Collator and Turkish (probably a bug)

Question

Hi,
Turkish has 2 unique letters pairs:
'\u0130' & 'i' ('&#304' & 'i') which correspond to English 'I', & 'i'
'I', & '\u0130' ('I' & 'ı') which don't exist as letters in English and represent back-vowel pairs of English 'I', & 'i'.
If you didn't get them above, you can check them out at:
http://www.prustinteractive.com/toolbox/font/
In other words, English I i are both with a dot in Turkish, and the back-vowel versions of them are both dotless.
My task at hand is:
sort them alphabetically, ignoring case, but naturally capturing the dot difference.
From the API it appears that either:
langCollator.setStrength(Collator.SECONDARY|Collator.CANONICAL_DECOMPOSITION);
or
langCollator.setStrength(Collator.SECONDARY);
should do the job.
However, all combinations of containing PRIMARY & SECONDARY fail to distinguish between the dotfulls and the dotless. The only thing that gets both of them to compare != 0 is TERTIARY or (a logical | with) Collator.FULL_DECOMPOSITION. But the moment i do that i am no longer able to ignore case.
I did a lot of testing, and am convinced that i'll need to submit a bug, but wanted to get any feedback first. If it is a bug, i'll have to figure out RuleBasedCollator, probably by listing the entire alphabet.
P.S. In fact, letters '�' and 'o', for example, don't compare to 0 in Turkish version of Collator even with Collator.PRIMARY, apparently because they are a part of the alphabet, whereas '�' and 'a' always compare to 0 even with SECONDARY, which is again a bug even though '�' and 'a' are not part of the alphabet.
I'm not really concerned whether PRIMARY or SECONDARY should compare the dotless and the dotfulls != 0, but right now none of them does, if there is a requirement of ignoring case.
Thanks for any feedback,
Reshat.

Answer 1

Hi,
a correction:
in the message below, in P.S.:
"whereas '�' and 'a' always compare to 0 even with SECONDARY, which is again a bug even though '�' and 'a' are not part of the alphabet."
should have been written as:
"but '�' and 'a' also != 0 even with PRIMARY, which is again a bug because '�' and 'a' are not part of the Turkish alphabet."
Sorry for my mistake in previous message.