Incorrect XPath returned by Article API
  • palakchokshi July 2011

    I think the Xpath returned from the Article API is incorrect.

    E.g. try the URL http://www.nytimes.com/2011/07/12/world/europe/12yard.html?_r=1&hp
    The Xpath returned is /HTML[1]/BODY[1]/DIV[1]/DIV[2]/DIV[3]/DIV[1]/DIV[1]/DIV[1]/DIV[1]/DIV[7]
    Now if you viewed the source at the URL you would see that

    /HTML[1]/BODY[1]/DIV[1] corresponds to

    As you can see that's an empty DIV hence anything that comes after /HTML[1]/BODY[1]/DIV[1] would be invalid. Hence when I use document.evaluate and pass the XPath returned by DiffBot I get a null.

    My guess is the Xpath generating code is considering zero based indexing instead of 1 based indexing (XPath uses 1 based indexing). Please let me know if I got something wrong in my understanding.

    Thanks

  • Mike September 2011

    Thanks for your question Palak. I suspect this has to do with a discrepancy in how the DOMs are parsed. Either nytimes.com is returning a different document to the Diffbot fetchers, or that document is parsed differently by the HTML parsers.

Add a Comment

Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Sign In Apply for Membership

Categories

In this Discussion

  • Mike September 2011

Tagged