cryptorugmunch
2f1eec2f78
Some checks failed
CI / lint (pull_request) Successful in 33s
CI / typecheck (pull_request) Failing after 1m42s
CI / test (pull_request) Failing after 2m33s
CI / security (pull_request) Failing after 39s
CI / gitleaks (pull_request) Successful in 36s
CI / commitlint (pull_request) Failing after 10s
66 lines
2 KiB
Python
66 lines
2 KiB
Python
"""Pry — Enrichment router (remaining api.py routes).
|
|
|
|
Auto-extracted from api.py during the router-split refactor.
|
|
"""
|
|
|
|
# SPDX-License-Identifier: MIT
|
|
# Copyright (c) 2026 Rug Munch Media LLC
|
|
|
|
from __future__ import annotations
|
|
|
|
import logging
|
|
from typing import Any
|
|
|
|
from fastapi import APIRouter, Body
|
|
|
|
logger = logging.getLogger(__name__)
|
|
|
|
router = APIRouter(tags=["Enrichment"])
|
|
|
|
|
|
@router.post(
|
|
"/v1/enrich",
|
|
tags=["Enrichment"],
|
|
summary="Enrich scraped data with company info, tech stack, social profiles",
|
|
)
|
|
async def enrich_data(
|
|
url: str = Body(...),
|
|
include_tech_stack: bool = Body(True),
|
|
include_social: bool = Body(True),
|
|
include_company: bool = Body(True),
|
|
) -> dict[str, Any]:
|
|
"""Enrich a URL with supplemental business intelligence.
|
|
|
|
Returns:
|
|
- Tech stack detection (CMS, framework, CDN, analytics, payments)
|
|
- Social media profiles (Twitter, LinkedIn, Facebook, Instagram, GitHub, etc.)
|
|
- Company information (email, phone, address, founded year, team size)
|
|
"""
|
|
from enrichment import enrich_url
|
|
|
|
result = await enrich_url(url)
|
|
filtered: dict[str, Any] = {"url": url}
|
|
if include_tech_stack:
|
|
filtered["tech_stack"] = result.get("tech_stack")
|
|
if include_social:
|
|
filtered["social_profiles"] = result.get("social_profiles")
|
|
if include_company:
|
|
filtered["company_info"] = result.get("company_info")
|
|
|
|
return {"success": True, "data": filtered}
|
|
|
|
|
|
@router.post(
|
|
"/v1/enrich/tech-stack", tags=["Enrichment"], summary="Detect technologies used on a website"
|
|
)
|
|
async def detect_tech(url: str = Body(...)) -> dict[str, Any]:
|
|
"""Detect what technologies a website is built with.
|
|
|
|
Detects: CMS (WordPress, Shopify, Wix), frameworks (Next.js, Django, Rails),
|
|
frontend (React, Vue, Angular), CDN (Cloudflare, Fastly), analytics (GA, Hotjar),
|
|
payments (Stripe, PayPal).
|
|
"""
|
|
from enrichment import enrich_url
|
|
|
|
result = await enrich_url(url)
|
|
return {"success": True, "data": result.get("tech_stack", {})}
|