Question

I use this xpath expression to list all “NON TEST” items:

/Items/Item[State!='TEST']/Name

Normally input XML looks like this and everything works fine:

<Items>
  <Item>
    <Name>Item1</Name>
    <State>ACTIVE</State>
  </Item>
  <Item>
    <Name>Item2</Name>
    <State>TEST</State>
  </Item>
</Items>

But when Item miss State element the xpath expression selects nothing:

<Items>
  <Item>
    <Name>Item1</Name>
  </Item>
  <Item>
    <Name>Item2</Name>
  </Item>
</Items>

Please, how should I change my xpath expression so that it also works when State element does not exists? If State element does not exists Item is considered as “NON TEST” item. I have no option to change xml structure.

Answer 1

Use the following expression:

/Items/Item[not(State='TEST')]/Name

The difference is that State!='TEST' selects those items having any State element whose string value is not equal to TEST, while not(State='TEST') selects all items not having a State element whose string value is equal to TEST (whether State exists or not). It’s a subtle but significant difference.

From the spec:

If one object to be compared is a node-set and the other is a string,
then the comparison will be true if and only if there is a node in the
node-set such that the result of performing the comparison on the
string-value of the node and the other string is true.

Furthermore, imagine the following input:

<Items>
    <Item>
        <Name>Item1</Name>
        <State>blah</State>
        <State>TEST</State>
    </Item>
    <Item>
        <Name>Item2</Name>
        <State>TEST</State>
    </Item>
</Items>

This expression:

/Items/Item[State!='TEST']/Name

…will select the first Item‘s Name element, because it contains at least one State element not having a value of TEST.

However, this expression:

/Items/Item[not(State='TEST')]/Name

…selects no items. This second form is very often what you want.

See this note from the spec:

NOTE: If $x is bound to a node-set, then $x="foo" does not mean the
same as not($x!="foo"): the former is true if and only if some node in
$x has the string-value foo; the latter is true if and only if all
nodes in $x have the string-value foo.