wiktor_ner.py

"""Module for parsing WiktorNER files."""

from typing import List, Tuple

from src.annotations import Annotation, MorphosyntacticAnnotation, NerAnnotation
from src.input_parsers.interface import InputParser


class WiktorNERInputParser(InputParser):
    """Parser for WiktorNER files.

    Example WiktorNER file:
    {
        "filename": "greeting-5b1401",
        "text": "Hello Tom!",
        "tokens": [
            {
                "index": 1,
                "position": [0,5],
                "orth": "Hello",
                "lex": [
                    {
                        "lemma": "hello",
                        "mstag": "interj"
                    }
                ]
            },
            {
                "index": 2,
                "position": [6,9],
                "orth": "Tom",
                "lex": [
                    {
                        "lemma": "Tom",
                        "mstag": "noun"
                    }
                ]
            },
            {
                "index": 3,
                "position": [9,10],
                "orth": "!",
                "lex": [
                    {
                        "lemma": "!",
                        "mstag": "interp"
                    }
                ]
            }
        ],
        "entities": [
            {
                "text": "Tom",
                "type": "nam_prs_human",
                "tokens": [2],
                "positions": [6,9]
            }
        ]
    }
    """

    def parse(self, content: str) -> Tuple[str, List[Tuple[int, int, Annotation]]]:
        """Parse wiktorner file into text and annotations.

        Annotations are returned as a dictionary with channel name as a key and list of
        tuples.

        Args:
            co z  (str): Path to file containing CCL.

        Returns:
            Tuple[str, List[Tuple[int, int, Annotation]]]: Text and annotations.

        """
        if content.text:
            text = content.text
        else:
            text = ""

        annotations = []
        # Morphosyntactic annotations
        if content.tokens():
            for token in content.tokens():
                if token.lexemes:
                    for lexeme in token.lexemes:
                        if lexeme.disamb and lexeme.disamb is True:
                            if lexeme.pos:
                                if lexeme.lemma:
                                    lemma = lexeme.lemma
                                else:
                                    lemma = None
                                annotations.append(
                                    (
                                        token.start,
                                        token.stop,
                                        MorphosyntacticAnnotation(
                                            lexeme.pos, lemma
                                        ),
                                    )
                                )
        # NER annotations
        if 'ner' in content.get_span_types():
            for entity in content.spans('ner'):
                if entity.type:
                    annotations.append(
                        (entity.start, entity.stop, NerAnnotation(entity.type))
                    )

        return text, annotations