Are case differences handled flexible enough?
Brought to you by:
jschrod
What does an author want if he indexes "Arm" and "arm"?
(1) in German, it's clear: these are two different words.
Uppercase is sorted in front of lowercase.
(2) in English: it may be
(a) That he uses automatic software to output the
index, and it was once at the start of the
sentence and once within a sentence. Then they
should be sorted together, we might want to use
a merge-rule "A" => "a".
(b) That the one is a name and these are two
different
entries.
Currently, these are always two different words.
Do we need variants of language modules?
Or do we need to solve that in a standard base module
that might be included?
I think it's the latter, as we would specify merge
rules, not sort rules.
example raw index