| from playwright.sync_api import sync_playwright |
|
|
| URL = "https://huggingface.co/hkblockchaintechnology/tiny-computer-agent" |
| with sync_playwright() as p: |
| browser = p.chromium.launch() |
| page = browser.new_page(viewport={"width": 1440, "height": 1100}) |
| page.goto(URL, wait_until="load", timeout=120000) |
| page.wait_for_timeout(6000) |
| html = page.content() |
| print("len(html):", len(html)) |
| for needle in ["gallery/", "<img", "article img", "Capability gallery", |
| "desk-click", "resolve/main/gallery"]: |
| print(f" {needle!r:24} occurrences in DOM: {html.count(needle)}") |
| print("total <img> on page:", page.eval_on_selector_all("img", "els => els.length")) |
| print("containers:", page.eval_on_selector_all( |
| "main, article, [class*=prose]", "els => els.slice(0,4).map(e => e.tagName + '.' + e.className.slice(0,40))")) |
| |
| print("has heading text:", page.locator("text=Capability gallery").count()) |
| browser.close() |
|
|
