atag.text.import.tei
Description
Import a TEI/XML document stored on a node property, controlled by an
import profile. In contrast to atag.text.import.xml, this procedure
runs the full pipeline: it validates the document element, resolves stand-off annotations
and entity declarations, maps element and attribute names with the profile's dictionary,
and turns entity references into relationships.
Parameters
| name | type | description | default value |
|---|---|---|---|
| startNode | node | content node holding the TEI document | |
| propertyKey | string | property name of the TEI document | |
| profile | map | import profile | {} |
| return value | node | new annotation nodes |
Stand-off markup
An annotation in <standOff> addresses a range of the text with the XPointer scheme
string-range(id, start, length) and an annotation on an annotation with a plain #id
pointer:
<standOff>
<listAnnotation>
<annotation target="#string-range(text-1,5,15)" xml:id="a-3" type="phrase"/>
<annotation target="#a-1" xml:id="c-1" type="commentary"/>
</listAnnotation>
<list type="entity">
<item xml:id="hildegard" n="Hildegard von Bingen" type="Person"/>
</list>
</standOff>
Both encodings result in the same kind of annotation node, which is what makes an inline and a stand-off encoding of the same relation interchangeable.
Example
MATCH (t:Text {uuid: 'text-1'})
CALL atag.text.import.tei(t, 'xml', {
dictionary: { elements: {persName: 'person-reference', s: 'sentence'} },
createMissingEntities: true
}) YIELD node
RETURN node
The worked example walks through this import step by step and shows the graph it produces.