Temporary (?) fix for multibyte char / record boundary issue.

2026-05-29 18:22:37 -04:00 · 2009-01-02 09:27:47 -05:00
parent c4582ed2a8
commit 35320295f8
1 changed files with 3 additions and 2 deletions
@@ -50,8 +50,9 @@ PALMDOC = 2
 HUFFDIC = 17480

 def encode(data):
-    #return data.encode('ascii', 'xmlcharrefreplace')
-    return data.encode('utf-8')
+    # Using UTF-8 means needing to worry about multibyte characters crossing
+    # record boundaries, so let's not for now.
+    return data.encode('ascii', 'xmlcharrefreplace')

 # Almost like the one for MS LIT, but not quite.
 def decint(value):