job-scraper/notify.py

from dataclasses import dataclass
from typing import Optional
import json

from db import StoredJob
from scrapers.base import Job


@dataclass
class ChangeReport:
    """Report of changes detected during a scrape."""
    company_name: str
    new_jobs: list[Job]
    removed_jobs: list[StoredJob]
    total_active: int


class Notifier:
    """Handles notifications for job changes."""

    def __init__(self, config: dict):
        self.config = config

    def notify(self, reports: list[ChangeReport]):
        """Send notifications for all changes."""
        # Filter to only reports with changes
        reports_with_changes = [r for r in reports if r.new_jobs or r.removed_jobs]

        if not reports_with_changes:
            print("\n✓ No changes detected across all companies.")
        else:
            # Console output for changes
            self._notify_console(reports_with_changes)

        # Email (if configured) - only sends when there are changes
        email_config = self.config.get("email")
        if email_config and reports_with_changes:
            self._notify_email(reports_with_changes, email_config)

        # msmtp (if configured - sends daily summary always)
        msmtp_config = self.config.get("msmtp")
        if msmtp_config:
            self._notify_msmtp_daily_summary(reports, msmtp_config)

        # Slack (if configured) - only sends when there are changes
        slack_config = self.config.get("slack")
        if slack_config and reports_with_changes:
            self._notify_slack(reports_with_changes, slack_config)

    def _notify_console(self, reports: list[ChangeReport]):
        """Print changes to console."""
        print("\n" + "=" * 60)
        print("JOB CHANGES DETECTED")
        print("=" * 60)

        total_new = sum(len(r.new_jobs) for r in reports)
        total_removed = sum(len(r.removed_jobs) for r in reports)

        print(f"\nSummary: {total_new} new jobs, {total_removed} removed jobs\n")

        for report in reports:
            print(f"\n📌 {report.company_name} ({report.total_active} active jobs)")
            print("-" * 40)

            if report.new_jobs:
                print(f"\n  🆕 NEW JOBS ({len(report.new_jobs)}):")
                for job in report.new_jobs:
                    location_str = f" [{job.location}]" if job.location else ""
                    remote_str = f" 🏠" if job.remote_type == "remote" else ""
                    print(f"     • {job.title}{location_str}{remote_str}")
                    print(f"       {job.url}")

            if report.removed_jobs:
                print(f"\n  ❌ REMOVED JOBS ({len(report.removed_jobs)}):")
                for job in report.removed_jobs:
                    print(f"     • {job.title}")

        print("\n" + "=" * 60)

    def _notify_email(self, reports: list[ChangeReport], config: dict):
        """Send email notification."""
        import smtplib
        from email.mime.text import MIMEText
        from email.mime.multipart import MIMEMultipart

        # Build email body
        body = self._build_html_report(reports)

        msg = MIMEMultipart("alternative")
        msg["Subject"] = f"Job Alert: {sum(len(r.new_jobs) for r in reports)} new positions"
        msg["From"] = config["from_addr"]
        msg["To"] = config["to_addr"]

        msg.attach(MIMEText(body, "html"))

        try:
            with smtplib.SMTP(config["smtp_host"], config["smtp_port"]) as server:
                server.starttls()
                server.login(config["username"], config["password"])
                server.send_message(msg)
            print("✓ Email notification sent")
        except Exception as e:
            print(f"✗ Failed to send email: {e}")

    def _notify_msmtp(self, reports: list[ChangeReport], config: dict):
        """Send email notification via system msmtp."""
        import subprocess
        from datetime import datetime

        to_addr = config.get("to_addr", "me@bastiangruber.ca")
        from_addr = config.get("from_addr", "admin@novanexus.ca")

        total_new = sum(len(r.new_jobs) for r in reports)
        total_removed = sum(len(r.removed_jobs) for r in reports)

        # Build subject line
        parts = []
        if total_new:
            parts.append(f"+{total_new} new")
        if total_removed:
            parts.append(f"-{total_removed} removed")
        subject = f"Job Board Update: {', '.join(parts)}"

        # Build plain text body
        body_lines = [
            "JOB BOARD CHANGES",
            f"{datetime.now().strftime('%Y-%m-%d %H:%M')}",
            "",
            f"Summary: {total_new} new jobs, {total_removed} removed jobs",
            "",
        ]

        for report in reports:
            body_lines.append(f"{report.company_name} ({report.total_active} active)")
            body_lines.append("-" * 40)

            if report.new_jobs:
                body_lines.append(f"  NEW ({len(report.new_jobs)}):")
                for job in report.new_jobs:
                    location_str = f" [{job.location}]" if job.location else ""
                    remote_str = " (Remote)" if job.remote_type == "remote" else ""
                    body_lines.append(f"    + {job.title}{location_str}{remote_str}")
                    body_lines.append(f"      {job.url}")

            if report.removed_jobs:
                body_lines.append(f"  REMOVED ({len(report.removed_jobs)}):")
                for job in report.removed_jobs:
                    body_lines.append(f"    - {job.title}")

            body_lines.append("")

        body_lines.append("---")
        body_lines.append("Generated by job-scraper")

        body = "\n".join(body_lines)

        # Build email message
        email_msg = f"""Subject: {subject}
From: {from_addr}
To: {to_addr}
Content-Type: text/plain; charset=UTF-8

{body}
"""

        try:
            result = subprocess.run(
                ["msmtp", to_addr],
                input=email_msg,
                capture_output=True,
                text=True,
            )
            if result.returncode == 0:
                print("✓ msmtp notification sent")
            else:
                print(f"✗ msmtp failed: {result.stderr}")
        except FileNotFoundError:
            print("✗ msmtp not found - install with: apt install msmtp")
        except Exception as e:
            print(f"✗ Failed to send msmtp notification: {e}")

    def _notify_msmtp_daily_summary(self, reports: list[ChangeReport], config: dict):
        """Send daily summary email via system msmtp (always sends)."""
        import subprocess
        from datetime import datetime

        to_addr = config.get("to_addr", "me@bastiangruber.ca")
        from_addr = config.get("from_addr", "admin@novanexus.ca")

        # Calculate totals
        total_companies = len([r for r in reports if r.total_active > 0])
        total_jobs = sum(r.total_active for r in reports)
        total_new = sum(len(r.new_jobs) for r in reports)
        total_removed = sum(len(r.removed_jobs) for r in reports)

        # Build subject line
        if total_new or total_removed:
            changes = []
            if total_new:
                changes.append(f"+{total_new}")
            if total_removed:
                changes.append(f"-{total_removed}")
            subject = f"Job Board: {', '.join(changes)} | {total_jobs} jobs"
        else:
            subject = f"Job Board: No changes | {total_jobs} jobs"

        # Build plain text body
        body_lines = [
            "JOB BOARD DAILY SUMMARY",
            f"{datetime.now().strftime('%Y-%m-%d %H:%M')}",
            "",
            "OVERVIEW",
            f"  Companies with jobs: {total_companies}",
            f"  Total jobs tracked:  {total_jobs}",
            "",
        ]

        # Changes section
        reports_with_changes = [r for r in reports if r.new_jobs or r.removed_jobs]

        if reports_with_changes:
            body_lines.append(f"CHANGES: +{total_new} new, -{total_removed} removed")
            body_lines.append("-" * 40)

            for report in reports_with_changes:
                if report.new_jobs:
                    for job in report.new_jobs:
                        location_str = f" [{job.location}]" if job.location else ""
                        remote_str = " (Remote)" if job.remote_type == "remote" else ""
                        body_lines.append(f"  + {report.company_name}: {job.title}{location_str}{remote_str}")

                if report.removed_jobs:
                    for job in report.removed_jobs:
                        body_lines.append(f"  - {report.company_name}: {job.title}")

            body_lines.append("")
        else:
            body_lines.append("CHANGES: No changes detected")
            body_lines.append("")

        body_lines.append("---")
        body_lines.append("https://jobs.novanexus.ca")

        body = "\n".join(body_lines)

        # Build email message
        email_msg = f"""Subject: {subject}
From: {from_addr}
To: {to_addr}
Content-Type: text/plain; charset=UTF-8

{body}
"""

        try:
            result = subprocess.run(
                ["msmtp", to_addr],
                input=email_msg,
                capture_output=True,
                text=True,
            )
            if result.returncode == 0:
                print("✓ Daily summary email sent")
            else:
                print(f"✗ msmtp failed: {result.stderr}")
        except FileNotFoundError:
            print("✗ msmtp not found - install with: apt install msmtp")
        except Exception as e:
            print(f"✗ Failed to send daily summary: {e}")

    def _notify_slack(self, reports: list[ChangeReport], config: dict):
        """Send Slack notification."""
        import httpx

        blocks = []

        # Header
        total_new = sum(len(r.new_jobs) for r in reports)
        blocks.append({
            "type": "header",
            "text": {"type": "plain_text", "text": f"🔔 {total_new} New Job Openings"}
        })

        for report in reports:
            if report.new_jobs:
                blocks.append({"type": "divider"})
                blocks.append({
                    "type": "section",
                    "text": {
                        "type": "mrkdwn",
                        "text": f"*{report.company_name}* ({len(report.new_jobs)} new)"
                    }
                })

                for job in report.new_jobs[:5]:  # Limit to 5 per company
                    location = f" • {job.location}" if job.location else ""
                    blocks.append({
                        "type": "section",
                        "text": {
                            "type": "mrkdwn",
                            "text": f"<{job.url}|{job.title}>{location}"
                        }
                    })

        payload = {"blocks": blocks}

        try:
            response = httpx.post(config["webhook_url"], json=payload)
            response.raise_for_status()
            print("✓ Slack notification sent")
        except Exception as e:
            print(f"✗ Failed to send Slack notification: {e}")

    def _build_html_report(self, reports: list[ChangeReport]) -> str:
        """Build HTML email body."""
        total_new = sum(len(r.new_jobs) for r in reports)

        html = f"""
        <html>
        <body style="font-family: Arial, sans-serif; max-width: 600px; margin: 0 auto;">
            <h1 style="color: #333;">🔔 {total_new} New Job Openings</h1>
        """

        for report in reports:
            if report.new_jobs:
                html += f"""
                <h2 style="color: #666; border-bottom: 1px solid #ddd; padding-bottom: 5px;">
                    {report.company_name}
                </h2>
                <ul>
                """
                for job in report.new_jobs:
                    location = f" <span style='color: #888;'>({job.location})</span>" if job.location else ""
                    html += f"""
                    <li style="margin: 10px 0;">
                        <a href="{job.url}" style="color: #0066cc; text-decoration: none;">
                            {job.title}
                        </a>
                        {location}
                    </li>
                    """
                html += "</ul>"

        html += """
        </body>
        </html>
        """
        return html