Review Request: C++-Parser: Eat typename

[Andreas Pakulat]

> On 2010-05-04 21:57:16, Andreas Pakulat wrote:
> > I'd rather have a proper fix in 4.1 than removing something during the parsing process.

Sorry, checked the "ship it" checkbox accidentally, please don't ship this.


- Andreas


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
http://reviewboard.kde.org/r/3845/#review5421
-----------------------------------------------------------


On 2010-04-29 12:49:57, The-User wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> http://reviewboard.kde.org/r/3845/
> -----------------------------------------------------------
> 
> (Updated 2010-04-29 12:49:57)
> 
> 
> Review request for KDevelop.
> 
> 
> Summary
> -------
> 
> At the kdevelop-ml somebody complained that the problems-tab is not useful. That is true when you are using stl-headers, then you get dozens of erros. But 95% of the errors are caused by the missing suppport of the typename-keyword in the C++-parser. It is difficult to implement a perfect solution, because you have to care about, that it will not be used outside of templates. But there is a very simple solution removing all those errors, although some wrong code will be accepted. However, I have never heard about anybody using too many typenames. So that should not be a problem, we are an IDE and not a compiler (alternatively we are using -fpermissive :D). The patch makes the problems-tab usable and it needs only two lines of code, so it would not be a problem to include it in the release. Btw.: It is not only a STL-problem, there are also KDE-headers with a lot of template-stuff.
> 
> 
> Diffs
> -----
> 
>   trunk/extragear/sdk/kdevelop/languages/cpp/parser/parser.cpp 1118581 
> 
> Diff: http://reviewboard.kde.org/r/3845/diff
> 
> 
> Testing
> -------
> 
> Yes, unit-tests passed, most errors are still detected.
> 
> 
> Thanks,
> 
> The-User
> 
>