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

* Add dependency post-process rule to ensure spaces are attached to neighbouring tokens, so that they can't be sentence boundaries

This commit is contained in:
Matthew Honnibal 2015-10-10 15:32:13 +11:00
parent 1521cf25c9
commit 8b39feefbe
1 changed files with 6 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
spacy/syntax/arc_eager.pyx
View File

@ -9,7 +9,8 @@ from .transition_system cimport do_func_t, get_cost_func_t
from .transition_system cimport move_cost_func_t, label_cost_func_t from .transition_system cimport move_cost_func_t, label_cost_func_t
from ..gold cimport GoldParse from ..gold cimport GoldParse
from ..gold cimport GoldParseC from ..gold cimport GoldParseC
from ..attrs cimport TAG, HEAD, DEP, ENT_IOB, ENT_TYPE from ..attrs cimport TAG, HEAD, DEP, ENT_IOB, ENT_TYPE, IS_SPACE
from ..lexeme cimport Lexeme
from libc.stdint cimport uint32_t from libc.stdint cimport uint32_t
from libc.string cimport memcpy from libc.string cimport memcpy
@ -380,7 +381,10 @@ cdef class ArcEager(TransitionSystem):
cdef int finalize_state(self, StateClass st) nogil: cdef int finalize_state(self, StateClass st) nogil:
for i in range(st.length): for i in range(st.length):
if st._sent[i].head == 0 and st._sent[i].dep == 0: # Always attach spaces to the previous word
if Lexeme.c_check_flag(st._sent[i].lex, IS_SPACE):
st._sent[i].head = -1 if (i >= 1) else 1
elif st._sent[i].head == 0 and st._sent[i].dep == 0:
st._sent[i].dep = self.root_label st._sent[i].dep = self.root_label
# If we're not using the Break transition, we segment via root-labelled # If we're not using the Break transition, we segment via root-labelled
# arcs between the root words. # arcs between the root words.