Open MaartenLMEM opened 3 years ago
Basic model for a product on Amazon.fr : 1st proposal based on our contribution practices by @zhinu
If product name = Serviette hygiénique
User enters in BO field :Serviette hygiénique
or Serviette-hygiénique
or Serviette_hygiénique
or Serviette.hygiénique
(we assume here contributor won't write Serviette hygienique
without accent.)
As a result the matching context machine could translate it this way :
Regex language = .*([^A-Za-zÀ-ÖØ-öø-ž]serviettes?([^A-Za-zÀ-ÖØ-öø-ž].*[^A-Za-zÀ-ÖØ-öø-ž]|[^A-Za-zÀ-ÖØ-öø-ž])hygi(e|é|%C3%A9)niques?([^A-Za-zÀ-ÖØ-öø-ž]|$))
Xpath language = //span[@id='productTitle']/text()[contains(translate(., 'ABCDEFGHIJKLMNOPQRSTUVWXYZÀÂÄÇÉÈÊËÎÏÙÛÜ' , 'abcdefghijklmnopqrstuvwxyzàâäçéèêëîïùûü'), 'serviette') contains(translate(., 'ABCDEFGHIJKLMNOPQRSTUVWXYZÀÂÄÇÉÈÊËÎÏÙÛÜ' , 'abcdefghijklmnopqrstuvwxyzàâäçéèêëîïùûü'), 'hygiénique')]
Following dis-moi/backend#448 in Backend
Main usecases of notices that will be filled in and that may be used here:
2019 (books)
323 (board games)
442 (lingerie)
526 (clothes)
761 (matress)
762 (pillow)
960 (cardboard box)
etc.