Commit a5bd1406 authored by Simon Will's avatar Simon Will
Browse files

Fix bug in reconstruct_verse_text_from_reading

parent 66973f50

allzweckmesser/corpus.py

+2 −1
Original line number Diff line number Diff line
@@ -114,7 +114,8 @@ def separate_punctuation(tokens): 




def reconstruct_verse_text_from_reading(reading):

    codepoints = [' ' for _ in range(reading.tokens[-1].span[1])]

    last_non_punct_token = [t for t in reading.tokens if not t.is_punct()]

    codepoints = [' ' for _ in range(last_non_punct_token.span[1])]

    for token in reading.tokens:

        codepoints[token.span[0]:token.span[1]] = token.text

    return ''.join(codepoints)