Review Request 127215: simplify code, reduce pointer dereferences

[David Faure]

> On Feb. 29, 2016, 8:14 a.m., David Faure wrote:
> > About the unittest failure, I'm available by email to debug that, tests shouldn't fail.
> > (first thing to check will be: did a ~/.qttest/share/applications/faketextapplication.desktop get created?)
> 
> Nick Shaforostoff wrote:
>     the test started to pass once i configured bash environment for kdesrc-buil correctly

cool


> On Feb. 29, 2016, 8:14 a.m., David Faure wrote:
> > src/sycoca/kbuildmimetypefactory.cpp, line 65
> > <https://git.reviewboard.kde.org/r/127215/diff/1/?file=445933#file445933line65>
> >
> >     Can you keep the order of the operands? I find
> >       if (var == constant)
> >     more readable than the other way around.
> >     ("is this car green?", not "is green the color of this car?"). In any case that's how it is everywhere in KF5 AFAIK.
> 
> Nick Shaforostoff wrote:
>     i have changed to a natural order. but the problem is that i couldn't find an appropriate operator== overload taking (qstringref, qlatin1string), there is only a (qlatin1string, qstringref) one. is that a qt bug that should be fixed?

Sounds like it.


> On Feb. 29, 2016, 8:14 a.m., David Faure wrote:
> > src/sycoca/kbuildsycoca.cpp, line 209
> > <https://git.reviewboard.kde.org/r/127215/diff/1/?file=445935#file445935line209>
> >
> >     what is the clazy warning that encourages to write such a long line of code, rather than the two lines as it was initially?
> 
> Nick Shaforostoff wrote:
>     since Q_FOREACH takes a copy anyways i thought that putting the call inside it is a little improvement.
>     i have reverted this part.

"copy" is a big word, it only increases a refcount.


- David


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/127215/#review92884
-----------------------------------------------------------


On Feb. 29, 2016, 12:27 p.m., Nick Shaforostoff wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://git.reviewboard.kde.org/r/127215/
> -----------------------------------------------------------
> 
> (Updated Feb. 29, 2016, 12:27 p.m.)
> 
> 
> Review request for KDE Frameworks.
> 
> 
> Repository: kservice
> 
> 
> Description
> -------
> 
> Changes are mostly related to containers/iterators, but there are also few QString related optimizations
> 
> note that i have made a switch from QLinkedList to just QList because it performs better for T=<void*>
> also i have changed QStringList to QSet<QString> in one place to make the code run faster
> 
> 
> Diffs
> -----
> 
>   src/services/kservice.cpp c75bce2 
>   src/services/kservicetype.cpp 2a73ccd 
>   src/sycoca/kbuildmimetypefactory.cpp 340f76a 
>   src/sycoca/kbuildservicefactory.cpp a74f2e8 
>   src/sycoca/kbuildsycoca.cpp e99e906 
>   src/sycoca/kmimeassociations.cpp 9423b27 
>   src/sycoca/ksycocadict.cpp f33d01a 
>   src/sycoca/ksycocafactory.cpp e410581 
>   src/sycoca/ksycocautils_p.h 8db26b0 
> 
> Diff: https://git.reviewboard.kde.org/r/127215/diff/
> 
> 
> Testing
> -------
> 
> all tests pass, except kmimeassociationstest (fakeApplicationService is NULL), but it fails also without my changes
> 
> 
> Thanks,
> 
> Nick Shaforostoff
> 
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/kde-frameworks-devel/attachments/20160301/37b35bb1/attachment-0001.html>