XPath – text() should use normalization

[FilipJirsak]

Input document can have one element divided into multiple text nodes (text() method than returns multiple nodes). XPath matching than doesn't work - it matches text only with first text node. Text nodes should be normalized before matching XPath.
Input text is divided into multiple text nodes for example when it contains entities - for example < or >.

[reluxa]

I guess it's belong to here: I was also trying to extract text() element form an xml document. the element looked liked the following

<element>TOOXYZ:sometext /TOOXYZ:otherText</element>

The above XML fragment is present many times in the xml document which I was trying to process. Interestingly the last occurrence of the XML could not get parsed correctly the text() has returned to nodes: "TOOXYZ:sometext /TOO" and XYZ:otherText

[FilipJirsak]

@reluxa This is correct, there can be multiple adjacent text nodes. You can call document.normalize() to settle text nodes.
This issue is about XPath matching - normalization should probably be done automatically before XPath matching because XPath expects normalized documents.