tools/linkedin_job_search.py

from smolagents.tools import Tool
import requests
from typing import List, Dict
import os
from dotenv import load_dotenv
from bs4 import BeautifulSoup

load_dotenv()

def scrape_indeed(position: str, location: str) -> List[Dict]:
    """
    Scrapes job postings from Indeed.
    """
    url = f"https://www.indeed.com/jobs?q=Odoo+{position}&l={location}"
    headers = {'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/91.0.4472.124 Safari/537.36'}
    try:
        response = requests.get(url, headers=headers)
        response.raise_for_status()
        soup = BeautifulSoup(response.content, 'html.parser')
        jobs = []
        for div in soup.find_all('div', class_='jobsearch-SerpJobCard'):
            title_element = div.find('a', title=True)
            title = title_element.text if title_element else 'N/A'
            company_element = div.find('span', class_='company')
            company = company_element.text.strip() if company_element else 'N/A'
            location_element = div.find('div', class_='location')
            location = location_element.text if location_element else 'N/A'
            link = 'https://www.indeed.com' + div.find('a', href=True)['href'] if div.find('a', href=True) else 'N/A'
            jobs.append({
                "Title": title,
                "Company": company,
                "Location": location,
                "Link": link,
                "Source": "Indeed"
            })
        return jobs
    except requests.exceptions.RequestException as e:
        print(f"Indeed scraping failed: {e}")
        return []

class LinkedInJobSearchTool(Tool):
    name = "linkedin_job_search"
    description = "Searches for job postings on LinkedIn and Indeed based on job title, location, and work mode (remote, hybrid, in-office) for Odoo profiles."
    
    inputs = {
        "position": {"type": "string", "description": "Job title (e.g., Data Scientist)"},
        "location": {"type": "string", "description": "City or country (e.g., Germany)"},
        "work_mode": {"type": "string", "description": "remote, hybrid, in-office"}
    }
    
    output_type = "array"

    def forward(self, position: str, location: str, work_mode: str) -> List[Dict]:
        """
        Fetches job listings from LinkedIn and Indeed and returns structured JSON.
        """
        BRAVE_API_KEY = os.getenv("BRAVE_API_KEY")
        if not BRAVE_API_KEY:
            return [{"Error": "Brave API key not found in .env file."}]
        
        linkedin_results = []
        indeed_results = []

        # LinkedIn Job Search
        base_url = "https://api.brave.com/v1/jobs"
        params = {
            "q": f"Odoo {position} {work_mode} jobs",
            "location": location,
            "api_key": BRAVE_API_KEY
        }

        try:
            response = requests.get(base_url, params=params)
            response.raise_for_status()
            data = response.json()
            linkedin_jobs = data.get("jobs", [])
            
            if linkedin_jobs:
                for job in linkedin_jobs:
                    linkedin_results.append({
                        "Title": job['title'],
                        "Company": job.get('company', 'N/A'),
                        "Location": job.get('location', 'N/A'),
                        "Posted": job.get('posted_date', 'N/A'),
                        "Link": job.get('url', 'N/A'),
                        "Source": "LinkedIn"
                    })
        except requests.exceptions.RequestException as e:
            linkedin_results = [{"Error": f"LinkedIn Error: {str(e)}"}]

        # Indeed Job Search
        indeed_results = scrape_indeed(position, location)

        # Combine results, prioritizing LinkedIn
        combined_results = linkedin_results + indeed_results

        # Format the results
        formatted_results = ""
        if combined_results:
            for job in combined_results:
                formatted_results += f"Title: {job.get('Title', 'N/A')}\n"
                formatted_results += f"Company: {job.get('Company', 'N/A')}\n"
                formatted_results += f"Location: {job.get('Location', 'N/A')}\n"
                formatted_results += f"Posted: {job.get('Posted', 'N/A')}\n"
                formatted_results += f"Link: {job.get('Link', 'N/A')}\n"
                formatted_results += f"Source: {job.get('Source', 'N/A')}\n"
                formatted_results += "---\n"
        else:
            formatted_results = "No jobs found. Try different keywords."

        return [{"Results": formatted_results}]