2014-09-15 05:22:40 +04:00
|
|
|
from cpython.ref cimport Py_INCREF
|
2014-09-18 01:09:24 +04:00
|
|
|
from cymem.cymem cimport Pool
|
2014-09-15 05:22:40 +04:00
|
|
|
|
2014-09-10 22:41:37 +04:00
|
|
|
|
2014-09-17 22:02:26 +04:00
|
|
|
cdef LexemeC* lexeme_init(Pool mem, unicode string, double prob, size_t cluster,
|
2014-09-10 22:41:37 +04:00
|
|
|
list views, set flags):
|
2014-09-17 22:02:26 +04:00
|
|
|
cdef LexemeC* lexeme = <LexemeC*>mem.alloc(1, sizeof(LexemeC))
|
2014-09-10 22:41:37 +04:00
|
|
|
lexeme.cluster = cluster
|
|
|
|
lexeme.prob = prob
|
2014-09-15 03:31:44 +04:00
|
|
|
lexeme.string = intern_and_encode(string, &lexeme.length)
|
2014-09-17 22:02:26 +04:00
|
|
|
lexeme.views = <char**>mem.alloc(len(views), sizeof(char*))
|
2014-09-15 03:31:44 +04:00
|
|
|
cdef size_t length = 0
|
2014-09-10 22:41:37 +04:00
|
|
|
for i, string in enumerate(views):
|
2014-09-15 03:31:44 +04:00
|
|
|
lexeme.views[i] = intern_and_encode(string, &length)
|
2014-09-10 22:41:37 +04:00
|
|
|
|
|
|
|
for active_flag in flags:
|
|
|
|
lexeme.flags |= (1 << active_flag)
|
|
|
|
return lexeme
|
|
|
|
|
|
|
|
|
2014-09-15 03:31:44 +04:00
|
|
|
cdef char* intern_and_encode(unicode string, size_t* length):
|
2014-09-15 08:34:45 +04:00
|
|
|
cdef bytes byte_string = string.encode('utf8')
|
|
|
|
cdef bytes utf8_string = intern(byte_string)
|
2014-09-15 05:22:40 +04:00
|
|
|
Py_INCREF(utf8_string)
|
2014-09-15 03:31:44 +04:00
|
|
|
length[0] = len(utf8_string)
|
2014-09-10 22:41:37 +04:00
|
|
|
return <char*>utf8_string
|
|
|
|
|
|
|
|
|
|
|
|
cdef bint lexeme_check_flag(LexemeC* lexeme, size_t flag_id):
|
|
|
|
return lexeme.flags & (1 << flag_id)
|
|
|
|
|
|
|
|
|
|
|
|
cdef unicode lexeme_string_view(LexemeC* lexeme, size_t view_id):
|
|
|
|
cdef bytes byte_string = lexeme.views[view_id]
|
|
|
|
return byte_string.decode('utf8')
|