UTF-8

[Tom]

Running zero under Fedora 7, I started loading source while debugging
and got msg that zero won't load a file because of invalid UTF-8.

As far as I know, my code is 7-bit us-ascii which is supposed to be
UTF-8 valid . Any hints as to how to debug text for bad UTF-8
characters?

Thanks.

[Tom Browder]

Well, I should have looked harder. There is a Perl module,
Search::Tools::UTF8, that solved the problem for me.

Thanks anyway.

-Tom

[C. Vlasceanu]

No problem. The main reason for having that check is to prevent accidental loading of binary files, any better idea would be appreciated.

And would there be any value in having a control in the UI (or some environmental variable in the least) to disable this check?

Best,
   Cristian

[Tom Browder]

On 10/13/07, C. Vlasceanu <cristi.v...@gmail.com> wrote:
> No problem. The main reason for having that check is to prevent accidental
> loading of binary files, any better idea would be appreciated.
>
> And would there be any value in having a control in the UI (or some
> environmental variable in the least) to disable this check?

Actually, Cristian, pointing to the place in the file where zero
detected the first non-utf-8 bytes would be helpful.

I actually found an apparent non-ascii character which displays as a
'u' with an umlaut but is apparently not UTF-8 (some other encoding I
guess). BTW, the 'u' was in Peter Kummel's name in the SafeFormat.h
header file of the Loki library.

The new zero is much improved over the earlier version I used some
time ago (I'm now running Fedora 7). It is so much easier to use than
DDD or SGI's debuggers.

Thanks for a great product.

-Tom

[C. Vlasceanu]

Actually, Cristian, pointing to the place in the file where zero
detected the first non-utf-8 bytes would be helpful.

How about doing that AND asking the user whether the non-UTF8 text is acceptable?

See attached screenshot.

Cheers,
     Cristian

[Tom Browder]

On 10/14/07, C. Vlasceanu <cristi.v...@gmail.com> wrote:
>
> > Actually, Cristian, pointing to the place in the file where zero
> > detected the first non-utf-8 bytes would be helpful.
> >
>
> How about doing that AND asking the user whether the non-UTF8 text is
> acceptable?

Tres bien!

That is perfect!

-Tom

[C. Vlasceanu]

Tres bien!

That is perfect!

I just uploaded a new build for Fedora7, give it a try.

--
the-free-meme.blogspot.com