is possible to have accent sensitive and case insensitive utf8 collation in mysql?

Question

How can I perform accent-sensitive but case-insensitive utf8 search in mysql? Utf8_bin is case sensitive, and utf8_general_ci is accent insensitive.

Answer 1

If you want to differ "café" from "cafe" You may use :

Select word from table_words WHERE Hex(word) LIKE Hex("café");

This way it will return 'café'.

Otherwise if you use :

Select word from table_words WHERE Hex(word) LIKE Hex("cafe");

it will return café. I'm using the latin1_german2_ci Collation.

Answer 2

There doesn't seem to be one because case sensitivity is tough to do in Unicode .

There is a utf8_general_cs collation but it seems to be experimental, and according to this bug report , doesn't do what it's expected to when using LIKE.

If your data consists of western umlauts only (ie. umlauts that are included in ISO-8859-1), you might be able to collate your search operation to latin1_german2_ci or create a separate search column with it (that specific collation is accent sensitive according to this page ; latin1_general_ci might be as well, I don't know and can't test right now).

Answer 3

You can use " hex " to make the search accent-sensitive. Then simply add lcase to make it case insensitive again. So that would give:

SELECT name FROM people WHERE HEX(LCASE(name)) = HEX(LCASE("René"))

You do throw all your indexes out of the window like that. If you want to avoid having to do a full table scan and you have an index on "name", also search for the same thing without the hex and lcase:

SELECT name FROM people WHERE name = "René" and HEX(LCASE(name)) = HEX(LCASE("René"))

This way the index on " name " will be used to find for example only the rows " René " and "Rene" and then the comparison with the " hex " needs to be done only on those two rows instead of on the complete table.