Rooba
/
spaCy
mirror of https://github.com/explosion/spaCy.git
1
1
Fork 0
Code Issues Packages Projects Releases Wiki Activity

* Replace from_package with load() classmethod in Vocab

This commit is contained in:
Matthew Honnibal 2015-12-29 16:56:51 +01:00
parent c5902f2b4b
commit 0e2498da00
1 changed files with 4 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
spacy/vocab.pyx
View File

@ -19,6 +19,7 @@ from .orth cimport word_shape
from .typedefs cimport attr_t from .typedefs cimport attr_t
from .cfile cimport CFile from .cfile cimport CFile
from .lemmatizer import Lemmatizer from .lemmatizer import Lemmatizer
from .util import MockPackage
from . import attrs from . import attrs
from . import symbols from . import symbols
@ -47,11 +48,12 @@ cdef class Vocab:
'''A map container for a language's LexemeC structs. '''A map container for a language's LexemeC structs.
''' '''
@classmethod @classmethod
def from_package(cls, package, get_lex_attr=None): def load(cls, pkg_or_str_or_file, get_lex_attr=None):
package = MockPackage.create_or_return(pkg_or_str_or_file)
tag_map = package.load_utf8(json.load, tag_map = package.load_utf8(json.load,
'vocab', 'tag_map.json') 'vocab', 'tag_map.json')
lemmatizer = Lemmatizer.from_package(package) lemmatizer = Lemmatizer.load(package)
serializer_freqs = package.load_utf8(json.load, serializer_freqs = package.load_utf8(json.load,
'vocab', 'serializer.json', 'vocab', 'serializer.json',
@ -67,7 +69,6 @@ cdef class Vocab:
if package.has_file('vocab', 'vec.bin'): # TODO: really optional? if package.has_file('vocab', 'vec.bin'): # TODO: really optional?
self.vectors_length = self.load_vectors_from_bin_loc( self.vectors_length = self.load_vectors_from_bin_loc(
package.file_path('vocab', 'vec.bin')) package.file_path('vocab', 'vec.bin'))
return self return self
def __init__(self, get_lex_attr=None, tag_map=None, lemmatizer=None, serializer_freqs=None): def __init__(self, get_lex_attr=None, tag_map=None, lemmatizer=None, serializer_freqs=None):