explosion/spaCy
1
1
Fork 0
mirror of https://github.com/explosion/spaCy.git synced 2025-01-13 02:36:32 +03:00
Code Issues Packages Projects Releases Wiki Activity

* Improve interface for PointerHash

Browse Source
This commit is contained in:
Matthew Honnibal 2014-09-13 17:29:58 +02:00
parent 45865be37e
commit f3393cf57c
3 changed files with 13 additions and 13 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
spacy/_hashing.pxd
View File

@ -1,7 +1,7 @@
from libc.stdint cimport uint64_t from libc.stdint cimport uint64_t
ctypedef uint64_t key_t ctypedef uint64_t key_t
ctypedef size_t val_t ctypedef void* val_t
cdef struct Cell: cdef struct Cell:
@ -15,6 +15,6 @@ cdef class PointerHash:
cdef Cell* _last cdef Cell* _last
cdef Cell* cells cdef Cell* cells
cdef val_t lookup(self, key_t key) cdef val_t get(self, key_t key)
cdef void insert(self, key_t key, val_t value) except * cdef void set(self, key_t key, val_t value) except *
cdef void resize(self, size_t new_size) except * cdef void resize(self, size_t new_size) except *

2
spacy/lang.pxd
View File

@ -25,7 +25,7 @@ cdef class Lexicon:
cpdef readonly size_t size cpdef readonly size_t size
cpdef Lexeme lookup(self, unicode string) cpdef Lexeme lookup(self, unicode string)
cdef size_t get(self, String* s) cdef LexemeC* get(self, String* s)
cdef PointerHash _dict cdef PointerHash _dict

18
spacy/lang.pyx
View File

@ -122,7 +122,6 @@ cdef class Language:
cdef int split cdef int split
cdef int remaining = string.n cdef int remaining = string.n
cdef String prefix cdef String prefix
cdef Cell* tmp_cell
while remaining >= 1: while remaining >= 1:
split = self._split_one(string.chars, string.n) split = self._split_one(string.chars, string.n)
remaining -= split remaining -= split
@ -194,10 +193,11 @@ cdef class Lexicon:
self._dict.set(string.key, lexeme) self._dict.set(string.key, lexeme)
self.size += 1 self.size += 1
cdef size_t get(self, String* string): cdef LexemeC* get(self, String* string):
cdef LexemeC* lex_addr = <LexemeC*>self._dict.get(string.key) cdef LexemeC* lexeme
if lex_addr != NULL: lexeme = <LexemeC*>self._dict.get(string.key)
return <size_t>lex_addr if lexeme != NULL:
return lexeme
cdef unicode uni_string = string.chars[:string.n] cdef unicode uni_string = string.chars[:string.n]
views = [string_view(uni_string, 0.0, 0, {}, {}) views = [string_view(uni_string, 0.0, 0, {}, {})
@ -207,10 +207,10 @@ cdef class Lexicon:
if flag_feature(uni_string, 0.0, {}, {}): if flag_feature(uni_string, 0.0, {}, {}):
flags.add(i) flags.add(i)
cdef LexemeC* lexeme = lexeme_init(uni_string, 0, 0, views, flags) lexeme = lexeme_init(uni_string, 0, 0, views, flags)
self._dict.set(string.key, lexeme) self._dict.set(string.key, lexeme)
self.size += 1 self.size += 1
return <size_t>lexeme return lexeme
cpdef Lexeme lookup(self, unicode uni_string): cpdef Lexeme lookup(self, unicode uni_string):
"""Retrieve (or create, if not found) a Lexeme for a string, and return it. """Retrieve (or create, if not found) a Lexeme for a string, and return it.
@ -223,8 +223,8 @@ cdef class Lexicon:
""" """
cdef String string cdef String string
string_from_unicode(&string, uni_string) string_from_unicode(&string, uni_string)
cdef size_t lexeme = self.get(&string) cdef LexemeC* lexeme = self.get(&string)
return Lexeme(lexeme) return Lexeme(<size_t>lexeme)
cdef void string_from_unicode(String* s, unicode uni): cdef void string_from_unicode(String* s, unicode uni):