Merge pull request #196 from megagonlabs/feature/improve_bunsetu_span

improve bunsetu span
megagonlabs · Oct 15, 2021 · ad46e8b · ad46e8b
2 parents 31a22bc + 185c4ae
commit ad46e8b
Show file tree

Hide file tree

Showing 4 changed files with 12 additions and 2 deletions.
diff --git a/README.md b/README.md
@@ -221,6 +221,11 @@ Please read the official documents to compile user dictionaries with `sudachipy`
 
 ### version 5.x
 
+#### ginza-5.0.3
+- 2021-10-15
+- Bug fix
+  - `Bunsetu span should not cross the sentence boundary` #195
+
 #### ginza-5.0.2
 - 2021-09-06
 - Bug fix

diff --git a/docs/index.md b/docs/index.md
@@ -225,6 +225,11 @@ Contains information from mC4 which is made available under the ODC Attribution
 
 ### version 5.x
 
+#### ginza-5.0.3
+- 2021-10-15
+- Bug fix
+  - `Bunsetu span should not cross the sentence boundary` #195
+
 #### ginza-5.0.2
 - 2021-09-06
 - Bug fix

diff --git a/ginza/bunsetu_recognizer.py b/ginza/bunsetu_recognizer.py
@@ -79,7 +79,7 @@ def bunsetu_span(token: Token) -> Span:
     start = token.i
     end = start + 1
     for idx in range(start, 0, -1):
-        if bunsetu_bi_list[idx] == "B":
+        if bunsetu_bi_list[idx] == "B" or token.doc[idx].is_sent_start:
             start = idx
             break
     else:

diff --git a/setup.py b/setup.py
@@ -26,5 +26,5 @@
     name="ginza",
     packages=find_packages(include=["ginza"]),
     url="https://github.com/megagonlabs/ginza",
-    version='5.0.2',
+    version='5.0.3',
 )