# remark: the web page we obtain will inevitably contains more than one article, # we keep only the first one: body: (//div[contains(concat(' ',normalize-space(@class),' '),' article__edito ')])[1] author: (//header[contains(concat(' ',normalize-space(@class),' '),' article__header ')])[1]//p[contains(concat(' ',normalize-space(@class),' '),' author-info__content__text ')]/a tidy: no prune: no strip: //script strip_id_or_class: support-block strip_id_or_class: subscription-box strip_id_or_class: container_hidden test_url: https://limprevu.fr/articles/17-05-2018/pour-embrasser-le-numerique-noubliez-pas-la-langue/