# HG changeset patch
# User Bernhard Reiter <bernhard@intevation.de>
# Date 1487682848 -3600
# Node ID 8b2f8f4398179b55bf146b4a79b96e6880136bb9
# Parent  81f75c9aac844f7088545339819753f00d3889e0
Improves: ding parser.

* Strips greater and lesser signs in the beginning and end of words
  when reading a ding directory. Words enclosed by those characters seem
  to be variants. This affects about 100 to 200 words for de in de-en 1.7.

diff -r 81f75c9aac84 -r 8b2f8f439817 ppgen.py
--- a/ppgen.py	Mon Feb 13 08:38:06 2017 +0100
+++ b/ppgen.py	Tue Feb 21 14:14:08 2017 +0100
@@ -102,7 +102,7 @@
             languageEntry = p[0] if useLeft else p[2]
 
             for word in splitter.split(languageEntry):
-                word = word.strip('(",.)\'!:;').rstrip('/')
+                word = word.strip('(",.)\'!:;<>').rstrip('/')
                 if len(word) > 2 and not word[0] in '[{/':
                     dset.add(word)