Re: [PATCH v2] riched20: Check for NULL in ME_InsertTextFromCursor and add test

[Huw Davies]

On Thu, Jan 05, 2017 at 05:06:42PM +0100, Fabian Maurer wrote:
> It already works with a null-pointer, but will crash if len is -1.
> Since null-pointer means no text, we just return to avoid crashes.
>
> v2: Followed suggestion by Nikolay Sivov, moved check and extend test

Actually, I'd prefer the NULL ptr check in fnTextSrv_TxSetText(), where
it was in your first version.  However, the extended test is much better.

Huw.

[PATCH v2] riched20: Check for NULL in ME_InsertTextFromCursor and add test

[Fabian Maurer]

On Monday, January 9, 2017 1:17:43 PM CET Huw Davies wrote:
> Actually, I'd prefer the NULL ptr check in fnTextSrv_TxSetText(), where
> it was in your first version.  However, the extended test is much better.

Thanks for the feedback, I re-sent the original patch with the extended tests.

I also added a patch to move calculation of the length out of 
ME_InsertTextFromCursor since I find the code flow easier to follow that way.
Now all checks are done before calling the function, and len always means 
length.
Not sure what you think about that idea though, so feel free to reject that.

Fabian