mirror of
https://github.com/explosion/spaCy.git
synced 2024-12-26 01:46:28 +03:00
Merge branch 'master' of https://github.com/honnibal/spaCy
This commit is contained in:
commit
4083059650
|
@ -38,6 +38,7 @@ def install_data(url, extract_path, download_path):
|
||||||
assert tmp == download_path
|
assert tmp == download_path
|
||||||
t = tarfile.open(download_path)
|
t = tarfile.open(download_path)
|
||||||
t.extractall(extract_path)
|
t.extractall(extract_path)
|
||||||
|
os.unlink(download_path)
|
||||||
|
|
||||||
|
|
||||||
@plac.annotations(
|
@plac.annotations(
|
||||||
|
|
|
@ -88,6 +88,8 @@ cdef class Morphology:
|
||||||
return orth
|
return orth
|
||||||
cdef unicode py_string = self.strings[orth]
|
cdef unicode py_string = self.strings[orth]
|
||||||
if pos != NOUN and pos != VERB and pos != ADJ and pos != PUNCT:
|
if pos != NOUN and pos != VERB and pos != ADJ and pos != PUNCT:
|
||||||
|
# TODO: This should lower-case
|
||||||
|
# return self.strings[py_string.lower()]
|
||||||
return orth
|
return orth
|
||||||
cdef set lemma_strings
|
cdef set lemma_strings
|
||||||
cdef unicode lemma_string
|
cdef unicode lemma_string
|
||||||
|
|
|
@ -473,7 +473,7 @@ cdef class Doc:
|
||||||
token.lex = lex
|
token.lex = lex
|
||||||
token.spacy = self.data[end-1].spacy
|
token.spacy = self.data[end-1].spacy
|
||||||
if tag in self.vocab.morphology.tag_map:
|
if tag in self.vocab.morphology.tag_map:
|
||||||
self.vocab.morphology.assign_tag(token, self.vocab.strings[tag])
|
self.vocab.morphology.assign_tag(token, tag)
|
||||||
else:
|
else:
|
||||||
token.tag = self.vocab.strings[tag]
|
token.tag = self.vocab.strings[tag]
|
||||||
token.tag = self.vocab.strings[tag]
|
token.tag = self.vocab.strings[tag]
|
||||||
|
|
Loading…
Reference in New Issue
Block a user