explosion/spaCy
1
1
Fork 0
mirror of https://github.com/explosion/spaCy.git synced 2024-12-25 17:36:30 +03:00
Code Issues Packages Projects Releases Wiki Activity

* Fix API for loading word vectors from a file.

Browse Source
This commit is contained in:
Matthew Honnibal 2015-09-23 23:51:08 +10:00
parent 46caf15bca
commit abf0d930af
1 changed files with 16 additions and 24 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

18
spacy/vocab.pyx
View File

@ -55,7 +55,7 @@ cdef class Vocab:
self.load_lexemes(path.join(data_dir, 'strings.txt'), path.join(data_dir, 'lexemes.bin'))
if path.exists(path.join(data_dir, 'vec.bin')):
self.vectors_length = self.load_vectors(path.join(data_dir, 'vec.bin'))
self.vectors_length = self.load_vectors_from_bin_loc(path.join(data_dir, 'vec.bin'))
return self
def __init__(self, get_lex_attr=None, tag_map=None, lemmatizer=None, serializer_freqs=None):
@ -258,25 +258,17 @@ cdef class Vocab:
i += 1
fp.close()
def load_vectors(self, loc):
if loc.endswith('bz2'):
vec_len = self.load_vectors_bz2(loc)
else:
vec_len = self.load_vectors_bin(loc)
return vec_len
def load_vectors_bz2(self, loc):
def load_vectors(self, loc_or_file):
cdef LexemeC* lexeme
cdef attr_t orth
cdef int32_t vec_len = -1
with bz2.BZ2File(loc, 'r') as file_:
for line_num, line in enumerate(file_):
for line_num, line in enumerate(loc_or_file):
pieces = line.split()
word_str = pieces.pop(0)
if vec_len == -1:
vec_len = len(pieces)
elif vec_len != len(pieces):
raise VectorReadError.mismatched_sizes(loc, line_num,
raise VectorReadError.mismatched_sizes(loc_or_file, line_num,
vec_len, len(pieces))
orth = self.strings[word_str]
lexeme = <LexemeC*><void*>self.get_by_orth(self.mem, orth)
@ -286,7 +278,7 @@ cdef class Vocab:
lexeme.repvec[i] = float(val_str)
return vec_len
def load_vectors_bin(self, loc):
def load_vectors_from_bin_loc(self, loc):
cdef CFile file_ = CFile(loc, b'rb')
cdef int32_t word_len
cdef int32_t vec_len = 0