Pull out the arbitrary factor of three as a named constant and
document its arbitrariness once.
Pull out the mask and bit used in each function's loop to the outer
layer of the loop, since they don't depend on the inner loop variable
(or the random value generated in that loop).
Use QTest::addRow() instead of constructing a string to pass to
newRow().
Change-Id: Ifacbcb390e00828fd47f51b0c73d0ad5f6bc8bdb
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
We've been requiring C++17 since Qt 6.0, and our qAsConst use finally
starts to bother us (QTBUG-99313), so time to port away from it
now.
Since qAsConst has exactly the same semantics as std::as_const (down
to rvalue treatment, constexpr'ness and noexcept'ness), there's really
nothing more to it than a global search-and-replace, with manual
unstaging of the actual definition and documentation in dist/,
src/corelib/doc/ and src/corelib/global/.
Task-number: QTBUG-99313
Change-Id: I4c7114444a325ad4e62d0fcbfd347d2bbfb21541
Reviewed-by: Ivan Solovev <ivan.solovev@qt.io>
Given a QTaggedPointer, users may write
taggedPtr = {};
to mean "reset it". This is error-prone: due to overload resolution,
this actually ends up calling QTaggedPointer<T>::operator=(T *),
which changes the pointer but *not* the tag, and not the implicitly
declared QTaggedPointer<T>:operator=(const QTaggedPointer<T> &)
which would reset both pointer and tag.
Given the idiomatic usage of {} is indeed to perform a full reset (cf.
std::exchange(obj, {}), std::take, etc.), work around this by disabling
the operator= overload for pointers in case an initializer list is
passed. In other words, make `={}` fall back to the implicitly
declared overload.
Note, this breaks some usages, such as
taggedPtr = {rawPtr};
but at least we get a compile error for these, and they don't look
common at all.
[ChangeLog][QtCore][QTaggedPointer] The operator assignment
taking a raw pointer has been reimplemented in order to avoid
subtle issues when assigning `{}` to a QTaggedPointer. This will
cause code that assigns a braced-init-list to a QTaggedPointer object
to stop compiling (for instance, `tagPtr = {ptr}` is now ill-formed).
Change-Id: I5e572a9b0f119ddb2df17f1797934933dff2ba7b
Task-number: QTBUG-106070
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Inserting the same key repeatedly with QMultiHash will not
test rehashing behavior because in Qt6 those entries all
end up in a linked list.
Pick-to: 6.4
Change-Id: I78c45eed0f35a13af6d6da75d7189a6933750f13
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
These tested results are all wrong and caused by internal overflows.
Note the behavior can not be fixed either as it involves moving an
already maximized QRect, which can not be done without overflow.
Change-Id: If35db68102889012c56eb149fe49bc48954d3422
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: hjk <hjk@qt.io>
This can be helpful when you calculate multiple hashes, store them in a
vector and you want to know which result belongs to which algorithm.
[ChangeLog][QtCore][QCryptographicHash] Added getter algorithm().
Change-Id: Ifcf78536f215619a6e2e3035a95598327d0ed733
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Marc Mutz <marc.mutz@qt.io>
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
This adds a move constructor, a move assignment operator and a swap
function to QCryptographicHash. This can (to name one example) be useful
when you want to store multiple hashes in a vector.
[ChangeLog][QtCore][QCryptographicHash] Added move constructor, move
assignment operator and swap() function.
Change-Id: Id54594fa69104ec25ad78581f962a021e85531c2
Reviewed-by: Marc Mutz <marc.mutz@qt.io>
And include qcore_mac_p.h where needed.
Task-number: QTBUG-99313
Change-Id: Idb1b005f1b5938e8cf329ae06ffaf0d249874db2
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
It does a check to ensure you aren't comparing outside the container.
Fixes: QTBUG-106001
Pick-to: 6.2 6.3 6.4
Change-Id: Ic6547f8247454b47baa8fffd170eef346b7f4f24
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
The QScopedPointer::take() call in comparison() test was used to
avoid a double-deletion error, because the test is creating two
QScopedPointer instances referencing the same memory.
Avoid the take() call by providing a custom DummyDeleter and
managing the memory by the extarnal std::unique_ptr.
As the test now has no test-cases for QScopedPointer::take()
calls, create a new test for this deprecated API, and guard
it with QT_DEPRECATED_SINCE checks.
Task-number: QTBUG-104858
Change-Id: Iecc28d44d76c9ce5835e6b1a1df7db30e2a9ca25
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The most common changes are:
* removing the explicit tests for deprecated APIs
* QMultiMap::insertMulti() -> QMultiMap::insert()
* QMultiMap::insert(QMultiMap) -> QMultiMap::unite(QMultiMap)
Add separate tests for the deprecated APIs, and guard them
with QT_DEPRECATED_SINCE() checks.
Task-number: QTBUG-104858
Change-Id: Ifb79212d07f20028d93d75f2b32ec3785cc93b22
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
These functions are marked as deprecated in future Qt releases.
Task-number: QTBUG-104858
Change-Id: I25d2932455d8c9e3e2d722b1c48fc2cfa2d1e679
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Those were workarounds to passing a comma to a macro, but there are ways
around it. The simplest is to just use variadic macros; another, which
has been applied to Q_DECLARE_METATYPE for a long time, is to define an
alias to the thing you're trying to use.
Change-Id: Ie4bb662dcb274440ab8bfffd17097fbf0c53eabc
Reviewed-by: Marc Mutz <marc.mutz@qt.io>
CMakeLists.txt and .cmake files of significant size
(more than 2 lines according to our check in tst_license.pl)
now have the copyright and license header.
Existing copyright statements remain intact
Task-number: QTBUG-88621
Change-Id: I3b98cdc55ead806ec81ce09af9271f9b95af97fa
Reviewed-by: Jörg Bornemann <joerg.bornemann@qt.io>
QScopedValueRollback has a few users that apply it on QAtomicInt,
which happens to work as QAtomicInt is copy-constructible and its
ctors are implicit.
But that's of course nonsense. We don't need to store the oldValue in
an atomic, nor do we need to pass the new value into the ctor as an
atomic.
So, add a QAtomicScopedValueRollback which works on std::atomic as
well as the Qt atomics, but distinguishes between the reference (which
is atomic) and the value (which isn't), and use it in one of the
users, tst_QList.
Keep it private until we know whether there's an actual need for this.
The test is a copy of tst_qscopedvaluefallback, so the occasional
oddity (like atomic op*=) should be ignored.
Task-number: QTBUG-103835
Change-Id: I3c05b3e51f465698657a02ca5521ed465386e9a6
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
Replace the current license disclaimer in files by
a SPDX-License-Identifier.
Files that have to be modified by hand are modified.
License files are organized under LICENSES directory.
Task-number: QTBUG-67283
Change-Id: Id880c92784c40f3bbde861c0d93f58151c18b9f1
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Reviewed-by: Jörg Bornemann <joerg.bornemann@qt.io>
When calling QHash::reserve(), or when creating the
internal QHashPrivate::Data structure, the value 0
for the size parameter is reserved for performing
the squeeze operation.
However commit 8a984ab772
broke it, by using the 0 value in QHashPrivate::Data
constructors as a mark that no resizing needs to be done.
This patch reverts the problematic commit (also applying
some later fixes to the code), and adds the missing
tests for Q[Multi]Hash::squeeze().
Pick-to: 6.3 6.2
Change-Id: Id644df7b2beb008e6a37b2c89b709adfbd893e25
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Calling Q[Multi]Hash::reserve(n) when n is much smaller than the
current amount of elements in the hash, could result in an infinite
loop, because at some point the algorithm could not find a free bucket
for the element.
Fixing it by returning early if the new desired capacity is less than
current.
Fixes: QTBUG-102067
Pick-to: 6.3 6.2
Change-Id: I38ef0b2168c4e2a317eedf91b2155b1fdffb1c27
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
QList and QString had them, so add them to QByteArray and
QVarLengthArray, too.
In the QVLA case, we need to jump though a hoop or two to avoid having
to duplicate all the reallocation logic. Nothing a few template tricks
cannot solve.
[ChangeLog][QtCore][QByteArray] Added resize(n, ch) overload.
[ChangeLog][QtCore][QVarLengthArray] Added resize(n, v) overload.
Fixes: QTBUG-102270
Change-Id: I0d281ae5b574f440f682e4a62427b434dcf5b687
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Don't mix unsigned and signed types in comparisons.
Pick-to: 6.3
Change-Id: Ia4ba9c114177425a21cadc8cafe8179928315a5d
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
That is, insert() doesn't overwrite an existing entry, and range
insert inserts the first of equivalent keys' values, not the last.
This allowed this author to optimize the implementation of
makeUnique() to a O(N) algorithm (was: O(N²)). Said optimization would
have been possible with the old semantics, too, but I wrote the
algorithm first and only then noticed the broken insert() behavior is
present on QFlatMap, too, so I decided not to let good code go to
waste and to fix both problems at the same time.
In order to give users a hint of the changed semantics, make the new
API opt-in until Qt 6.5, so Qt 6.4 ships with the both the old and the
new semantics disabled, where they contradict.
Fixes: QTBUG-100092
Change-Id: Ic96d8bfe6bed9068dbe8c0d7171bd8921050fd95
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
Our associative containers' iterator's value_type isn't a destructurable
type (yielding key/value). This means that something like
for (auto [k, v] : map)
doesn't even compile -- one can only "directly" iterate on the
values. For quite some time we've had QKeyValueIterator to allow
key/value iteration, but then one had to resort to a "traditional" for
loop:
for (auto i = map.keyValueBegin(), e = keyValueEnd(); i!=e; ++i)
This can be easily packaged in an adaptor class, which is what this
commmit does, thereby offering a C++17-compatible way to obtain
key/value iteration over associative containers.
Something possibly peculiar is the fact that the range so obtained is
a range of pairs of references -- not a range of references to pairs.
But that's easily explained by the fact that we have no pairs to build
references to; hence,
for (auto &[k, v] : map.asKeyValueRange())
doesn't compile (lvalue reference doesn't bind to prvalue pair).
Instead, both of these compile:
for (auto [k, v] : map.asKeyValueRange())
for (auto &&[k, v] : map.asKeyValueRange())
and in *both* cases one gets references to the keys/values in the map.
If the map is non-const, the reference to the value is mutable.
Last but not least, implement pinning for rvalue containers.
[ChangeLog][QtCore][QMap] Added asKeyValueRange().
[ChangeLog][QtCore][QMultiMap] Added asKeyValueRange().
[ChangeLog][QtCore][QHash] Added asKeyValueRange().
[ChangeLog][QtCore][QMultiHash] Added asKeyValueRange().
Task-number: QTBUG-4615
Change-Id: Ic8506bff38b2f753494b21ab76f52e05c06ffc8b
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
The existing API of QFlatMap did not allow efficient removal of
elements:
- std::remove_if does not apply, because it works by moving elements
back in the range onto those that need to be removed, which doesn't
work in flat_map's case, because, like for all associative
containers, the key in value_type is const.
- The node-based erase-loop (over it = cond ? c.erase(it) :
std::next(it)) works, but, unlike in traditional associative
containers, is quadratic, because flat_map::erase is a linear
operation.
According to Stepanov's principle of Efficient Computational Basis
(Elements of Programming, Section 1.4), we're therefore missing API.
Add it.
I couldn't make up my mind about the calling convention for the
predicate and, despite having authored a merged paper about erase_if,
can never remember what the predicate is supposed to take, so be fancy
and accept all: (*it), (it.key(), it.value()), (it.key()). This means
that unary predicates can either not be generic or must be properly
constrained to distinguish between pair<const K, V> and K, but that's
not necessarily a bad thing.
There's no reason to supply a Qt-ified removeIf on top of the standard
name, because this is private API and doubling the names would do
nothing except double the testing overhead.
Fixes: QTBUG-100983
Change-Id: I12545058958fc5d620baa770f92193c8de8b2d26
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Ulf Hermann <ulf.hermann@qt.io>
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
- current INTEGRITY development pack don't support denormals for float and double.
All values are rounded to 0.
Task-number: QTBUG-99123
Pick-to: 6.2 6.3
Change-Id: Iaaacdc4210c7ac2ec3ec337c61164a1ade0efb01
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
... which implements or assumes something about the
broken^Wnon-STL-compliant insertion behavior.
Once this has integrated into all module dependencies, we can
re-implement these APIs using STL-compatible semantics.
Task-number: QTBUG-100092
Change-Id: I54f4f5ce7addd9543866d2c399f48aff50983b88
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
The macOS standard library doesn't have std::contiguous_iterator yet, and
it doesn't seem like libc++ has it either.
Checking __cpp_lib_concepts for the C++20 official version appears to work.
Pick-to: 6.3 6.2
Change-Id: I8c31cd64de24c03b3a3f37cb393bb2f9b55a834d
Reviewed-by: Marc Mutz <marc.mutz@qt.io>
Remove Integrity and Android specific code that explicitly adds
test data to the resource files. qt_internal_add_test functions
implicitly adds test data to resources for Android and Integrity
platforms by default.
Change-Id: Ia1d58755b47442e1953462e38606f70fec262368
Reviewed-by: Assam Boudjelthia <assam.boudjelthia@qt.io>
Reviewed-by: Alexandru Croitor <alexandru.croitor@qt.io>
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
The value of __cplusplus has nothing to do with whether the library
implements wg21.link/P1115 (libstdc++ even before C++20) or not
(libc++, even in C++20).
Use the idiomatic check (#if defined(foo) && foo >= x) instead,
fixing the Android build.
Pick-to: 6.3 6.2
Change-Id: I11bcefe455a1f13865c15d4beecbd3fe32115328
Reviewed-by: Marc Mutz <marc.mutz@qt.io>
qt_internal_undefine_global_definition disables an internal global
definition that is defined by the qt_internal_add_global_definition
function for a specific target.
Remove the ability to set the custom "undefine" flag for the
definitions since it's hard to control it using the introduced
function.
Pick-to: 6.2 6.3
Task-number: QTBUG-100334
Change-Id: Ic1637d97aa51bbdd06c5b191c57a941aa208d4dc
Reviewed-by: Alexandru Croitor <alexandru.croitor@qt.io>
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
The code was trying to avoid a detach in the case no element needed to
be removed, by first running find_if() on const_iterators, and then,
after converting its result to (mutable) iterators, start the
remove_if() algorithm where find_if() left off.
But this applies the predicate to the element found by find_if() (if
any) _twice_: first just before we exit the first find_if() and then
just as we enter remove_if(), which will start by running find_if()
again, with the result of the initial find_if as 'first'.
Apart from being needlessly inefficient, this violates the
specification of Uniform Erasure, which defines sequential erase_if()
as being equivalent to remove_if() + container erase(), with the
former being specified to apply the predicate exactly once per
element.
Fix by writing the remove_if() part by hand.
Instead of doing the dance with the loop invariant documentation
twice, simply implement erase() via erase_if() (complicated a bit by
the weird passing of predicates by lvalue reference instead of by
value, as would be idiomatic). This exposes users to:
[ChangeLog][QtCore][Potentially Source-Incompatible Changes] A fix in
the implementation of the erase-like algorithms of sequential Qt
container may re-enable signed/unsigned comparison warnings previously
suppressed by having occurred in std library code. To fix, cast the
value to look for such that it has the same signedness as the
container's elements.
... but the issue would be the same had we inlined std::remove()
instead of passing a lambda to sequential_erase_if(), so it's nothing
we can, nor should, work around.
[ChangeLog][QtCore][Containers] Fixed a bug in the implementation of
most sequential Qt container's erase-like algorithms (member
removeAll()/removeIf() and free erase()/erase_if()) where the equality
operator or the predicate, respectively, was applied to the first
matching element twice. Each element is now tested exactly once.
Pick-to: 6.3 6.2
Change-Id: Ib6d24b01b40866c125406f1cd6042d4cd083ea0d
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
Restore the 'QT_NO_JAVA_STYLE_ITERATORS' and
'QT_NO_NARROWING_CONVERSIONS_IN_CONNECT' definitions for Qt
targets.
Add the function that adds global definitions for Qt targets according
to the provided scope and the target property-based switch to disable
the definition for a specific target.
Pick-to: 6.2 6.3
Task-number: QTBUG-100295
Change-Id: I28697e81f9aabc45c48d79aae1e5caea141e04e1
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Marc Mutz <marc.mutz@qt.io>
Reviewed-by: Alexandru Croitor <alexandru.croitor@qt.io>
Add a level of indirection via void_t such that
struct is_transparent {};
works, and not just
using is_transparent = <unspecified>;
Pick-to: 6.3
Change-Id: I3ca2af6a07e6989dc95abc10fb2d0078a5269e5b
Reviewed-by: Jörg Bornemann <joerg.bornemann@qt.io>
We can handle the UTF-8 case by reinterpreting it as Latin-1.
This way, the suffixIndex stays valid as a return value.
As a drive-by, optimize away toLatin1() calls by using a QVLA.
We really need a better way of converting UTF-16 -> L1 than
qt_to_latin1()...
[ChangeLog][QtCore][QVersionNumber] fromString() now takes
QAnyStringView (was: QString, QStringView, QLatin1String)
and a qsizetype pointer (was: int pointer).
Change-Id: I86abaadba2792658fbf93ccd1e0b86e3302c697c
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This completes the update to qsizetype in this class, adding a couple of
methods that need to be removed in Qt 7. They're only required where int
is not qsizetype (i.e., 64-bit platforms).
Change-Id: I0e5f6bec596a4a78bd3bfffd16c9de29bec4c637
Reviewed-by: Marc Mutz <marc.mutz@qt.io>
This does exactly what insert() on Qt associative containers does, but
allows to express the intent of using the STL-incompatible Qt insert()
semantics, in an STL-compatible way, instead of leaving the reader of
the code wondering what semantics are expected.
This is part of a very-long-term goal of fixing Qt associative
container's insert() behavior, in which QFlatMap, being an affected,
but private-API type, is used for proof-of-concept purposes.
Task-number: QTBUG-99651
Pick-to: 6.3 6.2
Change-Id: I69010285438259918aef659d3235180c1b5be696
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
We use raw pointers to the Nodes in the QHash which is
inherently fine, but we are then subject to invalidation when
nodes are moved around during deletion.
In trim() we don't actually need to iterate the linked-list
since the node we are interested in is always chain.prev
Pick-to: 6.3 6.2 6.2.3
Fixes: QTBUG-99710
Task-number: QTBUG-99224
Task-number: QTBUG-99240
Change-Id: I9c2ed69b29e3cadca013113a3553deb44d7382fc
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Reviewed-by: Jarek Kobus <jaroslaw.kobus@qt.io>
... in an attempt to foster the use of this data structure by making
it less onerous to spell.
Pick-to: 6.3 6.2
Change-Id: Ib9d17029c75278edde6ba90f65f68af179a6d230
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
QFlatMap, like its public brethren, features the broken Qt-style
insert() behavior (what the STL calls insert_or_assign()), which
makes its insert() unusable for actual STL-style insert() work,
with no replacement except the size-check-and-index-operator trick:
const auto oldSize = c.size();
auto &e = c[key];
if (c.size() != oldSize) {
// inserted
}
Even though QFlatMap::insert() appears to return the correct info,
it's useless, because the old value has been assigned over by the
time insert() returns.
Pick-to: 6.3 6.2
Change-Id: If4173c42523a128dfd22ab496dde0089ba73f41c
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
By using the bucketForHash function we can loop through and find
some appropriate keys to test the edge-case. This will then
automatically keep the test working even if some internals
of QHash changes.
We do this because certain changes which change the bucket the
pre-selected keys would end up in could make the test a no-op,
without warning. And recent and upcoming changes have changed
both this and erase(). We limit the search-space to
the minimum numBuckets * 4, where minimum numBuckets is current
128.
Change-Id: I13b0bce15ee884144e3248846be34667fb5d35cc
Reviewed-by: Marc Mutz <marc.mutz@qt.io>
- add test resources to binaries
- link Qt::Gui to tst_qpointer for static build case
Task-number: QTBUG-99123
Pick-to: 6.2 6.3
Change-Id: I311827b9c641eaf9537091b051c15f9fcbcb9f0c
Reviewed-by: Kimmo Ollila <kimmo.ollila@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
QHash::operator[] could grow the hash even if the key being
looked up already existed. This in turn invalidated all iterators.
Avoid this by refactoring findOrInsert() to not grow if the key
already exists.
Added advantage is that this should make lookups of existing keys
slightly faster.
Fixes: QTBUG-97752
Pick-to: 6.3 6.2
Change-Id: I9df30459797b42c434ba0ee299fd1d55af8d2313
Reviewed-by: Marc Mutz <marc.mutz@qt.io>
There's no sense in copying a ring buffer. Moving is enough. This
marks an important step on the way to preventing accidental copies of
ring buffer content, because the 'QList buffers' member can now no
longer be implicitly shared. While the compiler will still emit the
code for detach()ing, it will now never be executed.
Pick-to: 6.3
Change-Id: I968bfe3e50c46720ed4baca55c99c1f9c518f653
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Check, in tst_qvarlengtharray, that the forwarding header still works.
Pick-to: 6.3
Change-Id: Ia03cf48457f538287880bb676aea3fa44aeb255f
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
QHash changes some of its preconditions, so we must not call
findNode without verifying !isEmpty()
Task-number: QTBUG-91739
Task-number: QTBUG-98436
Change-Id: I2701b9a01187530f541a7c9a12db56c92f856d87
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
All Qt 6 containers have "fast" prepend these days. Except
QVLA. Instead of enabling "fast" prepend for QVLA, slowing down
idiomatic QVLA use, simply deprecate prepend().
There appear to be no users of this function in qtbase outside tests.
[ChangeLog][QtCore][Deprecation Notices][QVarLengthArray] Deprecated
prepend() because QVarLengthArray is the only Qt container without a
"fast" prepend. If you require that functionality, even though it's a
linear operation, then use insert(cbegin(), ~~~) instead.
Change-Id: I39ff1dd7d4de7fc08d5380a5a7450dd8c8996fe2
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Remove scaffolding in the test again.
[ChangeLog][QtCore][QList] Fixed a regression that caused the range
constructor to fail for pure input_iterator's.
Pick-to: 6.2
Fixes: QTBUG-99036
Change-Id: I72d01a9c44c3862c335d96538f26a453b4c7c554
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The old code had several bugs:
- it immediately clobbered *this with new state, before having copied
over the elements from the old to the new buffer
- when buffer relocation threw, it would keep the new (partially-filled)
buffer and throw away the old
- it unconditionally used std::move() for non-relocatable types, making
it impossible to restore the original buffer when a move throws
Instead of clobbering *this with new state, do all the work on the
side and change *this only once the reallocation has happened
successfully.
Also use q_uninitialized_relocate_n() and unique_ptr in the
implementation to simplify the code. The former got the necessary
update to use std::move_if_noexcept() instead of an unconditional
std::move() for the non-relocatable case.
[ChangeLog][QtCore][QVarLengthArray] The append()-like functions are
now strongly exception safe. This means reallocation will now use
copies instead of moves, unless the value_type has a noexcept move
constructor.
Fixes: QTBUG-99039
Change-Id: I031251b8d14ac045592d01caed59d4638c3d9892
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
If the in-place constructor throws, the old code had already updated
the container's size(). Fix by delaying the update to after the
in-place construction.
[ChangeLog][QtCore][QVarLengthArray] Fixed a bug whereby a failed
append() would leave the container with an inconsistent size().
Pick-to: 6.2 5.15
Change-Id: Ief1e668d945149bd8ba96c8af1398baaa7876880
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
There seems to have been no-one that checked a simple empty()/isEmpty()...
Pick-to: 6.2 5.15
Change-Id: I7fa567f556532dfa21db759719f1303a768a9732
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
When deleting the last item in a chain, without it being the last item
in the chain, then we re-use the iterator which was passed in as an
argument. This is wrong if we detached earlier in the function, and
means we return an iterator to the previously shared data.
Pick-to: 6.2
Change-Id: I7da6309e23a32073da59e7da0cbfd1d16734f1ca
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Marc Mutz <marc.mutz@qt.io>
... instead of QT_PREPEND_NAMESPACE(qHash), which is qualified (prepends at least '::'), and therefore disables ADL.
This is not a problem as long as we wrote our qHash() overloads as free functions (incl. non-hidden friends), but it should™ fail for hidden friends, so use the old using-std::swap() trick to bring QT_PREPEND_NAMESPACE(qHash) into scope, proceeding with an unqualified lookup.
Pick-to: 6.2
Change-Id: I00860b2313699849f86bfe3dd9f41db4ce993cd3
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Use variable args macros to swallow any extra commas in the
expression. To use this, the type of the exception has to be first.
Use Eddy's suggestion for a new name to avoid breaking the old macro.
[ChangeLog][QtTest] Added QVERIFY_THROWS_EXCEPTION, replacing
QVERIFY_EXCEPTION_THROWN, which has therefore been deprecated.
Change-Id: I16825c35bae0631c5fad5a9a3ace4d6edc067f83
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Q(Multi)Map mutating functions that take reference to a key and/or a
value (e.g. insert(), take(), etc.) must make sure that those references
are still valid -- that is, that the referred objects are still alive --
after the detach() call done inside those functions.
In fact, if the key/value are references into *this, one must take extra
steps in order to preserve them across the detach().
Consider the scenario where one has two shallow copies of QMap, each
accessed by a different thread, and each thread calls a mutating
function on its copy, using a reference into the map (e.g.
map.take(map.firstKey())). Let's call the shared payload of this QMap
SP, with its refcount of 2; it's important to note that the argument
(call it A) passed to the mutating function belongs to SP.
Each thread may then find the reference count to be different than 1 and
therefore do a detach() from inside the mutating function. Then this
could happen:
Thread 1: Thread 2:
detach() detach()
SP refcount != 1 => true SP refcount != 1 => true
deep copy from SP deep copy from SP
ref() the new copy ref() the new copy
SP.deref() => 1 => don't dealloc SP
set the new copy as payload
SP.deref() => 0 => dealloc SP
set the new copy as payload
use A to access the new copy use A to access the new copy
The order of ref()/deref() SP and the new copy in each thread doesn't
really matter here. What really matters is that SP has been destroyed
and that means A is a danging reference.
Fix this by keeping SP alive in the mutating functions before doing a
detach(). This can simply be realized by taking a local copy of the map
from within such functions.
remove() doesn't suffer from this because its implementation doesn't do
a bare detach() but something slightly smarter.
Change-Id: Iad974a1ad1bd5ee5d1e9378ae90947bef737b6bb
Pick-to: 6.2
Reviewed-by: Marc Mutz <marc.mutz@qt.io>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
If we detach from a shared hash while holding a reference to a key from
said shared hash then there is no guarantee for how long the reference
is valid (given a multi-thread environment).
Pick-to: 6.2
Change-Id: Ifb610753d24faca63e2c0eb8836c78d55a229001
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Marc Mutz <marc.mutz@qt.io>
We expect to produce all-bits-set in the combined OR'ed value of the
seed, so instead of counting how many bits got set and reporting that,
simply compare to -1 and count how long it took to get that far.
To make sure, I've increased the number of iterations by 50%.
Change-Id: I89446ea06b5742efb194fffd16ba37b2d93c19ef
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
QHashSeed is not a random number generator (though it uses one). It
returns the same value over and over again unless you reset it to a new,
random seed.
Fixes: QTBUG-98480
Change-Id: I89446ea06b5742efb194fffd16ba36601f08d794
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
We have some special handling in qt_windows.h,
use it instead of the original windows.h
Change-Id: I12fa45b09d3f2aad355573dce45861d7d28e1d77
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
QMultiHash::operator== crashes when comparing two unequal objects.
This patch fixes it.
Pick-to: 6.2
Fixes: QTBUG-98265
Change-Id: Ibf9fef3372a2b4581843be5f25e65cc9a55ef64d
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Beats a manual array with too wide strings. I thought even to simply
replace this with a switch (loc)... it's not like this is
performance-critical code, given it uses QString.
Change-Id: I2bbf422288924c198645fffd16a977778ff8d52d
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@gmx.de>
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
QArrayDataPointer<>::size is now a qsizetype, not the uint it used to be.
Pick-to: 6.2
Change-Id: I09d7e5a50401b46a12f29f93b2b39d646b771cfc
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Andrei Golubev <andrei.golubev@qt.io>
There's now another half of the seed which will be used by the hashers.
This is not stored in QHash, so it is never changed for the lifetime of
the application (not even when QHashSeed::setDeterministicGlobalSeed()
is called). However, we will not use it when we're in deterministic
mode.
This commit uses the compiler thread-safe statics to implement the
initialization of more than one atomic word, thus freeing us from having
to have a reserved value. As a bonus, the QT_HASH_SEED warning will only
be printed once.
Change-Id: Id2983978ad544ff79911fffd16723f1673f9a5b4
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Currently, we allocate memory for elements one by one which can get
pretty slow when adding many elements.
[ChangeLog][QtCore][QVarLengthArray] Reduced number of memory
allocations in emplace() by allocating more memory at once.
Fixes: QTBUG-97489
Pick-to: 6.2
Change-Id: Idfb5b5946b047d5215c8ed00770574249f9f5d40
Reviewed-by: Andrei Golubev <andrei.golubev@qt.io>
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
It doesn't make sense to have a recursive QSet with deleted operator==,
since it's not possible to add elements to it. Consequently declaring a
metatype for it also doesn't make sense. Remove the commented
compile-time check for it.
Task-number: QTBUG-96257
Change-Id: I74ebefb38adcbe36d5c2f317188743e1f37fe16d
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
When declaring metatypes, the metatype system tries to detect if the
comparison operators for the given type exist and automatically register
them. In case of QHash, the equality operator was enabled if the value
type provides one. But the implementation needs equality operator of
the key type as well. As a result, when the key type has no equality
operator, the metatype system detects that the equality operator is
available for the QHash itself, but the compilation for metatype
registration fails when trying to instantiate the code that uses
equality operator for the key. This is fixed by enabling equality
operators for the QHash only when both the key and value types provide
one.
The same issue existed also for QMultiHash, with the difference, that
QMultiHash didn't have the constraints even on the value type. So added
checks for both.
Fixes: QTBUG-96256
Pick-to: 6.2
Change-Id: Ib8b6d365223f2b3515cbcb1843524cd6f867a6ac
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Being a test, I'm going to abuse operator-> on end() to check
that we get what we want (a pointer past the end).
Change-Id: I7ab8d017b0fe320018820eff336d496328ade481
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
Because of the addition of the operator T*(), the expression "it + N"
where N was not exactly qsizetype but any other integer type was a
compilation failure because of ambiguous overload resolution.
With GCC it's apparently a warning:
warning: ISO C++ says that these are ambiguous, even though the worst conversion for the first is better than the worst conversion for the second:
note: candidate 1: ‘QList<T>::iterator QList<T>::iterator::operator+(qsizetype) const [with T = char; qsizetype = long long int]’
note: candidate 2: ‘operator+(char*, ptrdiff_t {aka long int})’ (built-in)
With Clang, it's an error:
error: use of overloaded operator '+' is ambiguous (with operand types 'QList<int>::const_iterator' and 'ptrdiff_t' (aka 'long'))
note: candidate function
inline const_iterator operator+(qsizetype j) const { return const_iterator(i+j); }
note: built-in candidate operator+(const int *, long)
Pick-to: 6.2
Fixes: QTBUG-96128
Change-Id: Ie72b0dd0fbe84d2caae0fffd16a06f23dd56b060
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
Less clunky due to having better constexpr support, plus fold
expressions.
Change-Id: I3eb1bd30e0124f89a052fffd16a6bc73ba79ec19
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
This patch removes most of the checks that are made using C++20
__cpp_* macros for features available in C++17 and earlier.
Library feature check macros (__cpp_lib_*) are unaffected.
Change-Id: I557b2bd0d4ff09b13837555e9880eb28e0355f64
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
- process environment/DNS are OFF for INTEGRITY
Task-number: QTBUG-96176
Pick-to: 6.2
Change-Id: I189a97f88c96a428586c31a66b8d250e04482900
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
It is no longer handled separately from Android.
This effectively reverts commit 6d50f746fe
Change-Id: Ic2d75b8c5a09895810913311ab2fe3355d4d2983
Reviewed-by: Assam Boudjelthia <assam.boudjelthia@qt.io>
This ensures that a command such as
$ ninja tst_qlocale && ninja tst_qlocale_check
will automagically build the syslocaleapp program that the test runs
from a subtest. Similar for testlib's selftests and tst_QProcess.
As a drive-by, pruned some legacy comments from when CMakeLists.txt
files were generated from .pro files.
Change-Id: I67691a8175aaef124d4104cf1898193993408bdf
Reviewed-by: Alexey Edelev <alexey.edelev@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Andrei Golubev <andrei.golubev@qt.io>
The operator checks cause compilation errors when trying to check for
their existence for recursive containers. This happens because of trying
to check for the operators on the template parameter type(s), that
inherit from the container itself, which leads to compilation errors.
Introduced alternative versions of the operator checks (with _container
suffix), that first check if the container is recursive, i.e. any of its
template parameter types inherits from the given container, and skips
the operator check, if that's the case.
The fix is done for all Qt container types that had the problem, except
for QVarLengthArray and QContiguousCache, which don't compile with
recursive parameter types for unrelated reasons.
Fixes: QTBUG-91707
Pick-to: 6.2 6.1
Change-Id: Ia1e7240b4ce240c1c44f00ca680717d182df7550
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
When trying to fix 0-length matches at the end of a QString,
be83ff65c4 actually introduced a
regression due to how lastIndexOf interprets its `from` parameter.
The "established" (=legacy) interpretation of a negative `from` is that
it is supposed to indicate that we want the last match at offset `from +
size()`. With the default from of -1, that means we want a match
starting at most at position `size() - 1` inclusive, i.e. *at* the last
position in the string. The aforementioned commit changed that, by
allowing a match at position `size()` instead, and this behavioral
change broke code.
The problem the commit tried to fix was that empty matches *are* allowed
to happen at position size(): the last match of regexp // inside the
string "test" is indeed at position 4 (the regexp matches 5 times).
Changing the meaning of negative from to include that last position (in
general: to include position `from+size()+1` as the last valid matching
position, in case of a negative `from`) has unfortunately broken client
code. Therefore, we need to revert it. This patch does that, adapting
the tests as necessary (drive-by: a broken #undef is removed).
Reverting the patch however is not sufficient. What we are facing here
is an historical API mistake that forces the default `from` (-1) to
*skip* the truly last possible match; the mistake is that thre is simply
no way to pass a negative `from` and obtain that match. This means that
the revert will now cause code like this:
str.lastIndexOf(QRE("")); // `from` defaulted to -1
NOT to return str.size(), which is counter-intuitive and wrong. Other
APIs expose this inconsistency: for instance, using
QRegularExpressionIterator would actually yield a last match at position
str.size(). Similarly, using QString::count would return `str.size()+1`.
Note that, in general, it's still possible for clients to call
str.lastIndexOf(~~~, str.size())
to get the "truly last" match.
This patch also tries to fix this case ("have our cake and eat it").
First and foremost, a couple of bugs in QByteArray and QString code are
fixed (when dealing with 0-length needles).
Second, a lastIndexOf overload is added. One overload is the "legacy"
one, that will honor the pre-existing semantics of negative `from`. The
new overload does NOT take a `from` parameter at all, and will actually
match from the truly end (by simply calling `lastIndexOf(~~~, size())`
internally).
These overloads are offered for all the existing lastIndexOf()
overloads, not only the ones taking QRE.
This means that code simply using `lastIndexOf` without any `from`
parameter get the "correct" behavior for 0-length matches, and code that
specifies one gets the legacy behavior. Matches of length > 0 are not
affected anyways, as they can't match at position size().
[ChangeLog][Important Behavior Changes] A regression in the behavior of
the lastIndexOf() function on text-related containers and views
(QString, QStringView, QByteArray, QByteArrayView, QLatin1String) has
been fixed, and the behavior made consistent and more in line with
user expectations. When lastIndexOf() is invoked with a negative `from`
position, the last match has now to start at the last character in the
container/view (before, it was at the position *past* the last
character). This makes a difference when using lastIndexOf() with a
needle that has 0 length (for instance an empty string, a regular
expression that can match 0 characters, and so on); any other case is
unaffected. To retrieve the "truly last" match, one can pass a
positive `from` offset to lastIndexOf() (basically, pass `size()` as the
`from` parameter). To make calls such as `text.lastIndexOf(~~~);`, that
do not pass any `from` parameter, behave properly, a new lastIndexOf()
overload has been added to all the text containers/views. This overload
does not take a `from` parameter at all, and will search starting from
one character past the end of the text, therefore returning a correct
result when used with needles that may yield 0-length matches. Client
code may need to be recompiled in order to use this new overload.
Conversely, client code that needs to skip the "truly last" match now
needs to pass -1 as the `from` parameter instead of relying on the
default.
Change-Id: I5e92bdcf1a57c2c3cca97b6adccf0883d00a92e5
Fixes: QTBUG-94215
Pick-to: 6.2
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Take the rvalue insert() function and turn it into the emplace()
function. Reformulate rvalue-insert using emplace(). Lvalue insert()
is using a different code path, so leave that alone. This way, we
don't need to go overboard with testing.
[ChangeLog][QtCore][QVarLengthArray] Added emplace(), emplace_back().
Change-Id: I3e1400820ae0dd1fe87fd4b4c518f7f40be39f8b
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
C++20 deprecated compound volatile statements such as pre- and
post-increments, to stress that they're not atomic. So instead of
volatile i;
~~~~;
++i;
you're now supposed to write
volatile i;
~~~~;
int j = i; // volatile load
++j;
i = j; // volatile store
which matches more closely what hardware does.
Instead of fixing every use of volatile pre- or post-increment in this
fashion individually, and realising that probably a few more Qt
modules will have the same kind of code patterns in them, write
QtPrivate functions to do the job centrally.
Change-Id: I838097bd484ef2118c071726963f103c080d2ba5
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
We missed the chance of deprecating them in 5.15, so
they'll just add to the pain of porting to 6.0. We
should not keep them around forever, though; QMap isn't
random access and so its iterators should only have
bidirectional APIs.
Pick-to: 6.2
Fixes: QTBUG-95334
Change-Id: I3577f7d25e8ab793722d2f220fd27bc85c622b0d
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
QHash::squeeze() was unconditionally calling reserve(0), which is
always allocating memory (even for 0 size).
This was leading to a confusing situation when calling squeeze() on
a default-constructed container with 0 capacity() actually allocated
memory. This is very misleading, as squeeze() is supposed to free
unneeded memory, not to allocate more.
This patch adds a check for non-zero capacity. As a result, nothing
is done for default-constructed container.
Note that this patch also affects the QSet::squeeze() behavior, because
QSet uses QHash as its underlying data type.
Task-number: QTBUG-91736
Pick-to: 6.2 6.1
Change-Id: Ib1c3c8b7b3de6ddeefea0e70b1ec71803e8fd3b3
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Andreas Buhr <andreas.buhr@qt.io>
This reverts commit c19695ab95.
Just because QSet has limited API doesn't mean we can't provide this
in an efficient way for std::unordered_set :P
Added tests.
Pick-to: 6.2
Change-Id: I4f8f0e60c810acdc666cf34f929845227ed87f3b
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
This patch introduces some test improvements to check the calls of
different methods on an empty default-constructed container.
Apart from that, many other tests are added to extend code coverage.
Task-number: QTBUG-91736
Pick-to: 6.2 6.1
Change-Id: Icc1f1342738603c9bed065b2a36c72ea60b48962
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Edward Welbourne
Marc Mutz
Giuseppe D'Angelo
Mårten Nordheim
Allan Sandfeld Jensen
Linus Jahn
Sona Kurazyan
Thiago Macieira
Ivan Solovev
Lucie Gérard
Alexandru Croitor
Dimitrios Apostolou
Volker Hilsheimer
Fabian Kosmale
Tatiana Borisova
Alexey Edelev
Andreas Buhr
Kai Köhne
Joerg Bornemann
Lars Knoll
Yuhang Zhao
Robert Löhning
Ievgenii Meshcheriakov
Marc Mutz