Fwd: KDE CI: Plasma ksshaskpass kf5-qt5 SUSEQt5.9 - Build # 11 - Still Failing!

[Ben Cooksley]

On Mon, Jan 22, 2018 at 7:30 PM, Martin Koller <kollix at aon.at> wrote:
> On Sonntag, 21. Jänner 2018 22:40:31 CET Ben Cooksley wrote:
>> On Mon, Jan 22, 2018 at 10:15 AM, Martin Koller <kollix at aon.at> wrote:
>> > Hi,
>>
>> Hi Martin,
>>
>> >
>> > On Sonntag, 21. Jänner 2018 20:43:42 CET Ben Cooksley wrote:
>> >> Hi Martin,
>> >>
>> >> Looks like your latest commit to ksshaskpass doesn't build.
>> >> Could you please take a look?
>> >
>> > I don't understand this.
>> >
>> > Why does my new line fail
>> >         QRegularExpression re3a("^(Password|Username) for \"(.*?)\"$");
>> >
>> > and all the others don't
>> >         QRegularExpression re3("^(Password|Username) for (.*?)[:] $");
>> >         QRegularExpression re2("^Bad passphrase, try again for (.*?)( \\(will confirm each use\\))?: $");
>> >
>> > etc.
>> >
>> > All of these use QString(const char*)
>> >
>> > Is the define QT_NO_CAST_FROM_ASCII used during compilation ?
>> > If so, all QRegularExpression's should fail.
>> >
>> > I can of course now wrap all strings in QLatin1String() but first I'd like to understand
>> > why only my new line fails.
>>
>> From my understanding Laurent has recently been making changes
>> throughout various parts of KDE applications which do have the effect
>> of enabling QT_NO_CAST_FROM_ASCII.
>
> nice :-(
>
>> The CI system doesn't do anything
>> special compared to anyone else when compiling software - any defines
>> such as this one will be set by CMake so this should be showing up
>> locally for you as well.
>
> no, I don't see it, but I just compile ksshaskpass using the locally installed
> KF5 5.42 libs.

Interesting. It is possible that previously one of the Frameworks it
uses was disabling that flag as well, as the CI will have a newer KF5
(it rebuilds once a week essentially)

>
>
>> I'm not sure why only the new line is failing, it could be the
>> compiler only complaining about the first instance it hits though?
>
> Yes, that might be possible.
> Nevertheless, I have now wrapped all exising strings in QStringLiteral
> and it compiles when I explicitely set QT_NO_CAST_FROM_ASCII

Thanks, I can confirm it's building again.

Cheers,
Ben

>
>
> --
> Best regards/Schöne Grüße
>
> Martin
> A: Because it breaks the logical sequence of discussion
> Q: Why is top posting bad?
>
> ()  ascii ribbon campaign - against html e-mail
> /\                        - against proprietary attachments
>
> Geschenkideen, Accessoires, Seifen, Kulinarisches: www.lillehus.at
>
>