import os
import datetime
from jobspy.google import Google
from jobspy.linkedin import LinkedIn
from jobspy.indeed import Indeed
from jobspy.ziprecruiter import ZipRecruiter
from jobspy.model import ScraperInput

def clean_text(text: str) -> str:
    """
    Cleans text for CSV output by removing or replacing characters
    that could break CSV formatting.
    """
    if not text:
        return ""
    # Remove commas, newlines, carriage returns and double quotes.
    cleaned = text.replace(",", " ") \
                  .replace("\n", " ") \
                  .replace("\r", " ") \
                  .replace('"', "'")
    # Collapse multiple spaces into one.
    return " ".join(cleaned.split())

# Define job sources
sources = {
    "google": Google,
    "linkedin": LinkedIn,
    "indeed": Indeed,
    "zip_recruiter": ZipRecruiter,
}

# Define search preferences
search_terms = ["Automation Engineer", "CRM Manager", "Implementation Specialist", "Automation", "CRM"]
results_wanted = 100  # Fetch more jobs
max_days_old = 2      # Fetch jobs posted in last 48 hours
target_state = "NY"   # Only keep jobs from New York

def scrape_jobs(search_terms, results_wanted, max_days_old, target_state):
    """Scrape jobs from multiple sources and filter by state."""
    all_jobs = []
    today = datetime.date.today()
    
    print("\n🔎 DEBUG: Fetching jobs for search terms:", search_terms)

    for search_term in search_terms:
        for source_name, source_class in sources.items():
            print(f"\n🚀 Scraping {search_term} from {source_name}...")

            scraper = source_class()
            search_criteria = ScraperInput(
                site_type=[source_name],
                search_term=search_term,
                results_wanted=results_wanted,
            )

            job_response = scraper.scrape(search_criteria)

            for job in job_response.jobs:
                # Normalize location fields
                location_city = job.location.city.strip() if job.location.city else "Unknown"
                location_state = job.location.state.strip().upper() if job.location.state else "Unknown"
                location_country = str(job.location.country) if job.location.country else "Unknown"

                # Debug: Show all jobs being fetched
                print(f"📍 Fetched Job: {job.title} - {location_city}, {location_state}, {location_country}")

                # Exclude jobs that don’t explicitly match the search terms
                if not any(term.lower() in job.title.lower() for term in search_terms):
                    print(f"🚫 Excluding: {job.title} (Doesn't match {search_terms})")
                    continue

                # Ensure the job is recent and in NY (or remote)
                if job.date_posted and (today - job.date_posted).days <= max_days_old:
                    if location_state == target_state or job.is_remote:
                        print(f"✅ MATCH: {job.title} - {location_city}, {location_state} (Posted {job.date_posted})")
                        all_jobs.append({
                            "Job ID": job.id,
                            "Job Title (Primary)": clean_text(job.title),
                            "Company Name": clean_text(job.company_name) if job.company_name else "Unknown",
                            "Industry": clean_text(job.company_industry) if job.company_industry else "Not Provided",
                            "Experience Level": clean_text(job.job_level) if job.job_level else "Not Provided",
                            "Job Type": job.job_type[0].name if job.job_type else "Not Provided",
                            "Is Remote": job.is_remote,
                            "Currency": job.compensation.currency if job.compensation else "",
                            "Salary Min": job.compensation.min_amount if job.compensation else "",
                            "Salary Max": job.compensation.max_amount if job.compensation else "",
                            "Date Posted": job.date_posted.strftime("%Y-%m-%d") if job.date_posted else "Not Provided",
                            "Location City": location_city,
                            "Location State": location_state,
                            "Location Country": location_country,
                            "Job URL": job.job_url,
                            "Job Description": clean_text(job.description) if job.description else "No description available",
                            "Job Source": source_name
                        })
                    else:
                        print(f"❌ Ignored (Wrong State): {job.title} - {location_city}, {location_state} (Posted {job.date_posted})")
                else:
                    print(f"⏳ Ignored (Too Old): {job.title} - {location_city}, {location_state} (Posted {job.date_posted})")

    print(f"\n✅ {len(all_jobs)} jobs retrieved in NY")
    return all_jobs

def save_jobs_to_csv(jobs, filename="jobspy_output.csv"):
    """Save job data to a CSV file with a custom delimiter."""
    if not jobs:
        print("⚠️ No jobs found matching criteria.")
        return

    # Remove old CSV file before writing
    if os.path.exists(filename):
        os.remove(filename)

    fieldnames = [
        "Job ID", "Job Title (Primary)", "Company Name", "Industry",
        "Experience Level", "Job Type", "Is Remote", "Currency",
        "Salary Min", "Salary Max", "Date Posted", "Location City",
        "Location State", "Location Country", "Job URL", "Job Description",
        "Job Source"
    ]

    # Define your custom delimiter
    delimiter = "|~|"

    with open(filename, mode="w", encoding="utf-8") as file:
        # Write header
        file.write(delimiter.join(fieldnames) + "\n")
        # Write each job record
        for job in jobs:
            # Convert all field values to string
            row = [str(job.get(field, "")) for field in fieldnames]
            file.write(delimiter.join(row) + "\n")

    print(f"✅ Jobs saved to {filename} ({len(jobs)} entries)")

# Run the scraper with multiple job searches
job_data = scrape_jobs(
    search_terms=search_terms,
    results_wanted=results_wanted,
    max_days_old=max_days_old,
    target_state=target_state
)

# Save results to CSV
save_jobs_to_csv(job_data)