spaCy/spacy/training/iob_utils.py

233 lines
8.6 KiB
Python
Raw Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

from typing import List, Dict, Tuple, Iterable, Union, Iterator, cast
import warnings
from ..errors import Errors, Warnings
from ..tokens import Span, Doc
def iob_to_biluo(tags: Iterable[str]) -> List[str]:
out: List[str] = []
tags = list(tags)
while tags:
out.extend(_consume_os(tags))
out.extend(_consume_ent(tags))
return out
def biluo_to_iob(tags: Iterable[str]) -> List[str]:
out = []
for tag in tags:
if tag is None:
out.append(tag)
else:
tag = tag.replace("U-", "B-", 1).replace("L-", "I-", 1)
out.append(tag)
return out
def _consume_os(tags: List[str]) -> Iterator[str]:
while tags and tags[0] == "O":
yield tags.pop(0)
def _consume_ent(tags: List[str]) -> List[str]:
if not tags:
return []
tag = tags.pop(0)
target_in = "I" + tag[1:]
target_last = "L" + tag[1:]
length = 1
while tags and tags[0] in {target_in, target_last}:
length += 1
tags.pop(0)
label = tag[2:]
if length == 1:
if len(label) == 0:
raise ValueError(Errors.E177.format(tag=tag))
return ["U-" + label]
else:
start = "B-" + label
end = "L-" + label
middle = [f"I-{label}" for _ in range(1, length - 1)]
return [start] + middle + [end]
def doc_to_biluo_tags(doc: Doc, missing: str = "O"):
return offsets_to_biluo_tags(
doc,
[(ent.start_char, ent.end_char, ent.label_) for ent in doc.ents],
missing=missing,
)
def offsets_to_biluo_tags(
doc: Doc, entities: Iterable[Tuple[int, int, Union[str, int]]], missing: str = "O"
) -> List[str]:
"""Encode labelled spans into per-token tags, using the
Begin/In/Last/Unit/Out scheme (BILUO).
doc (Doc): The document that the entity offsets refer to. The output tags
will refer to the token boundaries within the document.
entities (iterable): A sequence of `(start, end, label)` triples. `start`
and `end` should be character-offset integers denoting the slice into
the original string.
missing (str): The label used for missing values, e.g. if tokenization
doesnt align with the entity offsets. Defaults to "O".
RETURNS (list): A list of unicode strings, describing the tags. Each tag
string will be of the form either "", "O" or "{action}-{label}", where
action is one of "B", "I", "L", "U". The missing label is used where the
entity offsets don't align with the tokenization in the `Doc` object.
The training algorithm will view these as missing values. "O" denotes a
non-entity token. "B" denotes the beginning of a multi-token entity,
"I" the inside of an entity of three or more tokens, and "L" the end
of an entity of two or more tokens. "U" denotes a single-token entity.
EXAMPLE:
>>> text = 'I like London.'
>>> entities = [(len('I like '), len('I like London'), 'LOC')]
>>> doc = nlp.tokenizer(text)
>>> tags = offsets_to_biluo_tags(doc, entities)
>>> assert tags == ["O", "O", 'U-LOC', "O"]
"""
# Ensure no overlapping entity labels exist
tokens_in_ents: Dict[int, Tuple[int, int, Union[str, int]]] = {}
starts = {token.idx: token.i for token in doc}
ends = {token.idx + len(token): token.i for token in doc}
biluo = ["-" for _ in doc]
# Handle entity cases
for start_char, end_char, label in entities:
if not label:
for s in starts: # account for many-to-one
if s >= start_char and s < end_char:
biluo[starts[s]] = "O"
else:
for token_index in range(start_char, end_char):
if token_index in tokens_in_ents.keys():
raise ValueError(
Errors.E103.format(
span1=(
tokens_in_ents[token_index][0],
tokens_in_ents[token_index][1],
tokens_in_ents[token_index][2],
),
span2=(start_char, end_char, label),
)
)
tokens_in_ents[token_index] = (start_char, end_char, label)
start_token = starts.get(start_char)
end_token = ends.get(end_char)
# Only interested if the tokenization is correct
if start_token is not None and end_token is not None:
if start_token == end_token:
biluo[start_token] = f"U-{label}"
else:
biluo[start_token] = f"B-{label}"
for i in range(start_token + 1, end_token):
biluo[i] = f"I-{label}"
biluo[end_token] = f"L-{label}"
# Now distinguish the O cases from ones where we miss the tokenization
entity_chars = set()
for start_char, end_char, label in entities:
for i in range(start_char, end_char):
entity_chars.add(i)
for token in doc:
for i in range(token.idx, token.idx + len(token)):
if i in entity_chars:
break
else:
biluo[token.i] = missing
if "-" in biluo and missing != "-":
ent_str = str(entities)
warnings.warn(
Warnings.W030.format(
text=doc.text[:50] + "..." if len(doc.text) > 50 else doc.text,
entities=ent_str[:50] + "..." if len(ent_str) > 50 else ent_str,
)
)
return biluo
def biluo_tags_to_spans(doc: Doc, tags: Iterable[str]) -> List[Span]:
"""Encode per-token tags following the BILUO scheme into Span object, e.g.
to overwrite the doc.ents.
doc (Doc): The document that the BILUO tags refer to.
tags (iterable): A sequence of BILUO tags with each tag describing one
token. Each tag string will be of the form of either "", "O" or
"{action}-{label}", where action is one of "B", "I", "L", "U".
RETURNS (list): A sequence of Span objects. Each token with a missing IOB
tag is returned as a Span with an empty label.
"""
token_offsets = tags_to_entities(tags)
spans = []
for label, start_idx, end_idx in token_offsets:
span = Span(doc, start_idx, end_idx + 1, label=label)
spans.append(span)
return spans
def biluo_tags_to_offsets(
doc: Doc, tags: Iterable[str]
) -> List[Tuple[int, int, Union[str, int]]]:
"""Encode per-token tags following the BILUO scheme into entity offsets.
doc (Doc): The document that the BILUO tags refer to.
tags (iterable): A sequence of BILUO tags with each tag describing one
token. Each tags string will be of the form of either "", "O" or
"{action}-{label}", where action is one of "B", "I", "L", "U".
RETURNS (list): A sequence of `(start, end, label)` triples. `start` and
`end` will be character-offset integers denoting the slice into the
original string.
"""
spans = biluo_tags_to_spans(doc, tags)
return [(span.start_char, span.end_char, span.label_) for span in spans]
def tags_to_entities(tags: Iterable[str]) -> List[Tuple[str, int, int]]:
"""Note that the end index returned by this function is inclusive.
To use it for Span creation, increment the end by 1."""
entities = []
start = None
for i, tag in enumerate(tags):
if tag is None or tag.startswith("-"):
# TODO: We shouldn't be getting these malformed inputs. Fix this.
if start is not None:
start = None
else:
entities.append(("", i, i))
elif tag.startswith("O"):
pass
elif tag.startswith("I"):
if start is None:
raise ValueError(
Errors.E067.format(start="I", tags=list(tags)[: i + 1])
)
elif tag.startswith("U"):
entities.append((tag[2:], i, i))
elif tag.startswith("B"):
start = i
elif tag.startswith("L"):
if start is None:
raise ValueError(
Errors.E067.format(start="L", tags=list(tags)[: i + 1])
)
entities.append((tag[2:], start, i))
start = None
else:
raise ValueError(Errors.E068.format(tag=tag))
return entities
def split_bilu_label(label: str) -> Tuple[str, str]:
return cast(Tuple[str, str], label.split("-", 1))
def remove_bilu_prefix(label: str) -> str:
return label.split("-", 1)[1]
# Fallbacks to make backwards-compat easier
offsets_from_biluo_tags = biluo_tags_to_offsets
spans_from_biluo_tags = biluo_tags_to_spans
biluo_tags_from_offsets = offsets_to_biluo_tags