main.py

from __future__ import annotations

import argparse
from pathlib import Path

from http.cookiejar import MozillaCookieJar

from rich import print as rprint

from ref_collector.workflow import run_pipeline


def _load_cookies(path: str):
    """Load a Netscape/Mozilla cookies.txt file into a cookie jar for reuse in requests."""
    jar = MozillaCookieJar()
    jar.load(path, ignore_discard=True, ignore_expires=True)
    return jar


def parse_args() -> argparse.Namespace:
    """Parse CLI options for scraping, enrichment, and downloading steps."""
    parser = argparse.ArgumentParser(
        description="End-to-end helper to scrape Google Scholar, enrich with Crossref, download PDFs via Sci-Hub, and report missing papers."
    )
    source = parser.add_mutually_exclusive_group(required=False)
    source.add_argument("--profile-url", help="Google Scholar profile URL (e.g., https://scholar.google.com/citations?user=XXXX)")
    source.add_argument("--cited-url", help="Google Scholar 'cited by' search URL.")
    source.add_argument("--from-xlsx", help="Skip scraping and start from an existing metadata XLSX/CSV.")

    parser.add_argument("--workdir", default="_results/latest", help="Workspace folder to store outputs.")
    parser.add_argument("--pagesize", type=int, default=100, help="Items per page when scraping profiles (Scholar caps at 100).")
    parser.add_argument("--max-pages", type=int, help="Max pages to scrape (default: all).")
    parser.add_argument("--scrape-delay", type=float, default=2.5, help="Delay between Scholar page fetches (seconds).")
    parser.add_argument("--crossref-delay", type=float, default=1.0, help="Delay between Crossref API calls (seconds).")
    parser.add_argument("--download-delay", type=float, default=1.5, help="Delay between Sci-Hub download attempts (seconds).")
    parser.add_argument("--limit", type=int, help="Optional limit on number of papers to download (after scraping).")
    parser.add_argument("--mailto", help="Email for Crossref polite pool User-Agent (recommended).")
    parser.add_argument(
        "--name-template",
        default="{year}_{venue_abbrev}_{short_title}",
        help="Filename template for PDFs. Fields: year, venue, venue_abbrev, short_title, title, doi.",
    )
    parser.add_argument("--pdf-dir", help="Custom PDF output folder (default: <workdir>/pdfs).")
    parser.add_argument("--report-name", default="missing_or_failed.md", help="Filename for the missing/failed report in workdir.")
    parser.add_argument("--no-download", action="store_true", help="Only build metadata spreadsheet, skip PDF downloads.")
    parser.add_argument("--proxy", help="HTTP(S) proxy, e.g. http://user:pass@host:port . Applied to Scholar, Crossref, Sci-Hub.")
    parser.add_argument("--cookies-file", help="Netscape/Mozilla cookies.txt file to reuse solved Scholar captcha session.")
    return parser.parse_args()


def main() -> None:
    """Entry point: parse arguments, run the pipeline, and print a small summary."""
    args = parse_args()
    if not any([args.profile_url, args.cited_url, args.from_xlsx]):
        raise SystemExit("Provide --profile-url, --cited-url, or --from-xlsx.")

    # Auto-correct common mistake: cited-by URL supplied via --profile-url.
    if args.profile_url and ("cites=" in args.profile_url or "oi=bibs" in args.profile_url):
        if args.cited_url:
            raise SystemExit("Provide only one source. Detected cited-by URL in --profile-url and also --cited-url.")
        args.cited_url = args.profile_url
        args.profile_url = None

    result = run_pipeline(
        profile_url=args.profile_url,
        cited_url=args.cited_url,
        from_xlsx=Path(args.from_xlsx) if args.from_xlsx else None,
        workdir=Path(args.workdir),
        pagesize=args.pagesize,
        max_pages=args.max_pages,
        scrape_delay=args.scrape_delay,
        crossref_delay=args.crossref_delay,
        mailto=args.mailto,
        name_template=args.name_template,
        pdf_dir=Path(args.pdf_dir) if args.pdf_dir else None,
        download_delay=args.download_delay,
        limit=args.limit,
        report_name=args.report_name,
        skip_download=args.no_download,
        cookies=_load_cookies(args.cookies_file) if args.cookies_file else None,
        proxies={"http": args.proxy, "https": args.proxy} if args.proxy else None,
    )

    rprint("\n:checkered_flag: [bold green]Pipeline complete[/bold green]")
    rprint(f":page_facing_up: Metadata saved to : [cyan]{result['metadata_path']}[/cyan]")
    rprint(f":open_file_folder: PDFs saved to     : [cyan]{result['pdf_dir']}[/cyan]")
    rprint(f":memo: Missing/failed MD : [cyan]{result['report_path']}[/cyan]")
    rprint(f":bar_chart: Total papers      : [yellow]{result['total']}[/yellow]")
    rprint(f":dart: Attempted download: [yellow]{result['attempted']}[/yellow]")
    rprint(f":white_check_mark: Downloaded        : [green]{result['downloaded']}[/green]")
    rprint(f":x: Failed/missing    : [red]{result['failed']}[/red]")


if __name__ == "__main__":
    main()