# SPDX-License-Identifier: Apache-2.0
"""Download a file with the browser, then read it on the gateway and save a copy here.

The browser saves downloads in the user's own folder on the gateway; wp.files lists them, extracts text or rows
(PDF, DOCX, XLSX, CSV…) and streams them (resumable).

    export WEBPILOT_URL=http://localhost:8931 WEBPILOT_TOKEN=cbu_...
    python examples/fixture_site.py &
    python examples/python/download.py                                   # the fixture's report.csv
    python examples/python/download.py https://example.org/files "Annual report"   # any page and link text
"""

from __future__ import annotations

import os
import sys
import time

from webpilot import WebPilot

FIXTURE = os.environ.get("WEBPILOT_FIXTURE", "http://host.docker.internal:8765")
page = sys.argv[1] if len(sys.argv) > 1 else f"{FIXTURE}/files"
link_text = sys.argv[2] if len(sys.argv) > 2 else "Download report"

with WebPilot() as wp:
    before = {f.name for f in wp.files.list()}
    tab = wp.open(page, mode="act")
    try:
        link = tab.snapshot().first(link_text, role="link")
        tab.click(link.ref)
        for _ in range(30):  # the download finishes in the background; wait until it is listed
            new = [f for f in wp.files.list() if f.name not in before]
            if new:
                break
            time.sleep(1)
        else:
            sys.exit("no new file after 30 s")
        f = new[0]
        print(f"downloaded {f.name} ({f.size} bytes, {f.mime})")
        if f.name.lower().endswith((".csv", ".xlsx", ".xls")):
            for row in list(wp.files.rows(f.name))[:5]:
                print("  ", row)
        else:
            print(wp.files.text(f.name).text[:500])
        print("saved to", wp.files.save(f.name, "."))
        wp.files.delete(f.name)  # clean up on the gateway
    finally:
        tab.close()