Extract positions of indicated spans from indicated text. Used as a precursor to the step to convert named entities identified by alternative processes into a spaCy NER format

extract_span_start_end_positions.py

from typing import List

def extract_span_start_end_positions(text: str, spans: List[str]):
    """
    Extract positions of indicated spans from indicated text.
    Adapted from : https://www.programcreek.com/python/?CodeExample=convert+to+spans

    Args:
        text: The string to be searched
        spans: The spans of interest within the string. Can be single or
        multiple contiguous words.

    Returns:
        [list of (span, start, end) tuples] mapping each token to corresponding indices
        in the text.
    """
    cur_idx = 0
    spans_w_positions = []
    for span in spans:
        tmp = text.find(span, cur_idx)
        l = len(span)
        cur_idx = tmp
        spans_w_positions.append((span, cur_idx, cur_idx + l))
        cur_idx += l

    return spans_w_positions