explosion/spaCy
1
1
Fork 0
mirror of https://github.com/explosion/spaCy.git synced 2025-07-15 02:32:37 +03:00
Code Issues Packages Projects Releases Wiki Activity

* Switch parser to use nogil functions

Browse Source
This commit is contained in:
Matthew Honnibal 2016-01-30 20:27:07 +01:00
parent 10877a7791
commit 9410e74c92
2 changed files with 18 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
spacy/syntax/parser.pxd
View File

@ -9,9 +9,11 @@ from ..structs cimport TokenC
cdef class ParserModel(AveragedPerceptron): cdef class ParserModel(AveragedPerceptron):
cdef void set_featuresC(self, ExampleC* eg, StateClass stcls) except * cdef void set_featuresC(self, ExampleC* eg, StateClass stcls) nogil
cdef class Parser: cdef class Parser:
cdef readonly ParserModel model cdef readonly ParserModel model
cdef readonly TransitionSystem moves cdef readonly TransitionSystem moves
cdef void parseC(self, Doc tokens, StateClass stcls, Example eg) nogil

26
spacy/syntax/parser.pyx
View File

@ -65,7 +65,7 @@ def ParserFactory(transition_system):
cdef class ParserModel(AveragedPerceptron): cdef class ParserModel(AveragedPerceptron):
cdef void set_featuresC(self, ExampleC* eg, StateClass stcls) except *: cdef void set_featuresC(self, ExampleC* eg, StateClass stcls) nogil:
fill_context(eg.atoms, stcls) fill_context(eg.atoms, stcls)
eg.nr_feat = self.extracter.set_features(eg.features, eg.atoms) eg.nr_feat = self.extracter.set_features(eg.features, eg.atoms)
@ -95,7 +95,6 @@ cdef class Parser:
raise NotImplementedError( raise NotImplementedError(
"This should be here, but isn't yet =/. Use Parser.from_dir") "This should be here, but isn't yet =/. Use Parser.from_dir")
def __reduce__(self): def __reduce__(self):
return (Parser, (self.moves.strings, self.moves, self.model), None, None) return (Parser, (self.moves.strings, self.moves, self.model), None, None)
@ -103,11 +102,16 @@ cdef class Parser:
cdef StateClass stcls = StateClass.init(tokens.c, tokens.length) cdef StateClass stcls = StateClass.init(tokens.c, tokens.length)
self.moves.initialize_state(stcls) self.moves.initialize_state(stcls)
cdef Pool mem = Pool()
cdef Example eg = Example( cdef Example eg = Example(
nr_class=self.moves.n_moves, nr_class=self.moves.n_moves,
nr_atom=CONTEXT_SIZE, nr_atom=CONTEXT_SIZE,
nr_feat=self.model.nr_feat) nr_feat=self.model.nr_feat)
with nogil:
self.parseC(tokens, stcls, eg)
# Check for KeyboardInterrupt etc. Untested
PyErr_CheckSignals()
cdef void parseC(self, Doc tokens, StateClass stcls, Example eg) nogil:
while not stcls.is_final(): while not stcls.is_final():
self.model.set_featuresC(&eg.c, stcls) self.model.set_featuresC(&eg.c, stcls)
self.moves.set_valid(eg.c.is_valid, stcls) self.moves.set_valid(eg.c.is_valid, stcls)
@ -116,15 +120,15 @@ cdef class Parser:
guess = VecVec.arg_max_if_true(eg.c.scores, eg.c.is_valid, eg.c.nr_class) guess = VecVec.arg_max_if_true(eg.c.scores, eg.c.is_valid, eg.c.nr_class)
action = self.moves.c[guess] action = self.moves.c[guess]
if not eg.is_valid[guess]: if not eg.c.is_valid[guess]:
raise ValueError( with gil:
"Illegal action: %s" % self.moves.move_name(action.move, action.label) move_name = self.moves.move_name(action.move, action.label)
) raise ValueError("Illegal action: %s" % move_name)
action.do(stcls, action.label) action.do(stcls, action.label)
# Check for KeyboardInterrupt etc. Untested memset(eg.c.scores, 0, sizeof(eg.c.scores[0]) * eg.c.nr_class)
PyErr_CheckSignals() memset(eg.c.costs, 0, sizeof(eg.c.costs[0]) * eg.c.nr_class)
eg.reset_classes(eg.nr_class) for i in range(eg.c.nr_class):
eg.c.is_valid[i] = 1
self.moves.finalize_state(stcls) self.moves.finalize_state(stcls)
tokens.set_parse(stcls._sent) tokens.set_parse(stcls._sent)