LCase/UCase to work on .c

[Mistrel]

LCase and UCase only work on strings. What about chars?

[Fred]

Debug UCase(Chr('a')) ?

[Mistrel]

Is it possible to change the case of a unicode character without changing it to a string? From a performance perspective I parse using character arrays because it deals purely with integers. For a string search I convert the input string to lower case before running a comparison.

I think this operation would be faster if I could change the case of only the characters I parse. But this only seems feasible without having to change the character to a string first.

[Trond]

No, but you can always change it back with Asc().

[Mistrel]

Caught me while I was editing, Trond. See my clarification. Maybe you can provide a different solution.

[Trond]

You could have converted the characters to lowercase before putting them into the array if case is not significant.

[AND51]

Note, that the ASCII-characters have a special offset.
Between 'A' and 'a' are 32 others characters:

Code: Select all

Debug 'a'-'A'
Debug 'b'-'B'
; and so on...

So you can easily add or substract 32 to each .c-variable to turn it into upper or lower case.

With pointer you can even scan whole strings.

[Trond]

Note, that the ASCII-characters have a special offset.
Between 'A' and 'a' are 32 others characters:

Code: Select all

Debug 'a'-'A'
Debug 'b'-'B'
; and so on...

So you can easily add or substract 32 to each .c-variable to turn it into upper or lower case.

Remember to check if it's actually a letter first. Subtracting 32 from a symbol like backspace will not create a "lower-case backspace".

[flaith]

Or with XOR

Code: Select all

a.c = Asc("A")  ; = 65  ;1000001
; or
;a.c = Asc("a")  ; = 97  ;1100001

Debug Chr(a)

b.c = a ! %00100000

Debug Chr(b)

[Mistrel]

Excellent! I had no idea about this significant offset. Thank you.

[flaith]

you're welcome

[Trond]

Also it doesn't work for ß and such characters.