GitLab

- Remove usage of QT_PRIVATE_SLOT, and the _k_ prefix from method names, neither
  are needed any more

NO_CHANGELOG

NO_CHANGELOG

By deprecating it and adding a new signal KReplace::textReplaced(QString &,
int, int, int).

For the rationale see KF6 task https://phabricator.kde.org/T13986

And add a new signal textFoundAtId().

See KF6 task: https://phabricator.kde.org/T13986

And add a new signal, textFound(QString &, int, int); this is a more descriptive
name (thanks to dfaure for the idea); using a unique signal name makes it
easier to the signal in connect() calls without having to cast or use
qOverload/QOverload ...etc.

The next step is replacing the highlight(int, int, int) with a new signal,
textFoundAtId().

See KF6 task: https://phabricator.kde.org/T13986

This fixes a QRegularExpression porting bug, QRegExp is always Unicode-aware
whereas with QRegularExpression Unicode support is enabled by the setting
UseUnicodePropertiesOption flag.

Add a small unit test.

For more details see:
ktexteditor#10

GIT_SILENT

NO_CHANGELOG

GIT_SILENT

GIT_SILENT

Should help produce better diffs and clang-format won't squash the enum
on one line.

GIT_SILENT

See sdk/kdesrc-build!82

GIT_SILENT

Frameworks will be Frameworks 6 at some point and there is no good reason to have the major version in the variable name.

Given this is purely internal we can to this now, making it a bit more future-proof

GIT_SILENT

NO_CHANGELOG

In C++20 there are new classes that have member functions named emit().
c.f.:
https://lists.qt-project.org/pipermail/development/2020-February/038812.html
https://en.cppreference.com/w/cpp/io/basic_osyncstream/emit

GIT_SILENT

Namespace __ktextwidgets should ensure no clashes with member names of
subclasses, while not using d avoids danger of forgetting Q:D in
new methods.

NO_CHANGELOG

NO_CHANGELOG

Saves one memory allocation per subclass instance and one pointer member

NO_CHANGELOG

NO_CHANGELOG

KReplaceDialogPrivate::_k_slotOk() doesn't seem to be used/connected to
anything as far as I can see, I commented it out and everything build
just fine. Anyway, it's now been ported to QRegularExpression (but not
tested).

Also wrap "#include <QRegExp>"... etc with deprecation wrappers, this
was forgotten when the code was ported to QRegularExpression the first
time around.