import hermes
from hermes.tools import web_search, web_extract, sheets

class LeadResearchAgent:
    def __init__(self, icp):
        self.icp = icp  # Ideal Customer Profile
        self.leads = []

    def research(self, query, max_results=50):
        """Search for prospects matching ICP"""
        results = web_search(
            query=self._build_query(query),
            limit=max_results
        )
        return self._parse_results(results)

    def _build_query(self, base):
        return (
            f"{base} "
            f"{self.icp['industry']} "
            f"{self.icp['title']} "
            f"{self.icp['location']}"
        )

    def _parse_results(self, results):
        for r in results:
            lead = {
                'name': r.get('title'),
                'company': r.get('site'),
                'url': r.get('url'),
                'source': 'web_search'
            }
            self.leads.append(lead)
        return self.leads

# Usage
agent = LeadResearchAgent(icp={
    'industry': 'SaaS',
    'title': 'CTO OR VP Engineering',
    'location': 'United States'
})
leads = agent.research('engineering leadership')
sheets.append('Leads', leads)

import requests
from bs4 import BeautifulSoup
import re

def scrape_prospects(url):
    """Scrape company directory for leads"""
    headers = {
        'User-Agent': 'LeadResearchBot/1.0'
    }
    response = requests.get(url, headers=headers)
    soup = BeautifulSoup(response.text, 'html.parser')

    leads = []
    # Find all company cards
    for card in soup.select('.company-card'):
        lead = {
            'company': card.select_one('.company-name').text.strip(),
            'website': card.select_one('a')['href'],
            'industry': card.select_one('.industry').text.strip(),
            'size': card.select_one('.size').text.strip(),
            'location': card.select_one('.location').text.strip(),
        }
        leads.append(lead)

    return leads

# Scrape multiple pages
for page in range(1, 11):
    url = f"https://directory.example.com?page={page}"
    leads = scrape_prospects(url)
    # Save to database...

import re
from difflib import SequenceMatcher

def clean_leads(leads):
    """Clean and deduplicate lead list"""
    cleaned = []
    seen = set()

    for lead in leads:
        # Normalize email
        email = lead['email'].lower().strip()
        if not re.match(r'^[\w\.-]+@[\w\.-]+\.\w+$', email):
            continue  # Skip invalid emails

        # Create dedup key
        key = email.split('@')[0] + '@' + email.split('@')[1].split('.')[0]

        if key in seen:
            continue
        seen.add(key)

        # Clean fields
        lead['name'] = lead['name'].title().strip()
        lead['company'] = lead['company'].strip()
        lead['email'] = email
        cleaned.append(lead)

    return cleaned

def fuzzy_dedup(leads, threshold=0.85):
    """Remove near-duplicate company names"""
    unique = []
    for lead in leads:
        is_dup = False
        for existing in unique:
            ratio = SequenceMatcher(
                None,
                lead['company'].lower(),
                existing['company'].lower()
            ).ratio()
            if ratio > threshold:
                is_dup = True
                break
        if not is_dup:
            unique.append(lead)
    return unique

# automation.py — Run daily at 9 AM
import schedule
import time
from lead_agent import LeadResearchAgent
from scraper import scrape_prospects
from cleaner import clean_leads, fuzzy_dedup

def daily_research():
    """Full automated research pipeline"""
    agent = LeadResearchAgent(icp=ICP)

    # 1. Discover new prospects
    new_leads = agent.research('target keywords')

    # 2. Scrape additional data
    scraped = []
    for lead in new_leads:
        data = scrape_prospects(lead['url'])
        scraped.extend(data)

    # 3. Clean & deduplicate
    cleaned = clean_leads(scraped)
    unique = fuzzy_dedup(cleaned)

    # 4. Store in database
    sheets.append('Leads', unique)
    print(f"Added {len(unique)} new leads")

# Schedule daily at 9:00 AM
schedule.every().day.at("09:00").do(daily_research)

while True:
    schedule.run_pending()
    time.sleep(60)