Merge pull request #3 from nsaphra/master

Giza formatting fix
proycon · May 8, 2014 · da4e731 · da4e731
2 parents 087970c + 85c07bd
commit da4e731
Showing 1 changed file with 5 additions and 7 deletions.
diff --git a/formats/giza.py b/formats/giza.py
@@ -106,21 +106,19 @@ def __repr__(self):
 
 
     def getalignedtarget(self, index):
+        """Returns target range only if source index aligns to a single consecutive range of target tokens."""
         targetindices = []
         target = None
         foundindex = -1
         for sourceindex, targetindex in self.alignment:
             if sourceindex == index:
                 targetindices.append(targetindex)
         if len(targetindices) > 1:
-            consecutive = True
             for i in range(1,len(targetindices)):
                 if abs(targetindices[i] - targetindices[i-1]) != 1:
-                    consecutive  = False
-                    break
-            if consecutive:
-                foundindex = (min(targetindices), max(targetindices))
-                target = ' '.join(self.target[min(targetindices):max(targetindices)+1])
+                    break  # not consecutive
+            foundindex = (min(targetindices), max(targetindices))
+            target = ' '.join(self.target[min(targetindices):max(targetindices)+1])
         elif targetindices:
             foundindex = targetindices[0]
             target = self.target[foundindex]
@@ -134,7 +132,7 @@ def __init__(self, filename, encoding= 'utf-8'):
         elif filename.split(".")[-1] == "gz":
             self.f = gzip.GzipFile(filename,'r')
         else:
-            self.f = io.open(filename,'r',encoding)
+            self.f = io.open(filename,'r',encoding=encoding)
         self.nextlinebuffer = None