atag.text.xpath
Description
A function that evaluates an XPath 3.1 expression against an XML string. Matched elements are returned serialized - without XML declaration and without any re-indentation, so a subtree cut out of a document can be stored and imported on its own with its character offsets intact. Attributes, text nodes and atomic values are returned as their string value.
The *:name wildcard matches an element in any namespace, so no namespace declaration is
needed for TEI.
Parameters
| name | type | description | default value |
|---|---|---|---|
| xml | String | the document | |
| xpath | String | XPath 3.1 expression | |
| return value | list of strings | the matches, in document order; empty if none |
Example
MATCH (c:Corpus {uuid: 'sozinianer'})
UNWIND atag.text.xpath(c.xml, '/*:teiCorpus/*:TEI') AS letterXml
CREATE (l:Letter {uuid: atag.text.xpath(letterXml, '/*:TEI/@xml:id')[0], xml: letterXml})
CREATE (l)-[:PART_OF]->(c)
[0] of an empty list is null, so an attribute the element does not have sets no
property.
Taking a corpus apart is usually better declared than written out:
atag.text.import.corpus builds the same hierarchy from the
documents section of a profile, which is what the
second worked example does. This function is for the cases
the profile does not cover - inspecting a document, or cutting a fragment out of one for a
purpose of its own.