# 项目文件：demos/browser_collect.py
# 先登记响应等待再点击；接口完成后继续等页面卡片出现，最后读取当前 DOM。
import argparse
from urllib.parse import urlsplit

from playwright.sync_api import expect, sync_playwright
from crawlsite import running_site


def collect(base_url, channel=None):
    with sync_playwright() as playwright:
        browser = playwright.chromium.launch(headless=True, channel=channel)
        try:
            page = browser.new_page()
            response = page.goto(base_url + "/dynamic/", wait_until="domcontentloaded")
            if response is None or not response.ok:
                raise RuntimeError("目录页面没有正常打开")
            cards = page.locator("#results article")
            for number, label in [(1, "加载目录"), (2, "下一页")]:
                # 监听先于 click 建立，避免快速响应已经到达后才开始等待。
                with page.expect_response(
                    lambda r: (
                        urlsplit(r.url).path == "/api/catalog"
                        and urlsplit(r.url).query == f"page={number}"
                    )
                ) as pending:
                    page.get_by_role("button", name=label, exact=True).click()
                if not pending.value.ok:
                    raise RuntimeError("目录接口读取失败")
                expect(cards).to_have_count(number * 3)
            ids = [card.get_attribute("data-id") for card in cards.all()]
            return len(ids), sorted(set(ids))
        finally:
            browser.close()


if __name__ == "__main__":
    parser = argparse.ArgumentParser()
    parser.add_argument(
        "--channel", help="例如使用已安装的 msedge；省略则用 Playwright Chromium"
    )
    args = parser.parse_args()
    with running_site() as site:
        print(collect(site.base_url, args.channel))