kazu.steps.ner.entity_post_processing

Classes

NonContiguousEntitySplitter	Some simple rules to split non-contiguous entities into component entities.
SplitOnConjunctionPattern
SplitOnNumericalListPatternWithPrefix	split a string of numerically incrementing parts:

class kazu.steps.ner.entity_post_processing.NonContiguousEntitySplitter

Bases: object

Some simple rules to split non-contiguous entities into component entities.

__call__(entity, text)

Call self as a function.

Parameters:

• entity (Entity)

• text (str)

Return type:

list[Entity]

__init__(entity_conditions)

Parameters:

entity_conditions (dict[str, list[Callable[[Entity, str], list[Entity]]]])

class kazu.steps.ner.entity_post_processing.SplitOnConjunctionPattern

Bases: object

__call__(entity, text)

Call self as a function.

Parameters:

• entity (Entity)

• text (str)

Return type:

list[Entity]

__init__(path)

Analyse.

Parameters:

• pattern

• path (str)

run_conjunction_rules(doc, entity, text)

Parameters:

• doc (Doc)

• entity (Entity)

• text (str)

Return type:

list[Entity]

class kazu.steps.ner.entity_post_processing.SplitOnNumericalListPatternWithPrefix

Bases: object

split a string of numerically incrementing parts:

splitter = SplitOnNumericalListPatternWithPrefix()
ent = Entity.load_contiguous_entity(
    start=0, end=8, namespace="test", entity_class="gene", match="BRCA1/2/3"
)
print(splitter(ent, "BRCA1/2/3 are oncogenes"))

[BRCA1:gene:test:0:5, BRCA2:gene:test:0:7, BRCA3:gene:test:0:9]

__call__(entity, text)

Call self as a function.

Parameters:

• entity (Entity)

• text (str)

Return type:

list[Entity]

__init__(pattern='/')

Pattern to split the string on (typically “/”)

Parameters:

pattern (str)