explosion/spaCy
1
1
Fork 0
mirror of https://github.com/explosion/spaCy.git synced 2024-12-27 18:36:36 +03:00
Code Issues Packages Projects Releases Wiki Activity

* Add support for sentence breaks in stateclass

Browse Source
This commit is contained in:
Matthew Honnibal 2015-06-10 12:33:55 +02:00
parent b7b18c279d
commit 495f528709
2 changed files with 10 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
spacy/syntax/stateclass.pxd
View File

@ -16,10 +16,10 @@ cdef class StateClass:
cdef Entity* _ents cdef Entity* _ents
cdef TokenC _empty_token cdef TokenC _empty_token
cdef int length cdef int length
cdef bint at_sent_end
cdef int _s_i cdef int _s_i
cdef int _b_i cdef int _b_i
cdef int _e_i cdef int _e_i
cdef int _break
@staticmethod @staticmethod
cdef inline StateClass init(const TokenC* sent, int length): cdef inline StateClass init(const TokenC* sent, int length):
@ -93,6 +93,6 @@ cdef class StateClass:
cdef void set_ent_tag(self, int i, int ent_iob, int ent_type) nogil cdef void set_ent_tag(self, int i, int ent_iob, int ent_type) nogil
cdef void set_sent_end(self, int i) nogil cdef void set_break(self, int i) nogil
cdef void clone(self, StateClass src) nogil cdef void clone(self, StateClass src) nogil

12
spacy/syntax/stateclass.pyx
View File

@ -14,6 +14,7 @@ cdef class StateClass:
self._ents = <Entity*>mem.alloc(length, sizeof(Entity)) self._ents = <Entity*>mem.alloc(length, sizeof(Entity))
self.mem = mem self.mem = mem
self.length = length self.length = length
self._break = length
self._s_i = 0 self._s_i = 0
self._b_i = 0 self._b_i = 0
self._e_i = 0 self._e_i = 0
@ -104,10 +105,10 @@ cdef class StateClass:
return self._s_i <= 0 return self._s_i <= 0
cdef bint eol(self) nogil: cdef bint eol(self) nogil:
return self._b_i >= self.length or self.at_sent_end return self._b_i >= self._break
cdef bint is_final(self) nogil: cdef bint is_final(self) nogil:
return self.stack_depth() <= 1 and self.buffer_length() == 0 return self.stack_depth() <= 1 and self._b_i >= self.length
cdef bint has_head(self, int i) nogil: cdef bint has_head(self, int i) nogil:
return self.safe_get(i).head != 0 return self.safe_get(i).head != 0
@ -130,12 +131,14 @@ cdef class StateClass:
return self._s_i return self._s_i
cdef int buffer_length(self) nogil: cdef int buffer_length(self) nogil:
return self.length - self._b_i return self._break - self._b_i
cdef void push(self) nogil: cdef void push(self) nogil:
self._stack[self._s_i] = self.B(0) self._stack[self._s_i] = self.B(0)
self._s_i += 1 self._s_i += 1
self._b_i += 1 self._b_i += 1
if self._b_i >= self._break:
self._break = self.length
cdef void pop(self) nogil: cdef void pop(self) nogil:
self._s_i -= 1 self._s_i -= 1
@ -182,9 +185,10 @@ cdef class StateClass:
self._sent[i].ent_iob = ent_iob self._sent[i].ent_iob = ent_iob
self._sent[i].ent_type = ent_type self._sent[i].ent_type = ent_type
cdef void set_sent_end(self, int i) nogil: cdef void set_break(self, int i) nogil:
if 0 <= i < self.length: if 0 <= i < self.length:
self._sent[i].sent_end = True self._sent[i].sent_end = True
self._break = i
cdef void clone(self, StateClass src) nogil: cdef void clone(self, StateClass src) nogil:
memcpy(self._sent, src._sent, self.length * sizeof(TokenC)) memcpy(self._sent, src._sent, self.length * sizeof(TokenC))