Re: The naive test for equality

David Cressey wrote:
> The case insensitive equality test is a very good example. Better than
> "Rationals" and better than "floating point numbers".
>
> How many times have I had to write "where upcase (x.foobar) = upcase
> (y.foobar)" in Oracle? And how many times
> has it done a sort-merge join instead of using the index, because the index
> was built using the case sensitive form?

Coincidentally, I came across this thread today when looking for something else:
http://groups.google.co.uk/group/comp.databases.postgresql.general/browse_frm/thread/24b0fdd2e35914ed/1d93176ebc116af3?lnk=st&q=korean+postgresql+sort+order&rnum=6&hl=en#1d93176ebc116af3

Apparently Japanese has two ways of writing words: "katakana" and "hiragana" which can be treated as equal. This is supposedly similar to ignoring case but not exactly. To someone who doesn't read Japanese they would look like totally different words.

As would some uppercase and lowercase versions of words written in the roman alphabet, I suppose, to someone who only knows a language with a different character set.

Paul. Received on Thu Aug 11 2005 - 20:41:37 CEST