Normalization of user-provided spans

[Tomarchelone]

As I understand from the docs, the normalize() method can be called on text spans, that were built by tag_ner() method.

Is there a way to normalize a span of text, that does not appear in doc.spans? Can it be provided by the user?

[kuk]

There is no public API for that. One may use internal normalize https://github.com/natasha/natasha/blob/master/natasha/norm.py#L126 and syntax_normalize https://github.com/natasha/natasha/blob/master/natasha/norm.py#L119. Please read the code to learn how to prepare vocab and tokens arguments.