Code review comment for lp:~aaron-whitehouse/duplicity/08-merge-glob-parsers

Thanks Kenneth,

No, there should be no changes to the man page, as there should be no change to behaviour from how the functions should work. In practice I have not found any situations where the functional behaviour changed.

There are minor changes to responses to individual functions in certain cases, as these were returning different values depending on whether the glob string had a wildcard or not. In these cases, my view was that the non-wildcard responses were less correct than the wildcard responses, but it will be rare that these would have a real-world impact, e.g.:
- the select function returning "None" (silent on the file) vs 0 (do not include the file) for something included in an exclude line; or
- the selection function returning "1" (always include) vs "2" (include only if contents are marked for inclusion).

I mention all of this as there is some risk of changes to functional behaviour (which is why I waited until the 0.8 series), but there is no intended change to what it should be doing. Potentially it will fix some unexpected glob parsing behaviour.