Feedback from code review

This commit is contained in:
Paul O'Leary McCann 2022-07-06 14:03:09 +09:00
parent 63e27b5e44
commit 8f598d7b01
2 changed files with 2 additions and 2 deletions

View File

@ -205,7 +205,7 @@ def create_gold_scores(
return out
def _spans_to_offsets(doc):
def _spans_to_offsets(doc: Doc) -> List[List[Tuple[int, int]]]:
"""Convert doc.spans to nested list of ints for comparison.
The ints are character indices, and the spans groups are sorted by key first.

View File

@ -154,7 +154,7 @@ def test_tokenization_mismatch(nlp):
for key, cluster in ref.spans.items():
char_spans[key] = []
for span in cluster:
char_spans[key].append((span[0].idx, span[-1].idx + len(span[-1])))
char_spans[key].append((span.start_char, span.end_char))
with ref.retokenize() as retokenizer:
# merge "picked up"
retokenizer.merge(ref[2:4])