Scraper et stocker les données d'une page

// Scrape and Store — Extract data and save it
//
// This script navigates to a page, extracts structured data
// using DOM queries, and saves results to the workspace store.
// Combines browser automation, data extraction, and persistence.
// Requires an open workspace.

const targetUrl = "https://example.com";
const collection = "scraped-pages";

// Ensure we have a browser tab
await g.tabs.newBrowserTab();

g.log("Navigating to " + targetUrl + "...");
await g.nav.navigateToUrl(targetUrl);

// Extract data from the page
const title = await g.page.executeJS("document.title");
const heading = await g.dom.extractContent("h1");
const bodyText = await g.dom.extractContent("p");

// Build a structured record
const pageData = {
    url: targetUrl,
    title: title,
    heading: heading,
    excerpt: bodyText ? bodyText.substring(0, 200) : "",
    scrapedAt: new Date().toISOString()
};

g.log("Extracted: " + pageData.title);

// Save to the workspace store
const key = targetUrl.replace(/[^a-zA-Z0-9]/g, "-");
g.store.put(collection, key, pageData);
g.log("Saved to store: " + collection + "/" + key);

// Index for full-text search
g.search.index({
    _key: key,
    _collection: collection,
    title: pageData.title,
    heading: pageData.heading,
    excerpt: pageData.excerpt,
    url: pageData.url
});
g.log("Indexed for search");

// Verify: query it back
const results = g.search.query("example");
if (results && results.length > 0) {
    g.log("Search found " + results.length + " result(s)");
}