explosion/spaCy
1
1
Fork 0
mirror of https://github.com/explosion/spaCy.git synced 2025-02-04 21:50:35 +03:00
Code Issues Packages Projects Releases Wiki Activity

* Switch to dynamically allocating array, based on the document length

Browse Source
This commit is contained in:
Matthew Honnibal 2014-07-07 08:05:29 +02:00
parent 0575f16ade
commit 0074ae2fc0
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/spacy.pyx
View File

@ -1,6 +1,8 @@
# cython: profile=True # cython: profile=True
from __future__ import unicode_literals from __future__ import unicode_literals
from libc.stdlib cimport calloc, free
from ext.murmurhash cimport MurmurHash64A from ext.murmurhash cimport MurmurHash64A
from ext.murmurhash cimport MurmurHash64B from ext.murmurhash cimport MurmurHash64B
@ -60,9 +62,7 @@ cdef vector[Lexeme_addr] tokenize(Vocab* vocab, dict bacov, Splitter splitter,
cdef Py_UNICODE c cdef Py_UNICODE c
cdef vector[Lexeme_addr] tokens = vector[Lexeme_addr]() cdef vector[Lexeme_addr] tokens = vector[Lexeme_addr]()
cdef Py_UNICODE[1000] current cdef Py_UNICODE* current = <Py_UNICODE*>calloc(len(string), sizeof(Py_UNICODE))
for i in range(1000):
current[i] = 0
cdef size_t word_len = 0 cdef size_t word_len = 0
cdef Lexeme* token cdef Lexeme* token
for i in range(length): for i in range(length):
@ -84,6 +84,7 @@ cdef vector[Lexeme_addr] tokenize(Vocab* vocab, dict bacov, Splitter splitter,
while token != NULL: while token != NULL:
tokens.push_back(<Lexeme_addr>token) tokens.push_back(<Lexeme_addr>token)
token = token.tail token = token.tail
free(current)
return tokens return tokens
cdef inline bint _is_whitespace(Py_UNICODE c) nogil: cdef inline bint _is_whitespace(Py_UNICODE c) nogil: