Commit 3bed04eb authored by Grzegorz Kostkowski's avatar Grzegorz Kostkowski

Add possibility to set annotation for expression (more than one token)

parent 9328f692
Pipeline #2684 passed with stage
in 29 seconds
......@@ -13,6 +13,7 @@ __all__ = [
......@@ -195,12 +196,34 @@ def set_annotation_for_token(sentence, token, key, value=None, set_head=False):
value (int, bool): annotation number (convertible to integer)
idx = 0 if set_head else None
set_annotation_for_tokens(sentence, [token], key, value=value, head_index=idx)
def set_annotation_for_tokens(sentence, tokens, key, value=None, head_index=None):
Set annotation for a group of tokens from same sentence. Tokens are treated
as representation of single expression and thus they get same annotation
number (value).
sentence (Corpus2.Sentence)
tokens (list of Corpus2.Token)
key (str): a name for annotation channel
value (int, bool): annotation number (convertible to integer)
head_index(int): index of token from passed list (counting starts from 0),
which will be marked as a head of annotation. If not
given, then head will not be set.
if not tokens or not isinstance(tokens, list):
raise ValueError(f"List of tokens not given or invalid format: "
f"{tokens} (type: {type(tokens)})")
ann_sentence = annotate_sentence(sentence)
if key not in ann_sentence.all_channels():
channel = ann_sentence.get_channel(key)
token_index = _find_token(sentence, token)
if value is not None:
segment = int(value)
......@@ -208,9 +231,16 @@ def set_annotation_for_token(sentence, token, key, value=None, set_head=False):
raise Exception("Wrong value type - should be an integer.")
segment = channel.get_new_segment_index()
channel.set_segment_at(token_index, segment)
if set_head:
channel.set_head_at(token_index, True)
for i, token in enumerate(tokens):
token_index = _find_token(sentence, token)
channel.set_segment_at(token_index, segment)
if head_index is not None:
if head_index < 0 or head_index >= len(tokens):
raise ValueError(f"head_index ({head_index}) does not match "
f"passed list of tokens.")
if i == head_index:
channel.set_head_at(token_index, True)
def is_head_of(sentence, token, key):
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment