Web Scraper Experimental Debug

1from apify_client import ApifyClient
2
3# Initialize the ApifyClient with your Apify API token
4client = ApifyClient("<YOUR_API_TOKEN>")
5
6# Prepare the Actor input
7run_input = {
8    "startUrls": [{ "url": "https://apify.com" }],
9    "pseudoUrls": [{ "purl": "https://apify.com[(/[\\w-]+)?]" }],
10    "linkSelector": "a",
11    "pageFunction": """async function pageFunction(context) {
12    // See README for context properties. If the syntax is unfamiliar see the link
13    // https://javascript.info/destructuring-assignment#object-destructuring
14    const { request, log, jQuery } = context;
15
16    // To be able to use jQuery as $, one needs save it into a variable
17    // and select the inject jQuery option. We've selected it for you.
18    const $ = jQuery;
19    const title = $('title').text();
20
21    // This is yet another new feature of Javascript called template strings.
22    // https://javascript.info/string#quotes
23    log.info(`URL: ${request.url} TITLE: ${title}`);
24
25    // To save data just return an object with the requested properties.
26    return {
27        url: request.url,
28        title
29    };
30}""",
31    "proxyConfiguration": { "useApifyProxy": False },
32    "initialCookies": [],
33    "waitUntil": ["networkidle2"],
34    "customData": {},
35}
36
37# Run the Actor and wait for it to finish
38run = client.actor("mtrunkat/web-scraper-experimental-dbgr").call(run_input=run_input)
39
40# Fetch and print Actor results from the run's dataset (if there are any)
41print("💾 Check your data here: https://console.apify.com/storage/datasets/" + run["defaultDatasetId"])
42for item in client.dataset(run["defaultDatasetId"]).iterate_items():
43    print(item)
44
45# 📚 Want to learn more 📖? Go to → https://docs.apify.com/api/client/python/docs/quick-start