Re: [AD] 3 patches regarding unicode and buffer overfows (mostly)

["Eric Botcazou" <ebotcazou@xxxxxxxxxx>]

> If anybody using Unicode in some programs could test those, it
> would be great, as Unicode with latin characters does not help much.
> Most of these fixes are here to patch crashes when using 16 bit encoding.

Ok, I've attached what I expect is the final state of the 3 patches.

But after looking through file.c, I found that there are many places where
buffer size limitations are hard-coded, so that they propagate to the file
selector and I think the goal of Annie's 'nolimit' patch can't be reached.

Moreover, as the functions from file.c don't report truncated buffer
problems, I think there is no reliable way to detect them in the file
selector either...

So I gave up and didn't modify the return value of file_select_ex().
If someone want to change that, feel free... otherwise commit the patch.

--
Eric Botcazou
ebotcazou@xxxxxxxxxx

Attachment: unicode2c.zip
Description: Zip compressed data