Review Request 122071: Fix parsing environment hash

[Milian Wolff]

-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/122071/#review74101
-----------------------------------------------------------



duchain/clangparsingenvironment.cpp
<https://git.reviewboard.kde.org/r/122071/#comment51462>

    I'm not good with hash functions, so I'm reluctant to accept this. I certainly see the issue though. Maybe we could instead do the following, which wouldn't be more work /and/ save to use:
    
    convert the hash to a list of QByteArrays (or std::strings) of the form "-DFOO=Bar", i.e. just as we need it later on in ParseSession anyways. This list can then be sorted for the hash function and we could keep using qHash + KDevHash::operator<<.



duchain/clangparsingenvironment.cpp
<https://git.reviewboard.kde.org/r/122071/#comment51463>

    I'm afraid this maybe introducing subtle bugs. The order of the include paths is /not/ random. If you have the same file in two folders, the order of the include paths decides which one to prefer over the other, no?


- Milian Wolff


On Jan. 15, 2015, 10:26 a.m., Sergey Kalinichev wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://git.reviewboard.kde.org/r/122071/
> -----------------------------------------------------------
> 
> (Updated Jan. 15, 2015, 10:26 a.m.)
> 
> 
> Review request for KDevelop.
> 
> 
> Repository: kdev-clang
> 
> 
> Description
> -------
> 
> The hash should stay the same, no matter what order of includes/defines is.
> 
> With this patch applied subsequent project loading time reduced from 150 to 10 seconds for the kdev-clang project. 
> 
> 
> Diffs
> -----
> 
>   duchain/clangparsingenvironment.cpp f73822f 
>   duchain/clangparsingenvironmentfile.cpp 88cbbe0 
>   tests/test_duchain.h 22bacd5 
>   tests/test_duchain.cpp cd21635 
> 
> Diff: https://git.reviewboard.kde.org/r/122071/diff/
> 
> 
> Testing
> -------
> 
> 
> Thanks,
> 
> Sergey Kalinichev
> 
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/kdevelop-devel/attachments/20150115/1d51cf22/attachment.html>