On Wed, 23 Jun 2010, Josh Triplett wrote:

> I ran across an interesting code formatting issue when doing some
> cleanups using spatch.  If code has function arguments aligned, and a
> semantic patch modifies the first line such that the first argument no
> longer starts in the same column, the remaining arguments will no longer
> line up with the first.
> 
> For instance:
> 
>      if ((num = gen->mb_parse_list_num))
> -        new_list = (ParseInfo *) Xrealloc(gen->mb_parse_list,
> +        new_list = Xrealloc(gen->mb_parse_list,
>                                            (num + 2) * sizeof(ParseInfo));
> 
> This seems non-trivial to fix.  Nonetheless, I thought I'd report it as
> an interesting problem that might otherwise necessitate manual editing
> after applying a semantic patch.
> 
> (This coding style in general seems like a bad idea, primarily because
> it means more lines will have to change than really need to; however,
> that doesn't stop people from using it. :) )

The pretty printing is not so good at things that involve crossing the 
boundary between semantic patch code and C code, which this in some sense 
does...

julia
_______________________________________________
Cocci mailing list
[email protected]
http://lists.diku.dk/mailman/listinfo/cocci
(Web access from inside DIKUs LAN only)

Reply via email to