comphelper::string::replaceAt contained pointer arithmetic errors. In
testing, this function only worked at all in the trivial append case.
Inserting a non-empty string anywhere else in the buffer would either
only insert part of the string, corrupt the string in the buffer, or
both.
In addition to replacing the implementation, this change adds unit tests
so we can be more confident that this function works in the future.
Change-Id: I97e109dc513b08528d6d20e7dbf6eedb3040670c
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/205222
Tested-by: Jenkins
Reviewed-by: Jonathan Clark <jonathan@libreoffice.org>
It makes no sense to insist on a random type, only to reinterpret_cast
it everywhere. An opaque void fits the case perfectly.
Change-Id: If945b1c59d98c5f561350db649f29fe8036e7cdb
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/186982
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
+ improve SALT management by taking the max between encryptedHashInput.size()
and comphelper::roundUp(mInfo.saltSize, mInfo.blockSize)
+ add QA tests
Change-Id: I90dce73928e0b204500ac659eb80cd1499a4f065
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/184407
Tested-by: Jenkins
Reviewed-by: Julien Nabet <serval2412@yahoo.fr>
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
Issue spotted when trying encryption and decryption with OpenSSL,
which doesn't handle the shorter keys correctly.
Also can't use the same instance of Encrypt after an attempt that
failed (0 bytes written) as the blocks are inter-dependent and
it looks like OpenSSL already changes the state.
Change-Id: I061764e6727c7fbe600e11c5a5a6b8889d15c4e7
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/177305
Reviewed-by: Tomaž Vajngerl <quikee@gmail.com>
Tested-by: Jenkins
This adds AES256 test of encryption and decryption - an example
on how the Encrypt and Decrypt classes are used.
Also add a convenience function for Hash, accepting vector of
bytes as this is used in a lot of places.
Change-Id: I9b395c7afb8fac45cae8d7d8bd983f5daaafd64b
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/176887
Tested-by: Jenkins
Reviewed-by: Tomaž Vajngerl <quikee@gmail.com>
Also move the relevant tests and clean-up the names a bit to make
it mroe consistent.
Change-Id: I929ef9c13b954fd6a506471231d1bc41e4ef9980
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/176027
Tested-by: Jenkins
Reviewed-by: Tomaž Vajngerl <quikee@gmail.com>
tostring to convert a vector of bytes into a hex string is useful
in other tests and implementation, so move it into common hash.hxx
Change-Id: Iebc71919c6209222eb01b78c307420afddc8caca
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/176020
Tested-by: Jenkins
Reviewed-by: Tomaž Vajngerl <quikee@gmail.com>
1. Make it take string view: helps to avoid extra string allocation
e.g. in desktop::jsonToPropertyValuesVector, and will help more,
when C++26 stringstream ctor taking string view is available.
2. Factor out a function taking boost::property_tree::ptree, making
implementation simpler for [][]com.sun.star.beans.PropertyValue and
[]com.sun.star.beans.PropertyValue, without writing a child node to
a JSON string, and parsing it again.
Change-Id: I16ac2641633ea67a7c9c054c9df09a790500e6fb
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/175361
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
As part of the efforts in tdf#146266 to move parts of the codebase from
the older makefile.mk build system to the newer gbuild system, this
commit migrates the WeakBag test, and cleans it up a little.
The test can now be invoked using this command:
$ make CppunitTest_comphelper_weakbag
Change-Id: I4ac3771fa23e28abd38a2dab84cacd1ae57fb70f
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/165591
Tested-by: Hossein <hossein@libreoffice.org>
Reviewed-by: Hossein <hossein@libreoffice.org>
...introduced in 28cc0bff10 "Break
comphelper::adjustIndexToStartOfSurrogate out of o3tl::iterateCodePoints" in
preparation for bd5c358258 "a11y: Fix returning
unpaired surrogates when retrieving characters", but which never picked it up
Change-Id: I5a10a3053d2f51333e27b74a808a41df7c1912ba
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/152183
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
...as what they do is orthogonal (and it turned out that the use case that
motivated the addition of o3tl::iterateCodePoints in the first place needs them
independently, anyway)
Change-Id: Id33901a2f7ac627253654ee6d883305dcf5a456f
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/151415
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
Needed for an uncoming .uno:TextFormFields uno command where one of the
parameters has the []com.sun.star.beans.PropertyValues type at an UNO
level, and we can't provide JSON at the moment that would express that.
Change-Id: I288a540b2fcac0e5a4a82bca235199c559ba2d0c
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/143302
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
Tested-by: Jenkins
69e9925ded "sdext.pdfimport: resolves tdf#104597:
RTL script text runs are reversed" and f6004e1c45
"tdf#151546: RTL text is reversed (Writer pdfimport)" had introduced two calls
to comphelper::string::reverseString into sdext. That function reverts on the
basis of individual UTF-16 code units, not on the basis of Unicode code points.
And while at least some pre-existing callers of that function want the former
semantics (see below), these two new callers in sdext apparently want the latter
semantics. Therefore, introduce an additional function
comphelper::string::reverseCodePoints with the latter semantics.
I identified three other places that call comphelper::string::reverseString:
* SbRtl_StrReverse in basic/source/runtime/methods1.cxx apparently implements
some StrReverse Basic function, where a (presumably non-existing) Basic spec
would need to decide which of the two semantics is called for. So leave it
alone for now.
* SvtFileDialog::IsolateFilterFromPath_Impl in fpicker/source/office/iodlg.cxx
reverts a string, operates on it, then reverts (parts of) it back. Whether or
not that is the most elegant code, using the latter semantics here would
apparently be wrong, as double invocation of
comphelper::string::reverseCodePoints is not idempotent when the input is a
malformed sequence of UTF-16 code units containing a low surrogate followed by
a high surrogate.
* AccessibleCell::getCellName in svx/source/table/accessiblecell.cxx apparently
always operates on a string consisting only of Latin uppercase letters A--Z,
for which both semantics are equivalent. (So we can just as well stick with
the simpler comphelper::string::reverseString here.)
(Extending the tests in comphelper/qa/string/test_string.cxx ran into an issue
where loplugin:stringliteralvar warns about deliberate uses of sal_Unicode
arrays rather than UTF-16 string literals wrapped in OUStringLiteral, as those
arrays deliberately contain malformed UTF-16 code unit sequences and thus
converting them into UTF-16 string literals might be considered inappropriate,
see the newly added comment at
StringLiteralVar::isPotentiallyInitializedWithMalformedUtf16 in
compilerplugins/clang/stringliteralvar.cxx for details. So that loplugin had to
be improved here, too.)
Change-Id: I641cc32c76b0c5f6339ae44d8aa85df0022ffb05
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/142949
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
All call sites already effectively asked for an unsigned return type, including:
* The ThreadPool ctor took an nWorkers argument of type sal_Int32, but
internally stores that as std::size_t mnMaxWorkers.
* ZipOutputStream::reduceScheduledThreadTasksToGivenNumberOrLess apparently
benefits from an unsigned nThreadTasks parameter, getting rid of various casts
in its implementation that were necessary to silence signed vs. unsigned
comparison warnings.
The only drawback is that
comphelper::ThreadPool::getPreferredConcurrency() * 4
in package/source/zippackage/ZipPackageStream.cxx would now silently wrap around
instead of causing UB on overflow (which could be detected with appropriate
tools). Ideally, it would use some o3tl::saturating_mul if we had that, so add
a TODO comment for now.
While std::thread::hardware_concurrency returns unsigned, it looked more natural
to go with std::size_t here, as some call sites already used that (see above),
so the implementation of ThreadPool::getPreferredConcurrency is a natural place
to hide clamping std::thread::hardware_concurrency() to std::size_t (in the
unlikely case that std::size_t is of smaller rank than unsigned).
This required addition of o3tl::clamp_to_unsigned in o3tl/safeint.hxx.
Change-Id: I0a04a8b32e63ebfeb39f924c4b38520455a6fb38
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/135309
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
...presumably added back in the day to avoid warnings about signed vs. unsigned
comparisons with old compiler versions
Change-Id: Id95e272949e48adf0e75883368cbcd421e289b33
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/135292
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
Natural string compare function doesn't take into account that
the string can start with a number and in this case it treats it
like a conventional string compare. This change takes this case
into account.
This change also refactores the tests for NaturalStringSorter
class. The previous tet used a mock XBreakIterator and XCollator
implementations to test the functionallity. This is not needed as
we can just use a real one instead, which makes the test more
real as it actually uses a real implementation instead of a mock
implementation, which could differ. This change removes the mock
XCollator and XBreakIterator implementations and moves the test
into a new file - NaturalStringSortTest.cxx
The test is also extended with the new use case where the string
starts with a number.
Change-Id: I32ea055f914c2947e4d979093b32f56170a61102
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/134540
Tested-by: Jenkins
Reviewed-by: Tomaž Vajngerl <quikee@gmail.com>
Move each test case into its own test function. Also add the
missing test for ValueRestorationGuard.
Change-Id: I588ab67f82ba82ef67939dac3d22438e8799ce11
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133917
Tested-by: Jenkins
Reviewed-by: Tomaž Vajngerl <quikee@gmail.com>
...which is unnecessary (short of dubious explicit invocations like
`makePropertyValue<int const &>(i)`) as T can never be deduced to be a reference
type there. And add some test code (that also tests the bit-field scenario
mentioned in the commit message of 23cded985b
"Specialize comphelper::makePropertyValue for arithmetic types").
Change-Id: If41ce20ab643041dd9721d05373eeb32fd099ead
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/123805
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
The scenarios are:
1. Calling sequence's begin() and end() in pairs to pass to algorithms
(both calls use getArray(), which does the COW checks)
2. In addition to #1, calling end() again when checking result of find
algorithms, and/or begin() to calculate result's distance
3. Using non-const sequences in range-based for loops, which internally
do #1
4. Assigning sequence to another sequence variable, and then modifying
one of them
In many cases, the sequences could be made const, or treated as const
for the purposes of the algorithms (using std::as_const, std::cbegin,
and std::cend). Where algorithm modifies the sequence, it was changed
to only call getArray() once. For that, css::uno::toNonConstRange was
introduced, which returns a struct (sublclass of std::pair) with two
iterators [begin, end], that are calculated using one call to begin()
and one call to getLength().
To handle #4, css::uno::Sequence::swap was introduced, that swaps the
internal pointer to uno_Sequence. So when a local Sequence variable
should be assigned to another variable, and the latter will be modified
further, it's now possible to use swap instead, so the two sequences
are kept independent.
The modified places were found by temporarily removing non-const end().
Change-Id: I8fe2787f200eecb70744e8b77fbdf7a49653f628
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/123542
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
Instead of nestable b(egin) and e(nd) events, generate the allegedly
deprecated S(tart)' and F(inish) events.
Makes the code simpler. (And AsyncEvent is still unused. Unclear
whether adding it was based on a misunderstanding.)
Change-Id: Ic029b67e0951dda775c0d0af009f2e431ae55e53
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/117700
Tested-by: Jenkins
Reviewed-by: Tor Lillqvist <tml@collabora.com>
Andrea Gelmini