explosion/spaCy
1
1
Fork 0
mirror of https://github.com/explosion/spaCy.git synced 2025-01-26 17:24:41 +03:00
Code Issues Packages Projects Releases Wiki Activity

* Fix whitespace attachment, so that left/right children are consistent with head.

Browse Source
This commit is contained in:
Matthew Honnibal 2015-10-13 15:58:22 +11:00
parent 63df729edd
commit c70eb776ae
1 changed files with 4 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
spacy/syntax/arc_eager.pyx
View File

@ -384,7 +384,10 @@ cdef class ArcEager(TransitionSystem):
for i in range(st.length):
# 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
if i >= 1:
st.add_arc(i-1, i, st._sent[i].dep)
else:
st.add_arc(i+1, i, st._sent[i].dep)
if st._sent[i].sent_start and st._sent[i].head == -1:
st._sent[i].sent_start = False
# If we had this space token as the start of a sentence,