Doctolib

1import { ApifyClient } from 'apify-client';
2
3// Initialize the ApifyClient with your Apify API token
4// Replace the '<YOUR_API_TOKEN>' with your token
5const client = new ApifyClient({
6    token: '<YOUR_API_TOKEN>',
7});
8
9// Prepare Actor input
10const input = {
11    "startUrls": [
12        {
13            "url": "https://www.doctolib.fr/infectiologue/75001-paris"
14        }
15    ],
16    "pageFunction": async function pageFunction(context) {
17    
18        let data = {}
19        let userData = context.request.userData
20        data.url = context.request.url
21        data.label = userData.label
22        
23        if(userData && userData.label === 'doctor'){   
24            data.nom = await context.page.locator('#main-content h1').innerText({timeout:6000})
25            data.tarif = await context.innerTextwrapper(context,'#payment_means')
26            data.horaire_contact = await context.innerTextwrapper(context,'#openings_and_contact')
27            data.description = await context.innerTextwrapper(context,'.dl-profile-bio')
28            data.specialite = await context.innerTextwrapper(context,'.dl-profile-header-speciality')
29            data.expertise = await context.innerTextwrapper(context,'#skills')
30            try{
31                data.phones = await context.getPhones(data.horaire_contact)
32            }catch(e){
33                context.log.info('Phones not found',e);     
34            }
35            try{
36                data.image = await context.page.locator('.dl-profile img').first().getAttribute('src',{timeout:2000})
37                if(data.image.startsWith('/')){ data.image = 'https:' + data.image}
38            }catch(e){
39                context.log.info('Image not found',e);     
40            }        
41            
42        }else{
43            context.log.info('we are not on a doctor page: so a search or pagination page.');
44            userData.label = 'doctor';
45            const elements = context.page.locator('.search-result-card a[href]');
46            const links = await elements.evaluateAll(elems => elems.map(elem => elem.getAttribute('href')));
47            let extenstion = 'fr'
48            if(context.request.url.includes('doctolib.de')){ extenstion = 'de' }
49            if(context.request.url.includes('doctolib.it')){ extenstion = 'it' }
50            links.forEach(async link => {
51                if(link.startsWith('/')){ link = `https://www.doctolib.${extenstion}${link}` }
52                await context.enqueueRequest(link, userData , true);
53            })
54    
55        }
56        context.log.info(`ending this page now`);
57        delete data.label
58        return data;
59    }
60};
61
62// Run the Actor and wait for it to finish
63const run = await client.actor("anchor/doctolib").call(input);
64
65// Fetch and print Actor results from the run's dataset (if any)
66console.log('Results from dataset');
67console.log(`💾 Check your data here: https://console.apify.com/storage/datasets/${run.defaultDatasetId}`);
68const { items } = await client.dataset(run.defaultDatasetId).listItems();
69items.forEach((item) => {
70    console.dir(item);
71});
72
73// 📚 Want to learn more 📖? Go to → https://docs.apify.com/api/client/js/docs