[Dltk-dev] Problem with DLTK Indexing

[Gerald Rosenberg <gerald@xxxxxxxxxx>]

Is there a way to control the scope of DLTK indexing?

I have an "attic" folder in my project to hold script snippets, many 
of which contain completely invalid syntax.  That folder is not on 
the build path, yet it seems that DLTK still tries to index them all.

Looks like the indexing job is trying to run the source parser on the 
snippets and becomes inundated with parser generate errors.  The 
result is a not-quite-frozen workspace -- the CPU pegs and the Java 
process grows to over 2 Meg.  If I cancel the DLTK indexing process, 
the workbench becomes unstable and I have to restart.