Differential D20714

akonadi: fix some clazy warnings
ClosedPublic
Actions

Authored by dfaure on Apr 21 2019, 4:20 PM.

Details

Reviewers

dvratil
vkrause

Commits

R165:4463bb8a2003: akonadi: fix some clazy warnings

Summary

relation.h:30:1: warning: Move qHash(Akonadi::Relation) into Akonadi namespace for ADL lookup [-Wclazy-qhash-namespace]
itemcreatehandlertest.cpp:874:30: warning: Don't call QVector::operator[]() on temporary [-Wclazy-detaching-temporary]
tagmodeltest.cpp:266:5: warning: unused QList<QVariant> (removedTagList)

Test Plan

builds, ctest passes

Diff Detail

Repository

R165 Akonadi

Lint

Automatic diff as part of commit; lint not applicable.

Unit

Automatic diff as part of commit; unit tests not applicable.

dfaure created this revision.Apr 21 2019, 4:20 PM

Restricted Application added a project: KDE PIM. · View Herald TranscriptApr 21 2019, 4:20 PM

dfaure requested review of this revision.Apr 21 2019, 4:20 PM

Harbormaster completed remote builds in B11067: Diff 56666.Apr 21 2019, 4:20 PM

vkrause accepted this revision.Apr 21 2019, 7:23 PM

This revision is now accepted and ready to land.Apr 21 2019, 7:23 PM

Closed by commit R165:4463bb8a2003: akonadi: fix some clazy warnings (authored by dfaure). · Explain WhyApr 21 2019, 7:59 PM

This revision was automatically updated to reflect the committed changes.

dvratil added inline comments.Apr 23 2019, 9:33 AM

src/server/handler/itemcreatehandler.cpp
402	@dfaure This looks suspicious. There's a difference in how this is translated in the SQL query: the original version gets translated into `... WHERE gid = ""`, while the new version gets translated into `... WHERE gid = NULL`, which is always `false` (a null never compares to another null). Even if you would change it to `IS NULL` this would still yield different results, since an empty value is not the same as a null value in SQL.

src/server/handler/itemcreatehandler.cpp
402	@smartins Interesting. This means this is a clazy false positive. Did I miss a caveat somewhere explaining that the change should only be made if empty vs null makes no difference, unlike here? I'll revert and add a silence-clazy comment.

smartins added inline comments.Apr 27 2019, 4:07 PM

src/server/handler/itemcreatehandler.cpp
402	README-empty-qstringliteral.md says: "Note: Beware that `QString().isNull()` is `true` while both `QStringLiteral().isEmpty()` and `QStringLiteral("").isNull()` are `false`. So be sure not to introduce subtle bugs when doing such replacements."

smartins added inline comments.Apr 27 2019, 4:52 PM

src/server/handler/itemcreatehandler.cpp
402	Just QLatin1String("") even, I've checked with heaptrack and it doesn't allocate memory (unlike QString(""))

Revision Contents
Changeset List

			Path	Packages
M			autotests/libs/tagmodeltest.cpp (4 lines)
M			autotests/server/itemcreatehandlertest.cpp (2 lines)
M			src/core/relation.h (2 lines)
M			src/server/handler/itemcreatehandler.cpp (2 lines)