3af8d34f0f
* As a consequence we are also now using `same_string' instead of `is_equal' to compare strings which will handle any type of strings. * Enforced the rule that arguments are READABLE_STRING_GENERAL and queries are STRING_32 when unicode is expected. * The most delicate part of the change was the update of the encoding library. Now we have two queries:last_converted_string_32 and last_converted_string_8 in addition of the typeless last_converted_string. The idea is that if you know that you are converting to something where characters are at least 2 bytes wide, you know that you have to use STRING_32, otherwise it will be STRING_8. * Unlike STRING_GENERAL, READABLE_STRING_GENERAL does not have a conversion to STRING_32 and thus in a few places we had to use `as_string_32' for explicit conversion which I found better. git-svn-id: https://svn.origo.ethz.ch/eiffelstudio/trunk/Src/library/cURL@85350 8089f293-4706-0410-a29e-feb5c42a2edf