TC6: RegEx / Unicode bug?

[gurba]

Hi

I have a bunch of unicode files in a directory tree

When I search for alternatives (e.g. "(IsMsShipped|IsUserTable)" ) and check both the Unicode and Regex boxes I get 0 files, although I know there are plenty.
A plain search for one of the alternatives returns several files.

So my guess is that the combination of RegEx and Unicode doesn't work.

Another thing is:
If I search for "table", check unicode box I get more resultfiles than if I search for "table" check both unicode and regex.
Shouldn't the results be the same as long as the search criteria doesn't include any regex "special characters"?

regards,
gurba

[ghisler(Author)]

So my guess is that the combination of RegEx and Unicode doesn't work.

Your guess is right! I should disable the other checkbox somehow.

If I search for "table", check unicode box I get more resultfiles than if I search for "table" check both unicode and regex.
Shouldn't the results be the same as long as the search criteria doesn't include any regex "special characters"?

Hmm, can you give an example of a string which isn't found this way?

[Thany]

Your guess is right! I should disable the other checkbox somehow.

You should fix the problem at its root, not work around it

I heard you are using some kind of free regex lib for delphi. I use one too in my own apps, and AFAIK, it works perfectly well with unicode strings. Or maybe you could adapt the library (if allowed by its license).

[pdavit]

I use one too in my own apps, and AFAIK, it works perfectly well with unicode strings.

So... are you going to help Christian or what!

I don't see any link with the source code!

[Thany]

Alright, here ya go...

This is the one I use

[ghisler(Author)]

Currently I have no plans to add RegEx support for Unicode, this feature would again eat a lot of memory. Btw, I use the same RegEx library, but it doesn't seem to support ANSI and Unicode at the same time - it has to be recompiled for Unicode, or included twice = twice the memory.