -1,357 +1,150 @@
|
||||
# anayasa_mcp_module/client.py
|
||||
# This client is for Norm Denetimi: https://normkararlarbilgibankasi.anayasa.gov.tr
|
||||
# Norm Denetimi client backed by the new KBB JSON API (see api_client.py).
|
||||
#
|
||||
# The Anayasa Mahkemesi sites were rebuilt as a single-page app; the old
|
||||
# HTML-scraping endpoints on normkararlarbilgibankasi.anayasa.gov.tr/Ara now
|
||||
# return HTTP 404. This client maps the rich legacy request model onto the new
|
||||
# free-text "query" search and rebuilds the legacy response models from the JSON
|
||||
# payload so existing tooling keeps working.
|
||||
|
||||
import asyncio
|
||||
import httpx
|
||||
from bs4 import BeautifulSoup
|
||||
from typing import Dict, Any, List, Optional, Tuple
|
||||
import logging
|
||||
import html
|
||||
import re
|
||||
import io
|
||||
from urllib.parse import urlencode, urljoin, quote
|
||||
from markitdown import MarkItDown
|
||||
import math # For math.ceil for pagination
|
||||
import math
|
||||
from typing import List, Optional
|
||||
|
||||
from .api_client import (
|
||||
AnayasaApiClient,
|
||||
KARAR_TIPI_NORM,
|
||||
DOCUMENT_MARKDOWN_CHUNK_SIZE,
|
||||
build_document_url,
|
||||
parse_document_url,
|
||||
convert_icerik_to_markdown,
|
||||
strip_html_text,
|
||||
)
|
||||
from .models import (
|
||||
AnayasaNormDenetimiSearchRequest,
|
||||
AnayasaDecisionSummary,
|
||||
AnayasaReviewedNormInfo,
|
||||
AnayasaSearchResult,
|
||||
AnayasaDocumentMarkdown, # Model for Norm Denetimi document
|
||||
AnayasaDocumentMarkdown,
|
||||
)
|
||||
|
||||
logger = logging.getLogger(__name__)
|
||||
if not logger.hasHandlers():
|
||||
logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(name)s - %(levelname)s - %(message)s')
|
||||
|
||||
|
||||
def _build_query(params: AnayasaNormDenetimiSearchRequest) -> str:
|
||||
"""Derive the free-text query string the new API expects from the legacy model.
|
||||
|
||||
The new endpoint only supports a single full-text "query" field, so the
|
||||
keyword lists are flattened. Esas/Karar numbers are appended when no keyword
|
||||
is provided so number-based lookups still return results.
|
||||
"""
|
||||
terms: List[str] = []
|
||||
for bucket in (params.keywords_all, params.keywords_any):
|
||||
if bucket:
|
||||
terms.extend(t for t in bucket if t)
|
||||
if not terms:
|
||||
for value in (params.case_number_esas, params.decision_number_karar):
|
||||
if value:
|
||||
terms.append(value)
|
||||
return " ".join(terms).strip()
|
||||
|
||||
|
||||
class AnayasaMahkemesiApiClient:
|
||||
BASE_URL = "https://normkararlarbilgibankasi.anayasa.gov.tr"
|
||||
SEARCH_PATH_SEGMENT = "Ara"
|
||||
DOCUMENT_MARKDOWN_CHUNK_SIZE = 5000 # Character limit per page
|
||||
"""Norm Denetimi search/document client over the KBB JSON API."""
|
||||
|
||||
def __init__(self, request_timeout: float = 60.0):
|
||||
self.http_client = httpx.AsyncClient(
|
||||
base_url=self.BASE_URL,
|
||||
headers={
|
||||
"Accept": "text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8",
|
||||
"Accept-Language": "tr-TR,tr;q=0.9,en-US;q=0.8,en;q=0.7",
|
||||
"User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36"
|
||||
},
|
||||
timeout=request_timeout,
|
||||
verify=True,
|
||||
follow_redirects=True
|
||||
)
|
||||
|
||||
def _build_search_query_params_for_aym(self, params: AnayasaNormDenetimiSearchRequest) -> List[Tuple[str, str]]:
|
||||
query_params: List[Tuple[str, str]] = []
|
||||
if params.keywords_all:
|
||||
for kw in params.keywords_all: query_params.append(("KelimeAra[]", kw))
|
||||
if params.keywords_any:
|
||||
for kw in params.keywords_any: query_params.append(("HerhangiBirKelimeAra[]", kw))
|
||||
if params.keywords_exclude:
|
||||
for kw in params.keywords_exclude: query_params.append(("BulunmayanKelimeAra[]", kw))
|
||||
if params.period and params.period and params.period != "ALL": query_params.append(("Donemler_id", params.period))
|
||||
if params.case_number_esas: query_params.append(("EsasNo", params.case_number_esas))
|
||||
if params.decision_number_karar: query_params.append(("KararNo", params.decision_number_karar))
|
||||
if params.first_review_date_start: query_params.append(("IlkIncelemeTarihiIlk", params.first_review_date_start))
|
||||
if params.first_review_date_end: query_params.append(("IlkIncelemeTarihiSon", params.first_review_date_end))
|
||||
if params.decision_date_start: query_params.append(("KararTarihiIlk", params.decision_date_start))
|
||||
if params.decision_date_end: query_params.append(("KararTarihiSon", params.decision_date_end))
|
||||
if params.application_type and params.application_type and params.application_type != "ALL": query_params.append(("BasvuruTurler_id", params.application_type))
|
||||
if params.applicant_general_name: query_params.append(("BasvuranGeneller_id", params.applicant_general_name))
|
||||
if params.applicant_specific_name: query_params.append(("BasvuranOzeller_id", params.applicant_specific_name))
|
||||
if params.attending_members_names:
|
||||
for name in params.attending_members_names: query_params.append(("Uyeler_id[]", name))
|
||||
if params.rapporteur_name: query_params.append(("Raportorler_id", params.rapporteur_name))
|
||||
if params.norm_type and params.norm_type and params.norm_type != "ALL": query_params.append(("NormunTurler_id", params.norm_type))
|
||||
if params.norm_id_or_name: query_params.append(("NormunNumarasiAdlar_id", params.norm_id_or_name))
|
||||
if params.norm_article: query_params.append(("NormunMaddeNumarasi", params.norm_article))
|
||||
if params.review_outcomes:
|
||||
for outcome_val in params.review_outcomes:
|
||||
if outcome_val and outcome_val != "ALL": query_params.append(("IncelemeTuruKararSonuclar_id[]", outcome_val))
|
||||
if params.reason_for_final_outcome and params.reason_for_final_outcome and params.reason_for_final_outcome != "ALL":
|
||||
query_params.append(("KararSonucununGerekcesi", params.reason_for_final_outcome))
|
||||
if params.basis_constitution_article_numbers:
|
||||
for article_no in params.basis_constitution_article_numbers: query_params.append(("DayanakHukmu[]", article_no))
|
||||
if params.official_gazette_date_start: query_params.append(("ResmiGazeteTarihiIlk", params.official_gazette_date_start))
|
||||
if params.official_gazette_date_end: query_params.append(("ResmiGazeteTarihiSon", params.official_gazette_date_end))
|
||||
if params.official_gazette_number_start: query_params.append(("ResmiGazeteSayisiIlk", params.official_gazette_number_start))
|
||||
if params.official_gazette_number_end: query_params.append(("ResmiGazeteSayisiSon", params.official_gazette_number_end))
|
||||
if params.has_press_release and params.has_press_release and params.has_press_release != "ALL": query_params.append(("BasinDuyurusu", params.has_press_release))
|
||||
if params.has_dissenting_opinion and params.has_dissenting_opinion and params.has_dissenting_opinion != "ALL": query_params.append(("KarsiOy", params.has_dissenting_opinion))
|
||||
if params.has_different_reasoning and params.has_different_reasoning and params.has_different_reasoning != "ALL": query_params.append(("FarkliGerekce", params.has_different_reasoning))
|
||||
|
||||
# Add pagination and sorting parameters as query params instead of URL path
|
||||
if params.results_per_page and params.results_per_page != 10:
|
||||
query_params.append(("SatirSayisi", str(params.results_per_page)))
|
||||
|
||||
if params.sort_by_criteria and params.sort_by_criteria != "KararTarihi":
|
||||
query_params.append(("Siralama", params.sort_by_criteria))
|
||||
|
||||
if params.page_to_fetch and params.page_to_fetch > 1:
|
||||
query_params.append(("page", str(params.page_to_fetch)))
|
||||
return query_params
|
||||
self.api = AnayasaApiClient(request_timeout)
|
||||
|
||||
async def search_norm_denetimi_decisions(
|
||||
self,
|
||||
params: AnayasaNormDenetimiSearchRequest
|
||||
params: AnayasaNormDenetimiSearchRequest,
|
||||
) -> AnayasaSearchResult:
|
||||
# Use simple /Ara endpoint - the complex path structure seems to cause 404s
|
||||
request_path = f"/{self.SEARCH_PATH_SEGMENT}"
|
||||
|
||||
final_query_params = self._build_search_query_params_for_aym(params)
|
||||
logger.info(f"AnayasaMahkemesiApiClient: Performing Norm Denetimi search. Path: {request_path}, Params: {final_query_params}")
|
||||
query = _build_query(params)
|
||||
payload = await self.api.search(
|
||||
karar_tipi=KARAR_TIPI_NORM,
|
||||
query=query,
|
||||
page=params.page_to_fetch,
|
||||
size=params.results_per_page,
|
||||
)
|
||||
|
||||
try:
|
||||
response = await self.http_client.get(request_path, params=final_query_params)
|
||||
response.raise_for_status()
|
||||
html_content = response.text
|
||||
except httpx.RequestError as e:
|
||||
logger.error(f"AnayasaMahkemesiApiClient: HTTP request error during Norm Denetimi search: {e}")
|
||||
raise
|
||||
except Exception as e:
|
||||
logger.error(f"AnayasaMahkemesiApiClient: Error processing Norm Denetimi search request: {e}")
|
||||
raise
|
||||
|
||||
soup = BeautifulSoup(html_content, 'html.parser')
|
||||
|
||||
total_records = None
|
||||
bulunan_karar_div = soup.find("div", class_="bulunankararsayisi")
|
||||
if not bulunan_karar_div: # Fallback for mobile view
|
||||
bulunan_karar_div = soup.find("div", class_="bulunankararsayisiMobil")
|
||||
|
||||
if bulunan_karar_div:
|
||||
match_records = re.search(r'(\d+)\s*Karar Bulundu', bulunan_karar_div.get_text(strip=True))
|
||||
if match_records:
|
||||
total_records = int(match_records.group(1))
|
||||
|
||||
processed_decisions: List[AnayasaDecisionSummary] = []
|
||||
decision_divs = soup.find_all("div", class_="birkarar")
|
||||
|
||||
for decision_div in decision_divs:
|
||||
link_tag = decision_div.find("a", href=True)
|
||||
doc_url_path = link_tag['href'] if link_tag else None
|
||||
decision_page_url_str = urljoin(self.BASE_URL, doc_url_path) if doc_url_path else None
|
||||
|
||||
title_div = decision_div.find("div", class_="bkararbaslik")
|
||||
ek_no_text_raw = title_div.get_text(strip=True, separator=" ").replace('\xa0', ' ') if title_div else ""
|
||||
ek_no_match = re.search(r"(E\.\s*\d+/\d+\s*,\s*K\.\s*\d+/\d+)", ek_no_text_raw)
|
||||
ek_no_text = ek_no_match.group(1) if ek_no_match else ek_no_text_raw.split("Sayılı Karar")[0].strip()
|
||||
|
||||
keyword_count_div = title_div.find("div", class_="BulunanKelimeSayisi") if title_div else None
|
||||
keyword_count_text = keyword_count_div.get_text(strip=True).replace("Bulunan Kelime Sayısı", "").strip() if keyword_count_div else None
|
||||
keyword_count = int(keyword_count_text) if keyword_count_text and keyword_count_text.isdigit() else None
|
||||
|
||||
info_div = decision_div.find("div", class_="kararbilgileri")
|
||||
info_parts = [part.strip() for part in info_div.get_text(separator="|").split("|")] if info_div else []
|
||||
|
||||
app_type_summary = info_parts[0] if len(info_parts) > 0 else None
|
||||
applicant_summary = info_parts[1] if len(info_parts) > 1 else None
|
||||
outcome_summary = info_parts[2] if len(info_parts) > 2 else None
|
||||
dec_date_raw = info_parts[3] if len(info_parts) > 3 else None
|
||||
decision_date_summary = dec_date_raw.replace("Karar Tarihi:", "").strip() if dec_date_raw else None
|
||||
|
||||
reviewed_norms_list: List[AnayasaReviewedNormInfo] = []
|
||||
details_table_container = decision_div.find_next_sibling("div", class_=re.compile(r"col-sm-12")) # The details table is in a sibling div
|
||||
if details_table_container:
|
||||
details_table = details_table_container.find("table", class_="table")
|
||||
if details_table and details_table.find("tbody"):
|
||||
for row in details_table.find("tbody").find_all("tr"):
|
||||
cells = row.find_all("td")
|
||||
if len(cells) == 6:
|
||||
reviewed_norms_list.append(AnayasaReviewedNormInfo(
|
||||
norm_name_or_number=cells[0].get_text(strip=True) or None,
|
||||
article_number=cells[1].get_text(strip=True) or None,
|
||||
review_type_and_outcome=cells[2].get_text(strip=True) or None,
|
||||
outcome_reason=cells[3].get_text(strip=True) or None,
|
||||
basis_constitution_articles_cited=[a.strip() for a in cells[4].get_text(strip=True).split(',') if a.strip()] if cells[4].get_text(strip=True) else [],
|
||||
postponement_period=cells[5].get_text(strip=True) or None
|
||||
))
|
||||
|
||||
processed_decisions.append(AnayasaDecisionSummary(
|
||||
decision_reference_no=ek_no_text,
|
||||
decision_page_url=decision_page_url_str,
|
||||
keywords_found_count=keyword_count,
|
||||
application_type_summary=app_type_summary,
|
||||
applicant_summary=applicant_summary,
|
||||
decision_outcome_summary=outcome_summary,
|
||||
decision_date_summary=decision_date_summary,
|
||||
reviewed_norms=reviewed_norms_list
|
||||
total_records = int(payload.get("total") or 0)
|
||||
decisions: List[AnayasaDecisionSummary] = []
|
||||
for item in payload.get("data") or []:
|
||||
esas_no = item.get("esasNo") or ""
|
||||
karar_no = item.get("kararNo") or ""
|
||||
if esas_no and karar_no:
|
||||
reference = f"E.{esas_no}, K.{karar_no}"
|
||||
else:
|
||||
reference = esas_no or karar_no or ""
|
||||
decisions.append(AnayasaDecisionSummary(
|
||||
decision_reference_no=reference,
|
||||
decision_page_url=build_document_url(KARAR_TIPI_NORM, item.get("id", "")),
|
||||
keywords_found_count=item.get("highlightCount") or 0,
|
||||
application_type_summary=item.get("basvuruTuruLabel") or "",
|
||||
applicant_summary=item.get("basvuranGenelLabel") or "",
|
||||
decision_outcome_summary=strip_html_text(item.get("kararKonusu")),
|
||||
decision_date_summary=item.get("kararTarihi") or "",
|
||||
reviewed_norms=[],
|
||||
))
|
||||
|
||||
return AnayasaSearchResult(
|
||||
decisions=processed_decisions,
|
||||
decisions=decisions,
|
||||
total_records_found=total_records,
|
||||
retrieved_page_number=params.page_to_fetch
|
||||
retrieved_page_number=params.page_to_fetch,
|
||||
)
|
||||
|
||||
def _convert_html_to_markdown_norm_denetimi(self, full_decision_html_content: str) -> Optional[str]:
|
||||
"""Converts direct HTML content from an Anayasa Mahkemesi Norm Denetimi decision page to Markdown."""
|
||||
if not full_decision_html_content:
|
||||
return None
|
||||
|
||||
processed_html = html.unescape(full_decision_html_content)
|
||||
soup = BeautifulSoup(processed_html, "html.parser")
|
||||
html_input_for_markdown = ""
|
||||
|
||||
karar_tab_content = soup.find("div", id="Karar") # "KARAR" tab content
|
||||
if karar_tab_content:
|
||||
karar_metni_div = karar_tab_content.find("div", class_="KararMetni")
|
||||
if karar_metni_div:
|
||||
# Remove scripts and styles
|
||||
for script_tag in karar_metni_div.find_all("script"): script_tag.decompose()
|
||||
for style_tag in karar_metni_div.find_all("style"): style_tag.decompose()
|
||||
# Remove "Künye Kopyala" button and other non-content divs
|
||||
for item_div in karar_metni_div.find_all("div", class_="item col-sm-12"): item_div.decompose()
|
||||
for modal_div in karar_metni_div.find_all("div", class_="modal fade"): modal_div.decompose() # If any modals
|
||||
|
||||
word_section = karar_metni_div.find("div", class_="WordSection1")
|
||||
html_input_for_markdown = str(word_section) if word_section else str(karar_metni_div)
|
||||
else:
|
||||
html_input_for_markdown = str(karar_tab_content)
|
||||
else:
|
||||
# Fallback if specific structure is not found
|
||||
word_section_fallback = soup.find("div", class_="WordSection1")
|
||||
if word_section_fallback:
|
||||
html_input_for_markdown = str(word_section_fallback)
|
||||
else:
|
||||
# Last resort: use the whole body or the raw HTML
|
||||
body_tag = soup.find("body")
|
||||
html_input_for_markdown = str(body_tag) if body_tag else processed_html
|
||||
|
||||
markdown_text = None
|
||||
try:
|
||||
# Ensure the content is wrapped in basic HTML structure if it's not already
|
||||
if not html_input_for_markdown.strip().lower().startswith(("<html", "<!doctype")):
|
||||
html_content = f"<html><head><meta charset=\"UTF-8\"></head><body>{html_input_for_markdown}</body></html>"
|
||||
else:
|
||||
html_content = html_input_for_markdown
|
||||
|
||||
# Convert HTML string to bytes and create BytesIO stream
|
||||
html_bytes = html_content.encode('utf-8')
|
||||
html_stream = io.BytesIO(html_bytes)
|
||||
|
||||
# Pass BytesIO stream to MarkItDown to avoid temp file creation
|
||||
md_converter = MarkItDown()
|
||||
conversion_result = md_converter.convert(html_stream)
|
||||
markdown_text = conversion_result.text_content
|
||||
except Exception as e:
|
||||
logger.error(f"AnayasaMahkemesiApiClient: MarkItDown conversion error: {e}")
|
||||
return markdown_text
|
||||
|
||||
async def get_decision_document_as_markdown(
|
||||
self,
|
||||
document_url: str,
|
||||
page_number: int = 1
|
||||
page_number: int = 1,
|
||||
) -> AnayasaDocumentMarkdown:
|
||||
"""
|
||||
Retrieves a specific Anayasa Mahkemesi (Norm Denetimi) decision,
|
||||
converts its content to Markdown, and returns the requested page/chunk.
|
||||
"""
|
||||
full_url = urljoin(self.BASE_URL, document_url) if not document_url.startswith("http") else document_url
|
||||
logger.info(f"AnayasaMahkemesiApiClient: Fetching Norm Denetimi document for Markdown (page {page_number}) from URL: {full_url}")
|
||||
karar_tipi, uuid = parse_document_url(document_url)
|
||||
if karar_tipi is None:
|
||||
karar_tipi = KARAR_TIPI_NORM
|
||||
|
||||
decision_ek_no_from_page = None
|
||||
decision_date_from_page = None
|
||||
official_gazette_from_page = None
|
||||
|
||||
try:
|
||||
# Use a new client instance for document fetching if headers/timeout needs to be different,
|
||||
# or reuse self.http_client if settings are compatible. For now, self.http_client.
|
||||
get_response = await self.http_client.get(full_url, headers={"Accept": "text/html"})
|
||||
get_response.raise_for_status()
|
||||
html_content_from_api = get_response.text
|
||||
|
||||
if not isinstance(html_content_from_api, str) or not html_content_from_api.strip():
|
||||
logger.warning(f"AnayasaMahkemesiApiClient: Received empty or non-string HTML from URL {full_url}.")
|
||||
return AnayasaDocumentMarkdown(
|
||||
source_url=full_url, markdown_chunk=None, current_page=page_number, total_pages=0, is_paginated=False
|
||||
)
|
||||
|
||||
# Extract metadata from the page content (E.K. No, Date, RG)
|
||||
soup = BeautifulSoup(html_content_from_api, "html.parser")
|
||||
karar_metni_div = soup.find("div", class_="KararMetni") # Usually within div#Karar
|
||||
if not karar_metni_div: # Fallback if not in KararMetni
|
||||
karar_metni_div = soup.find("div", class_="WordSection1")
|
||||
|
||||
# Initialize with empty string defaults
|
||||
decision_ek_no_from_page = ""
|
||||
decision_date_from_page = ""
|
||||
official_gazette_from_page = ""
|
||||
|
||||
if karar_metni_div:
|
||||
# Attempt to find E.K. No (Esas No, Karar No)
|
||||
# Norm Denetimi pages often have this in bold <p> tags directly or in the WordSection1
|
||||
# Look for patterns like "Esas No.: YYYY/NN" and "Karar No.: YYYY/NN"
|
||||
|
||||
esas_no_tag = karar_metni_div.find(lambda tag: tag.name == "p" and tag.find("b") and "Esas No.:" in tag.find("b").get_text())
|
||||
karar_no_tag = karar_metni_div.find(lambda tag: tag.name == "p" and tag.find("b") and "Karar No.:" in tag.find("b").get_text())
|
||||
karar_tarihi_tag = karar_metni_div.find(lambda tag: tag.name == "p" and tag.find("b") and "Karar tarihi:" in tag.find("b").get_text()) # Less common on Norm pages
|
||||
resmi_gazete_tag = karar_metni_div.find(lambda tag: tag.name == "p" and ("Resmî Gazete tarih ve sayısı:" in tag.get_text() or "Resmi Gazete tarih/sayı:" in tag.get_text()))
|
||||
|
||||
|
||||
if esas_no_tag and esas_no_tag.find("b") and karar_no_tag and karar_no_tag.find("b"):
|
||||
esas_str = esas_no_tag.find("b").get_text(strip=True).replace('Esas No.:', '').strip()
|
||||
karar_str = karar_no_tag.find("b").get_text(strip=True).replace('Karar No.:', '').strip()
|
||||
decision_ek_no_from_page = f"E.{esas_str}, K.{karar_str}"
|
||||
|
||||
if karar_tarihi_tag and karar_tarihi_tag.find("b"):
|
||||
decision_date_from_page = karar_tarihi_tag.find("b").get_text(strip=True).replace("Karar tarihi:", "").strip()
|
||||
elif karar_metni_div: # Fallback for Karar Tarihi if not in specific tag
|
||||
date_match = re.search(r"Karar Tarihi\s*:\s*([\d\.]+)", karar_metni_div.get_text()) # Norm pages often use DD.MM.YYYY
|
||||
if date_match: decision_date_from_page = date_match.group(1).strip()
|
||||
|
||||
|
||||
if resmi_gazete_tag:
|
||||
# Try to get the bold part first if it exists
|
||||
bold_rg_tag = resmi_gazete_tag.find("b")
|
||||
rg_text_content = bold_rg_tag.get_text(strip=True) if bold_rg_tag else resmi_gazete_tag.get_text(strip=True)
|
||||
official_gazette_from_page = rg_text_content.replace("Resmî Gazete tarih ve sayısı:", "").replace("Resmi Gazete tarih/sayı:", "").strip()
|
||||
|
||||
|
||||
full_markdown_content = await asyncio.to_thread(self._convert_html_to_markdown_norm_denetimi, html_content_from_api)
|
||||
|
||||
if not full_markdown_content:
|
||||
return AnayasaDocumentMarkdown(
|
||||
source_url=full_url,
|
||||
decision_reference_no_from_page=decision_ek_no_from_page,
|
||||
decision_date_from_page=decision_date_from_page,
|
||||
official_gazette_info_from_page=official_gazette_from_page,
|
||||
markdown_chunk=None,
|
||||
current_page=page_number,
|
||||
total_pages=0,
|
||||
is_paginated=False
|
||||
)
|
||||
|
||||
content_length = len(full_markdown_content)
|
||||
total_pages = math.ceil(content_length / self.DOCUMENT_MARKDOWN_CHUNK_SIZE)
|
||||
if total_pages == 0: total_pages = 1
|
||||
|
||||
current_page_clamped = max(1, min(page_number, total_pages))
|
||||
start_index = (current_page_clamped - 1) * self.DOCUMENT_MARKDOWN_CHUNK_SIZE
|
||||
end_index = start_index + self.DOCUMENT_MARKDOWN_CHUNK_SIZE
|
||||
markdown_chunk = full_markdown_content[start_index:end_index]
|
||||
record = await self.api.get_decision(karar_tipi, uuid) if uuid else None
|
||||
|
||||
if not record:
|
||||
logger.warning("AnayasaMahkemesiApiClient: No record for document_url %s", document_url)
|
||||
return AnayasaDocumentMarkdown(
|
||||
source_url=full_url,
|
||||
decision_reference_no_from_page=decision_ek_no_from_page,
|
||||
decision_date_from_page=decision_date_from_page,
|
||||
official_gazette_info_from_page=official_gazette_from_page,
|
||||
markdown_chunk=markdown_chunk,
|
||||
current_page=current_page_clamped,
|
||||
total_pages=total_pages,
|
||||
is_paginated=(total_pages > 1)
|
||||
source_url=document_url, markdown_chunk=None,
|
||||
current_page=page_number, total_pages=0, is_paginated=False,
|
||||
)
|
||||
|
||||
except httpx.RequestError as e:
|
||||
logger.error(f"AnayasaMahkemesiApiClient: HTTP error fetching Norm Denetimi document from {full_url}: {e}")
|
||||
raise
|
||||
except Exception as e:
|
||||
logger.error(f"AnayasaMahkemesiApiClient: General error processing Norm Denetimi document from {full_url}: {e}")
|
||||
raise
|
||||
esas_no = record.get("esasNo") or ""
|
||||
karar_no = record.get("kararNo") or ""
|
||||
reference = f"E.{esas_no}, K.{karar_no}" if (esas_no and karar_no) else (esas_no or karar_no or "")
|
||||
rg_tarihi = record.get("resmiGazeteTarihi") or ""
|
||||
rg_sayisi = record.get("resmiGazeteSayisi")
|
||||
official_gazette = f"{rg_tarihi} / {rg_sayisi}".strip(" /") if (rg_tarihi or rg_sayisi) else ""
|
||||
|
||||
full_markdown = convert_icerik_to_markdown(record.get("icerik"))
|
||||
if not full_markdown:
|
||||
return AnayasaDocumentMarkdown(
|
||||
source_url=document_url,
|
||||
decision_reference_no_from_page=reference,
|
||||
decision_date_from_page=record.get("kararTarihi") or "",
|
||||
official_gazette_info_from_page=official_gazette,
|
||||
markdown_chunk=None, current_page=page_number, total_pages=0, is_paginated=False,
|
||||
)
|
||||
|
||||
total_pages = max(1, math.ceil(len(full_markdown) / DOCUMENT_MARKDOWN_CHUNK_SIZE))
|
||||
current_page = max(1, min(page_number, total_pages))
|
||||
start = (current_page - 1) * DOCUMENT_MARKDOWN_CHUNK_SIZE
|
||||
chunk = full_markdown[start:start + DOCUMENT_MARKDOWN_CHUNK_SIZE]
|
||||
|
||||
return AnayasaDocumentMarkdown(
|
||||
source_url=document_url,
|
||||
decision_reference_no_from_page=reference,
|
||||
decision_date_from_page=record.get("kararTarihi") or "",
|
||||
official_gazette_info_from_page=official_gazette,
|
||||
markdown_chunk=chunk,
|
||||
current_page=current_page,
|
||||
total_pages=total_pages,
|
||||
is_paginated=(total_pages > 1),
|
||||
)
|
||||
|
||||
async def close_client_session(self):
|
||||
if hasattr(self, 'http_client') and self.http_client and not self.http_client.is_closed:
|
||||
await self.http_client.aclose()
|
||||
logger.info("AnayasaMahkemesiApiClient (Norm Denetimi): HTTP client session closed.")
|
||||
await self.api.close()
|
||||
logger.info("AnayasaMahkemesiApiClient (Norm Denetimi): HTTP client session closed.")
|
||||
|
||||
@@ -140,9 +140,10 @@ class AnayasaDocumentMarkdown(BaseModel):
|
||||
# --- Models for Anayasa Mahkemesi - Bireysel Başvuru Karar Raporu ---
|
||||
|
||||
class AnayasaBireyselReportSearchRequest(BaseModel):
|
||||
"""Model for Anayasa Mahkemesi (Bireysel Başvuru) 'Karar Arama Raporu' search request."""
|
||||
keywords: Optional[List[str]] = Field(default_factory=list, description="Keywords for AND logic (KelimeAra[]).")
|
||||
"""Model for Anayasa Mahkemesi (Bireysel Başvuru) search request."""
|
||||
keywords: Optional[List[str]] = Field(default_factory=list, description="Keywords joined into the full-text query.")
|
||||
page_to_fetch: int = Field(1, ge=1, description="Page number to fetch for the report (page). Default is 1.")
|
||||
results_per_page: int = Field(10, ge=1, le=100, description="Results per page.")
|
||||
|
||||
class AnayasaBireyselReportDecisionDetail(BaseModel):
|
||||
"""Details of a specific right/claim within a Bireysel Başvuru decision summary in a report."""
|
||||
@@ -191,26 +192,19 @@ class AnayasaBireyselBasvuruDocumentMarkdown(BaseModel):
|
||||
|
||||
# --- Unified Models ---
|
||||
class AnayasaUnifiedSearchRequest(BaseModel):
|
||||
"""Unified search request for both Norm Denetimi and Bireysel Başvuru."""
|
||||
"""Unified search request for both Norm Denetimi and Bireysel Başvuru.
|
||||
|
||||
The KBB API only exposes a single free-text "query" field plus pagination,
|
||||
so the keyword lists below are flattened into that query.
|
||||
"""
|
||||
decision_type: Literal["norm_denetimi", "bireysel_basvuru"] = Field(..., description="Decision type: norm_denetimi or bireysel_basvuru")
|
||||
|
||||
|
||||
# Common parameters
|
||||
keywords: List[str] = Field(default_factory=list, description="Keywords to search for")
|
||||
keywords: List[str] = Field(default_factory=list, description="Keywords to search for (joined into a single full-text query)")
|
||||
keywords_all: List[str] = Field(default_factory=list, description="Additional keywords to include in the query")
|
||||
keywords_any: List[str] = Field(default_factory=list, description="Additional alternative keywords to include in the query")
|
||||
page_to_fetch: int = Field(1, ge=1, le=100, description="Page number to fetch (1-100)")
|
||||
results_per_page: int = Field(10, ge=1, le=100, description="Results per page (1-100)")
|
||||
|
||||
# Norm Denetimi specific parameters (ignored for bireysel_basvuru)
|
||||
keywords_all: List[str] = Field(default_factory=list, description="All keywords must be present (norm_denetimi only)")
|
||||
keywords_any: List[str] = Field(default_factory=list, description="Any of these keywords (norm_denetimi only)")
|
||||
decision_type_norm: Literal["ALL", "1", "2", "3"] = Field("ALL", description="Decision type for norm denetimi")
|
||||
application_date_start: str = Field("", description="Application start date (norm_denetimi only)")
|
||||
application_date_end: str = Field("", description="Application end date (norm_denetimi only)")
|
||||
|
||||
# Bireysel Başvuru specific parameters (ignored for norm_denetimi)
|
||||
decision_start_date: str = Field("", description="Decision start date (bireysel_basvuru only)")
|
||||
decision_end_date: str = Field("", description="Decision end date (bireysel_basvuru only)")
|
||||
norm_type: Literal["ALL", "1", "2", "3", "4", "5", "6", "7", "8", "9", "10", "11", "12", "13", "14", "0"] = Field("ALL", description="Norm type (bireysel_basvuru only)")
|
||||
subject_category: str = Field("", description="Subject category (bireysel_basvuru only)")
|
||||
|
||||
class AnayasaUnifiedSearchResult(BaseModel):
|
||||
"""Unified search result containing decisions from either system."""
|
||||
|
||||
@@ -1,172 +1,118 @@
|
||||
# anayasa_mcp_module/unified_client.py
|
||||
# Unified client for both Norm Denetimi and Bireysel Başvuru
|
||||
# Unified client for both Norm Denetimi and Bireysel Başvuru, backed by the new
|
||||
# KBB JSON API. Routing between the two is by the "decision_type" discriminator
|
||||
# on search, and by the document URL (?type=...) on document retrieval.
|
||||
|
||||
import logging
|
||||
from typing import Optional, Tuple
|
||||
from urllib.parse import urlparse, urlunparse
|
||||
|
||||
from .models import (
|
||||
AnayasaUnifiedSearchRequest,
|
||||
AnayasaUnifiedSearchResult,
|
||||
AnayasaUnifiedSearchResult,
|
||||
AnayasaUnifiedDocumentMarkdown,
|
||||
# Removed AnayasaDecisionTypeEnum - now using string literals
|
||||
AnayasaNormDenetimiSearchRequest,
|
||||
AnayasaBireyselReportSearchRequest
|
||||
AnayasaBireyselReportSearchRequest,
|
||||
)
|
||||
from .client import AnayasaMahkemesiApiClient
|
||||
from .bireysel_client import AnayasaBireyselBasvuruApiClient
|
||||
from .api_client import (
|
||||
KARAR_TIPI_NORM,
|
||||
KARAR_TIPI_BIREYSEL,
|
||||
parse_document_url,
|
||||
)
|
||||
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
# Canonical hosts per decision type. Norm Denetimi (/ND/) documents live on the
|
||||
# "norm" subdomain; Bireysel Başvuru (/BB/) documents on the plain subdomain.
|
||||
# Callers (or upstream search links) sometimes supply the wrong host for a given
|
||||
# path, which makes the AYM server return 404. We re-key the host off the path.
|
||||
_NORM_HOST = "normkararlarbilgibankasi.anayasa.gov.tr"
|
||||
_BIREYSEL_HOST = "kararlarbilgibankasi.anayasa.gov.tr"
|
||||
|
||||
|
||||
def normalize_anayasa_document_url(document_url: str) -> Tuple[Optional[str], str]:
|
||||
"""Detect the AYM decision type from the URL path and force the correct host.
|
||||
"""Detect the AYM decision type from a document URL.
|
||||
|
||||
Detection is path-based (``/ND/`` vs ``/BB/``) because the path is
|
||||
unambiguous, whereas the supplied host may be wrong. Query params and
|
||||
fragment are preserved (they are harmless for document fetches).
|
||||
|
||||
Returns ``(decision_type, normalized_url)`` where ``decision_type`` is
|
||||
Returns ``(decision_type, document_url)`` where ``decision_type`` is
|
||||
``"norm_denetimi"``, ``"bireysel_basvuru"``, or ``None`` if it cannot be
|
||||
determined (URL returned unchanged in that case).
|
||||
determined. The URL is returned unchanged (kept for backwards compatibility
|
||||
with callers that expect a possibly-normalized URL).
|
||||
"""
|
||||
parsed = urlparse(document_url)
|
||||
path = parsed.path or ""
|
||||
|
||||
if "/ND/" in path:
|
||||
decision_type, host = "norm_denetimi", _NORM_HOST
|
||||
elif "/BB/" in path:
|
||||
decision_type, host = "bireysel_basvuru", _BIREYSEL_HOST
|
||||
else:
|
||||
# Fall back to host-based detection when the path is uninformative.
|
||||
if "normkararlarbilgibankasi" in parsed.netloc:
|
||||
return "norm_denetimi", document_url
|
||||
if "kararlarbilgibankasi" in parsed.netloc:
|
||||
return "bireysel_basvuru", document_url
|
||||
return None, document_url
|
||||
|
||||
normalized = urlunparse((
|
||||
parsed.scheme or "https",
|
||||
host,
|
||||
parsed.path,
|
||||
parsed.params,
|
||||
parsed.query,
|
||||
parsed.fragment,
|
||||
))
|
||||
return decision_type, normalized
|
||||
karar_tipi, _ = parse_document_url(document_url)
|
||||
if karar_tipi == KARAR_TIPI_NORM:
|
||||
return "norm_denetimi", document_url
|
||||
if karar_tipi == KARAR_TIPI_BIREYSEL:
|
||||
return "bireysel_basvuru", document_url
|
||||
return None, document_url
|
||||
|
||||
|
||||
class AnayasaUnifiedClient:
|
||||
"""Unified client that handles both Norm Denetimi and Bireysel Başvuru searches."""
|
||||
|
||||
|
||||
def __init__(self, request_timeout: float = 60.0):
|
||||
self.norm_client = AnayasaMahkemesiApiClient(request_timeout)
|
||||
self.bireysel_client = AnayasaBireyselBasvuruApiClient(request_timeout)
|
||||
|
||||
|
||||
async def search_unified(self, params: AnayasaUnifiedSearchRequest) -> AnayasaUnifiedSearchResult:
|
||||
"""Unified search that routes to appropriate client based on decision_type."""
|
||||
|
||||
"""Unified search that routes to the appropriate client based on decision_type."""
|
||||
|
||||
if params.decision_type == "norm_denetimi":
|
||||
# Convert to norm denetimi request
|
||||
norm_params = AnayasaNormDenetimiSearchRequest(
|
||||
keywords_all=params.keywords_all or params.keywords,
|
||||
keywords_any=params.keywords_any,
|
||||
application_type=params.decision_type_norm,
|
||||
page_to_fetch=params.page_to_fetch,
|
||||
results_per_page=params.results_per_page
|
||||
results_per_page=params.results_per_page,
|
||||
)
|
||||
|
||||
result = await self.norm_client.search_norm_denetimi_decisions(norm_params)
|
||||
|
||||
# Convert to unified format
|
||||
decisions_list = [decision.model_dump() for decision in result.decisions]
|
||||
|
||||
|
||||
return AnayasaUnifiedSearchResult(
|
||||
decision_type="norm_denetimi",
|
||||
decisions=decisions_list,
|
||||
decisions=[d.model_dump() for d in result.decisions],
|
||||
total_records_found=result.total_records_found,
|
||||
retrieved_page_number=result.retrieved_page_number
|
||||
retrieved_page_number=result.retrieved_page_number,
|
||||
)
|
||||
|
||||
|
||||
elif params.decision_type == "bireysel_basvuru":
|
||||
# Convert to bireysel başvuru request
|
||||
bireysel_params = AnayasaBireyselReportSearchRequest(
|
||||
keywords=params.keywords,
|
||||
decision_start_date=params.decision_start_date,
|
||||
decision_end_date=params.decision_end_date,
|
||||
norm_type=params.norm_type,
|
||||
subject_category=params.subject_category,
|
||||
keywords=params.keywords or params.keywords_all,
|
||||
page_to_fetch=params.page_to_fetch,
|
||||
results_per_page=params.results_per_page
|
||||
results_per_page=params.results_per_page,
|
||||
)
|
||||
|
||||
result = await self.bireysel_client.search_bireysel_basvuru_report(bireysel_params)
|
||||
|
||||
# Convert to unified format
|
||||
decisions_list = [decision.model_dump() for decision in result.decisions]
|
||||
|
||||
|
||||
return AnayasaUnifiedSearchResult(
|
||||
decision_type="bireysel_basvuru",
|
||||
decisions=decisions_list,
|
||||
decisions=[d.model_dump() for d in result.decisions],
|
||||
total_records_found=result.total_records_found,
|
||||
retrieved_page_number=result.retrieved_page_number
|
||||
retrieved_page_number=result.retrieved_page_number,
|
||||
)
|
||||
|
||||
else:
|
||||
raise ValueError(f"Unsupported decision type: {params.decision_type}")
|
||||
|
||||
|
||||
raise ValueError(f"Unsupported decision type: {params.decision_type}")
|
||||
|
||||
async def get_document_unified(self, document_url: str, page_number: int = 1) -> AnayasaUnifiedDocumentMarkdown:
|
||||
"""Unified document retrieval that auto-detects the appropriate client."""
|
||||
|
||||
# Auto-detect decision type from the path and force the correct host.
|
||||
# This repairs malformed URLs (e.g. a /ND/ path on the bireysel host),
|
||||
# which otherwise 404 against the AYM server.
|
||||
decision_type, normalized_url = normalize_anayasa_document_url(document_url)
|
||||
if normalized_url != document_url:
|
||||
logger.info(
|
||||
f"AnayasaUnifiedClient: Normalized document URL "
|
||||
f"'{document_url}' -> '{normalized_url}'"
|
||||
)
|
||||
"""Unified document retrieval that auto-detects the decision type from the URL."""
|
||||
|
||||
if decision_type == "norm_denetimi":
|
||||
result = await self.norm_client.get_decision_document_as_markdown(normalized_url, page_number)
|
||||
|
||||
return AnayasaUnifiedDocumentMarkdown(
|
||||
decision_type="norm_denetimi",
|
||||
source_url=result.source_url,
|
||||
document_data=result.model_dump(),
|
||||
markdown_chunk=result.markdown_chunk,
|
||||
current_page=result.current_page,
|
||||
total_pages=result.total_pages,
|
||||
is_paginated=result.is_paginated
|
||||
)
|
||||
|
||||
elif decision_type == "bireysel_basvuru":
|
||||
result = await self.bireysel_client.get_decision_document_as_markdown(normalized_url, page_number)
|
||||
decision_type, _ = normalize_anayasa_document_url(document_url)
|
||||
|
||||
if decision_type == "bireysel_basvuru":
|
||||
result = await self.bireysel_client.get_decision_document_as_markdown(document_url, page_number)
|
||||
return AnayasaUnifiedDocumentMarkdown(
|
||||
decision_type="bireysel_basvuru",
|
||||
source_url=result.source_url,
|
||||
document_data=result.model_dump(),
|
||||
document_data=result.model_dump(mode="json"),
|
||||
markdown_chunk=result.markdown_chunk,
|
||||
current_page=result.current_page,
|
||||
total_pages=result.total_pages,
|
||||
is_paginated=result.is_paginated
|
||||
is_paginated=result.is_paginated,
|
||||
)
|
||||
|
||||
else:
|
||||
raise ValueError(f"Cannot determine document type from URL: {document_url}")
|
||||
|
||||
# Default to norm_denetimi (also covers explicit norm_denetimi detection).
|
||||
result = await self.norm_client.get_decision_document_as_markdown(document_url, page_number)
|
||||
return AnayasaUnifiedDocumentMarkdown(
|
||||
decision_type="norm_denetimi",
|
||||
source_url=result.source_url,
|
||||
document_data=result.model_dump(mode="json"),
|
||||
markdown_chunk=result.markdown_chunk,
|
||||
current_page=result.current_page,
|
||||
total_pages=result.total_pages,
|
||||
is_paginated=result.is_paginated,
|
||||
)
|
||||
|
||||
async def close_client_session(self):
|
||||
"""Close both client sessions."""
|
||||
if hasattr(self.norm_client, 'close_client_session'):
|
||||
await self.norm_client.close_client_session()
|
||||
if hasattr(self.bireysel_client, 'close_client_session'):
|
||||
await self.bireysel_client.close_client_session()
|
||||
await self.norm_client.close_client_session()
|
||||
await self.bireysel_client.close_client_session()
|
||||
|
||||
Reference in New Issue
Block a user