Get a page, or a browser?

/** * Get a page, or a browser? The same site two ways: fetch (one call, the page as Markdown: for reading) and a session * (a browser that stays open between calls: for clicking through pages, forms and sign-ins). * *   npx tsx node/index.ts            (BOXLINE_API_KEY in the environment; PAGE_URL for another site) */import { mkdirSync, writeFileSync } from "node:fs";import { join } from "node:path";import { Boxline } from "@boxline/sdk";const out = process.env.OUTPUT_DIR ?? "output";const url = process.env.PAGE_URL ?? "https://quotes.toscrape.com/";const bx = new Boxline();// 1. fetch: the page is rendered in a real browser inside a sandbox and comes back as Markdown, with its links.let t0 = Date.now();const page = await bx.fetch(url, { format: "markdown", links: true });const fetched = { title: page.title, status: page.status, chars: page.content.length, links: page.links?.length ?? 0, ms: Date.now() - t0 };console.log(`fetch: "${page.title}" (HTTP ${page.status}), ${fetched.chars} characters of Markdown and ${fetched.links} links in ${fetched.ms} ms`);// 2. A session: the browser keeps its page, cookies and history between calls, so it can click on to page 3.const session = await bx.sessions.create({ timeout: 300, userMetadata: { example: "fetch-or-browser" } });console.log(`Session: ${session.id}`);try {  t0 = Date.now();  await session.goto(url);  const pages: { url: string; authors: string[] }[] = [];  for (let i = 0; i < 2; i++) {    const results = await session.actions([      { action: "click", selector: "li.next a" },      { action: "wait", selector: ".quote" },      { action: "evaluate", expression: "({ url: location.href, authors: [...document.querySelectorAll('.quote .author')].map((a) => a.textContent) })" },    ]);    const read = results.at(-1)!.value as { url: string; authors: string[] };    console.log(`session: ${read.url}: ${read.authors.length} quotes, the first by ${read.authors[0]}`);    pages.push(read);  }  const browsed = { ms: Date.now() - t0, pages };  mkdirSync(out, { recursive: true });  writeFileSync(join(out, "result.json"), JSON.stringify({ url, fetch: { ...fetched, hasEinstein: page.content.includes("Albert Einstein") }, session: browsed }, null, 2));  console.log("Use fetch to read a page; use a session when you need to act on it (click, type, sign in, download).");} finally {  await session.stop();}

"""Get a page, or a browser? The same site two ways: fetch (one call, the page as Markdown: for reading) and a session(a browser that stays open between calls: for clicking through pages, forms and sign-ins).    python python/main.py            (BOXLINE_API_KEY in the environment; PAGE_URL for another site)"""import jsonimport osimport timefrom pathlib import Pathfrom boxline import Boxlineout = Path(os.environ.get("OUTPUT_DIR", "output"))url = os.environ.get("PAGE_URL", "https://quotes.toscrape.com/")bx = Boxline()# 1. fetch: the page is rendered in a real browser inside a sandbox and comes back as Markdown, with its links.t0 = time.monotonic()page = bx.fetch(url, format="markdown", links=True)fetched = {    "title": page["title"],    "status": page["status"],    "chars": len(page["content"]),    "links": len(page.get("links") or []),    "ms": round((time.monotonic() - t0) * 1000),    "hasEinstein": "Albert Einstein" in page["content"],}print(f'fetch: "{page["title"]}" (HTTP {page["status"]}), {fetched["chars"]} characters of Markdown and {fetched["links"]} links in {fetched["ms"]} ms')# 2. A session: the browser keeps its page, cookies and history between calls, so it can click on to page 3.with bx.sessions.create(timeout=300, user_metadata={"example": "fetch-or-browser"}) as session:    print(f"Session: {session.id}", flush=True)    t0 = time.monotonic()    session.goto(url)    pages = []    for _ in range(2):        results = session.actions([            {"action": "click", "selector": "li.next a"},            {"action": "wait", "selector": ".quote"},            {"action": "evaluate", "expression": "({ url: location.href, authors: [...document.querySelectorAll('.quote .author')].map((a) => a.textContent) })"},        ])        read = results[-1]["value"]        print(f"session: {read['url']}: {len(read['authors'])} quotes, the first by {read['authors'][0]}")        pages.append(read)    browsed = {"ms": round((time.monotonic() - t0) * 1000), "pages": pages}    out.mkdir(parents=True, exist_ok=True)    (out / "result.json").write_text(json.dumps({"url": url, "fetch": fetched, "session": browsed}, indent=2, ensure_ascii=False))    print("Use fetch to read a page; use a session when you need to act on it (click, type, sign in, download).")

# Node 18 or newer, in the example's folderexport BOXLINE_API_KEY=…   # your API keynpm install @boxline/sdk tsxnpx tsx node/index.ts

# Python 3.9 or newer, in the example's folderexport BOXLINE_API_KEY=…   # your API keypip install boxline-sdkpython python/main.py