Search code examples
pythonstanford-nlpparse-tree

Convert constituent string to Tree object (Stanza)

I am doing some experiments in Python with Stanza and I have converted a lot of sentences into ParseTree objects and saved them into a file like this:

# sent_id = Not relevant to the question
# text = Might be ok for some , but not for me
(S (NONE Might be) (OPINION_1_Negative_Standard (POLAR_EXPRESSION ok)) (NONE for some , but) (OPINION_0_Negative_Standard (POLAR_EXPRESSION not)) (NONE for) (ALL_01 (SOURCE me)))

And now, I want to read that file, get the strings of the Tree and convert them again into a ParseTree object, but I am not able to find how it should be done properly.

I checked Data Conversion, but I can not find what I am trying to do, and the Constituency Parser examples show how to get the string representation but not how to convert that string into a Tree.

Regards.

Solution

  • I have found the solution reading the source code in their GitHub.

    Inside the constituency module, in the tree_reader file there is a method called read_trees(text) that it does exactly what I wanted.

    Regards.