Rooba
/
spaCy
mirror of https://github.com/explosion/spaCy.git
1
1
Fork 0
Code Issues Packages Projects Releases Wiki Activity

* Rename sent_start to sent_end, to reflect its new usage in the Break transition

This commit is contained in:
Matthew Honnibal 2015-06-23 05:39:23 +02:00
parent 34c0ef2ee8
commit a7bf7b0626
4 changed files with 12 additions and 15 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
spacy/structs.pxd
View File

@ -68,7 +68,7 @@ cdef struct TokenC:
int sense int sense
int head int head
int dep int dep
bint sent_end bint sent_start
uint32_t l_kids uint32_t l_kids
uint32_t r_kids uint32_t r_kids

10
spacy/syntax/arc_eager.pyx
View File

@ -114,7 +114,7 @@ cdef bint _is_gold_root(const GoldParseC* gold, int word) nogil:
cdef class Shift: cdef class Shift:
@staticmethod @staticmethod
cdef bint is_valid(StateClass st, int label) nogil: cdef bint is_valid(StateClass st, int label) nogil:
return st.buffer_length() >= 2 and not st.shifted[st.B(0)] and not st.B_(0).sent_end return st.buffer_length() >= 2 and not st.shifted[st.B(0)] and not st.B_(0).sent_start
@staticmethod @staticmethod
cdef int transition(StateClass st, int label) nogil: cdef int transition(StateClass st, int label) nogil:
@ -163,7 +163,7 @@ cdef class Reduce:
cdef class LeftArc: cdef class LeftArc:
@staticmethod @staticmethod
cdef bint is_valid(StateClass st, int label) nogil: cdef bint is_valid(StateClass st, int label) nogil:
return not st.B_(0).sent_end return not st.B_(0).sent_start
@staticmethod @staticmethod
cdef int transition(StateClass st, int label) nogil: cdef int transition(StateClass st, int label) nogil:
@ -196,7 +196,7 @@ cdef class LeftArc:
cdef class RightArc: cdef class RightArc:
@staticmethod @staticmethod
cdef bint is_valid(StateClass st, int label) nogil: cdef bint is_valid(StateClass st, int label) nogil:
return not st.B_(0).sent_end return not st.B_(0).sent_start
@staticmethod @staticmethod
cdef int transition(StateClass st, int label) nogil: cdef int transition(StateClass st, int label) nogil:
@ -367,9 +367,9 @@ cdef class ArcEager(TransitionSystem):
return t return t
cdef int initialize_state(self, StateClass st) except -1: cdef int initialize_state(self, StateClass st) except -1:
# Ensure sent_end is set to 0 throughout # Ensure sent_start is set to 0 throughout
for i in range(st.length): for i in range(st.length):
st._sent[i].sent_end = False st._sent[i].sent_start = False
st.fast_forward() st.fast_forward()
cdef int finalize_state(self, StateClass st) except -1: cdef int finalize_state(self, StateClass st) except -1:

2
spacy/syntax/stateclass.pyx
View File

@ -219,7 +219,7 @@ cdef class StateClass:
cdef void set_break(self, int _) nogil: cdef void set_break(self, int _) nogil:
if 0 <= self.B(0) < self.length: if 0 <= self.B(0) < self.length:
self._sent[self.B(0)].sent_end = True self._sent[self.B(0)].sent_start = True
self._break = self._b_i self._break = self._b_i
cdef void clone(self, StateClass src) nogil: cdef void clone(self, StateClass src) nogil:

13
spacy/tokens.pyx
View File

@ -186,15 +186,12 @@ cdef class Tokens:
""" """
cdef int i cdef int i
cdef Tokens sent = Tokens(self.vocab, self._string[self.data[0].idx:]) cdef Tokens sent = Tokens(self.vocab, self._string[self.data[0].idx:])
start = None start = 0
for i in range(self.length): for i in range(1, self.length):
if start is None: if self.data[i].sent_start:
yield Span(self, start, i)
start = i start = i
if self.data[i].sent_end: yield Span(self, start, self.length)
yield Span(self, start, i+1)
start = None
if start is not None:
yield Span(self, start, self.length)
cdef int push_back(self, int idx, LexemeOrToken lex_or_tok) except -1: cdef int push_back(self, int idx, LexemeOrToken lex_or_tok) except -1:
if self.length == self.max_length: if self.length == self.max_length: