feat: Add NIP lookup services for membership application

- Add ceidg_api_service.py with fetch_ceidg_by_nip() function
- Add KRSApiService class with search_by_nip() method
- KRS lookup uses rejestr.io API (unofficial) or database fallback
- CEIDG lookup uses official dane.biznes.gov.pl API

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
This commit is contained in:
Maciej Pienczyn 2026-02-01 14:22:01 +01:00
parent 3e23bd3e4e
commit 28affce99f
2 changed files with 271 additions and 0 deletions

156
ceidg_api_service.py Normal file
View File

@ -0,0 +1,156 @@
#!/usr/bin/env python3
"""
CEIDG API Service
==================
Service module for fetching company data from CEIDG (Centralna Ewidencja
i Informacja o Działalności Gospodarczej) using the official API at
dane.biznes.gov.pl.
Provides fetch_ceidg_by_nip function for membership application workflow.
"""
import os
import logging
import requests
from typing import Optional, Dict, Any
from dotenv import load_dotenv
# Load environment variables
load_dotenv()
logger = logging.getLogger(__name__)
# API Configuration
CEIDG_API_V3_URL = "https://dane.biznes.gov.pl/api/ceidg/v3/firmy"
CEIDG_API_KEY = os.getenv("CEIDG_API_KEY")
CEIDG_TIMEOUT = 15 # seconds
def fetch_ceidg_by_nip(nip: str) -> Optional[Dict[str, Any]]:
"""
Fetch company data from CEIDG API by NIP.
Args:
nip: NIP number (10 digits, no dashes)
Returns:
Dictionary with company data or None if not found
"""
if not CEIDG_API_KEY:
logger.warning("CEIDG_API_KEY not configured - CEIDG lookup disabled")
return None
# Clean NIP
nip = nip.strip().replace('-', '').replace(' ', '')
if not nip or len(nip) != 10 or not nip.isdigit():
logger.warning(f"Invalid NIP format: {nip}")
return None
headers = {
"Authorization": f"Bearer {CEIDG_API_KEY}",
"Accept": "application/json"
}
try:
logger.info(f"Fetching CEIDG data for NIP {nip}")
response = requests.get(
CEIDG_API_V3_URL,
params={"nip": nip},
headers=headers,
timeout=CEIDG_TIMEOUT
)
if response.status_code == 401:
logger.error("CEIDG API authentication failed - check CEIDG_API_KEY")
return None
if response.status_code == 404:
logger.info(f"NIP {nip} not found in CEIDG")
return None
if response.status_code != 200:
logger.error(f"CEIDG API error: {response.status_code} - {response.text[:200]}")
return None
data = response.json()
# Handle response format - can be list or dict
if isinstance(data, list):
if not data:
logger.info(f"NIP {nip} not found in CEIDG (empty list)")
return None
firma = data[0]
elif isinstance(data, dict):
if 'firmy' in data:
firmy = data.get('firmy', [])
if not firmy:
logger.info(f"NIP {nip} not found in CEIDG")
return None
firma = firmy[0]
else:
firma = data
else:
logger.error(f"Unexpected CEIDG response format: {type(data)}")
return None
# Extract address
adres = firma.get('adresDzialalnosci', {}) or firma.get('adres', {}) or {}
if isinstance(adres, str):
adres = {'full': adres}
# Build normalized result
result = {
'firma': firma.get('nazwa') or firma.get('nazwaSkrocona'),
'nip': firma.get('nip'),
'regon': firma.get('regon'),
'adresDzialalnosci': {
'kodPocztowy': adres.get('kodPocztowy') or adres.get('kod'),
'miejscowosc': adres.get('miejscowosc') or adres.get('miasto'),
'ulica': adres.get('ulica'),
'budynek': adres.get('budynek') or adres.get('nrDomu') or adres.get('nrBudynku'),
'lokal': adres.get('lokal') or adres.get('nrLokalu'),
},
'email': firma.get('email') or firma.get('adresEmail'),
'stronaWWW': firma.get('stronaWWW') or firma.get('www') or firma.get('strona'),
'telefon': firma.get('telefon'),
'dataRozpoczeciaDzialalnosci': firma.get('dataRozpoczeciaDzialalnosci') or firma.get('dataWpisuDoCeidg'),
'status': firma.get('status'),
'raw': firma
}
logger.info(f"CEIDG data found for NIP {nip}: {result['firma']}")
return result
except requests.exceptions.Timeout:
logger.error(f"CEIDG API timeout for NIP {nip}")
return None
except requests.exceptions.RequestException as e:
logger.error(f"CEIDG API request error for NIP {nip}: {e}")
return None
except Exception as e:
logger.error(f"Error fetching CEIDG data for NIP {nip}: {e}")
return None
# For testing
if __name__ == '__main__':
import sys
import json
if len(sys.argv) < 2:
print("Usage: python ceidg_api_service.py <NIP>")
print("Example: python ceidg_api_service.py 5881571773")
sys.exit(1)
nip = sys.argv[1]
print(f"Pobieranie danych z CEIDG API dla NIP: {nip}")
print("=" * 60)
data = fetch_ceidg_by_nip(nip)
if data:
print(json.dumps(data, indent=2, ensure_ascii=False, default=str))
else:
print(f"Nie znaleziono firmy o NIP {nip} w CEIDG")

View File

@ -440,6 +440,121 @@ def format_address(krs_data: KRSCompanyData) -> str:
return ', '.join(parts)
# ============================================================
# KRS API Service Class (for search_by_nip compatibility)
# ============================================================
class KRSApiService:
"""
KRS API Service class providing unified interface for KRS lookups.
Note: KRS Open API doesn't support direct NIP lookup.
This class uses rejestr.io unofficial API as a fallback.
"""
REJESTR_IO_URL = "https://rejestr.io/api/v2/org"
REJESTR_IO_TIMEOUT = 10
def search_by_nip(self, nip: str) -> Optional[Dict[str, Any]]:
"""
Search KRS by NIP number.
Since KRS Open API doesn't support NIP lookup, this method:
1. First tries rejestr.io API (unofficial but reliable)
2. Falls back to checking our database for KRS number
3. Then fetches full data from KRS Open API
Args:
nip: NIP number (10 digits)
Returns:
Dictionary with company data or None if not found
"""
import logging
logger = logging.getLogger(__name__)
# Clean NIP
nip = nip.strip().replace('-', '').replace(' ', '')
if not nip or len(nip) != 10 or not nip.isdigit():
return None
# Try rejestr.io first (supports NIP lookup)
krs_number = self._get_krs_from_rejestr_io(nip)
if not krs_number:
# Try our database
krs_number = self._get_krs_from_database(nip)
if not krs_number:
logger.info(f"No KRS found for NIP {nip}")
return None
# Fetch full data from KRS Open API
logger.info(f"Found KRS {krs_number} for NIP {nip}, fetching details")
krs_data = get_company_from_krs(krs_number)
if not krs_data:
return None
# Return as dict with expected format
return krs_data.to_dict()
def _get_krs_from_rejestr_io(self, nip: str) -> Optional[str]:
"""Try to get KRS number from rejestr.io API."""
import logging
logger = logging.getLogger(__name__)
try:
# rejestr.io API endpoint for NIP lookup
url = f"{self.REJESTR_IO_URL}"
params = {"nip": nip}
response = requests.get(url, params=params, timeout=self.REJESTR_IO_TIMEOUT)
if response.status_code == 200:
data = response.json()
# Handle different response formats
if isinstance(data, list) and data:
krs = data[0].get('krs')
if krs:
return str(krs).zfill(10)
elif isinstance(data, dict):
krs = data.get('krs')
if krs:
return str(krs).zfill(10)
logger.debug(f"rejestr.io: No KRS found for NIP {nip}")
return None
except Exception as e:
logger.debug(f"rejestr.io lookup failed: {e}")
return None
def _get_krs_from_database(self, nip: str) -> Optional[str]:
"""Check our database for KRS number."""
import logging
logger = logging.getLogger(__name__)
try:
from database import SessionLocal, Company
db = SessionLocal()
try:
company = db.query(Company).filter(Company.nip == nip).first()
if company and company.krs:
logger.debug(f"Found KRS {company.krs} in database for NIP {nip}")
return company.krs
finally:
db.close()
except Exception as e:
logger.debug(f"Database KRS lookup failed: {e}")
return None
# Singleton instance for import
krs_api_service = KRSApiService()
# CLI for testing
if __name__ == '__main__':
import sys