FrenchJobMarketCrawler

A simple crawler for French job market information - backed by Pôle Emploi

MIT License

Stars
7
Committers
2

Get French job market data with Ple Emploi

Hello all,

this is a simple crawler we use at Whire.

It replicates data from Ple Emploi candidates website directly to your "./dest" repository! Syncing it every other week will give you fresh insights on how market is evolving.

So, what do you get?

For each state department and each OGR code, you get a JSON file of data we could fetch (salary, contract types, tension between offer and demand...), representing the state of last week French job market.

Here is a WYSIWYG :

 { moinsDe35: [ 1500, 4650 ],
   plusDe35: [ 1750, 7750 ],
   lastWeek: [ 294, 1563 ],
   lastYear: [ 12, 10 ],
   saison: false,
   months: [ 'Janvier', 'Mars', 'Octobre' ],
   families: [ 'Attachs commerciaux' ],
   charts:
    [ 'CDI:80.35',
      'Autre:8.93',
      'CDD  3 mois:7.05',
      'CDD <= 3 mois:3.1',
      'Intrim:0.57' ],
   code: '11383',
   zone: '75' }

   /**

   {
    moinsDe35: min to max salary for people under the age of 35
     plusDe35: min to max salary for people over the age of 35
     lastWeek: index0 is the number of offers last week, index1 the number of candidates
     lastYear: last year, this was the actual ratio between number of offers and number of candidates
     saison: is it a seasonal job?
     months: these are the months where recruitment intensifies
     families: families used internally at Ple Emploi
     charts: ordered by appearance, it gives you the percentage of contract types for successful recruitments over last year
     code: this is your OGR code, use at your own convenience or not (probably not)
     zone: the state department where we got the data from
   }
   */

To launch it,

simply

npm install

then,

npm start

to launch it silently, or

npm run verbose

to get the logs of each added file.

Feel free to use PM2 or to bind it to your DB rather than pushing to "./dest"!

Questions or collaboration?

Send me an email : hugo at whire dot me!

Pull requests would also be greatly appreciated, especially to parallelize the process!