Question

I maybe be trying to do something invalid here, but maybe someone smarter than me knows the correct syntax to solve my problem.
Given:

    <group code="vehicle">
        <line code="car">
            <cell>
                <box code="price">1000.00</box>
            </cell>
        </line>
        <line code="car">
            <cell code="sports">
                <box code="price">500.00</box>
            </cell>
        </line>
    </group>

If I use //*[@code="vehicle"]//*[@code="car"]//*[@code="price"], I will get both boxes returned (1000.00 and 500.00)–as expected but not what I want.

Is there an xpath syntax that will evaluate against all nodes that have an attribute of @code rather than skipping it if it doesn’t match with the end result being that I only get back the first box (price of 1000.00)? Like asking, choose the first node with @code and that @code must equal “vehicle”, then choose the next node with @code and that @code must equal “car”, then choose the next node with @code and @code must equal “price”.

Answer 1

Use:

   "//box
       [(@code='car' or @code='price' or @code='vehicle')
      and
       not(
             ancestor-or-self::*
                   [
                       @code
                      and
                       not(@code='car'
                              or
                               @code='price'
                              or
                               @code='vehicle'
                           )

                   ]
               )
       ]