Age | Commit message (Collapse) | Author |
|
found by inspecting call sites of OUString::getToken
Change-Id: I4269c7476c7aa46fac39528227e350568f0eb34a
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/132644
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
|
|
Change-Id: Idcf537a838bb0e47cfa31db8bc09e477b277ee37
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133875
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
|
|
found by examining the call sites of OString::getToken
which means I needed to add a new o3tl::equalsAscii function
Change-Id: I7dc0ea1cf5ce8090a708d44f2cf7c938fa200c5f
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133826
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
|
|
I should have removed this in bfb6a42e26de28d88ced43eb3a2c63ce323.
Change-Id: Ieb4b5fabcc1eecf82dbacef66bef8288b6c51761
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133815
Tested-by: Jenkins
Reviewed-by: Luboš Luňák <l.lunak@collabora.com>
|
|
Change-Id: Id1b908e06ebaa7f6ba5981abff51584623770215
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133814
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
|
|
Change-Id: I5e18242b3a18f0c9ca1891ab4e29bd43aaab1f26
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133811
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
|
|
that is used in the traditional code path, this will fix vcl RTL
scrollbars in otherwise LTR UI
Change-Id: I1451f7e17b93b0339ded6d33147df6415274ebfc
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133780
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
|
|
Change-Id: Id10d68f2eb016671be6842dfaa82909207b0708d
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133754
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
|
|
so if we have a RTL SvTreeListBox in a LTR environ then we get RTL
scrollbars
Change-Id: Ia9aaab81f273e31c61cfe4fd4e138709414ad599
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133778
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
|
|
Change-Id: I51c446c2e1e4b02df34200357b2d4ccafd69c7f6
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133777
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
|
|
A mistake from bfb6a42e26de28d88ced43eb3a2c63ce323d33aa.
Change-Id: Iddf885d41177e6dbbfdc1c916b45843f4d7ba499
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133796
Tested-by: Jenkins
Reviewed-by: Luboš Luňák <l.lunak@collabora.com>
|
|
The document has parts of paragraphs e.g. underlined, which changes
the vcl::Font part of the hash key. So if a paragraph starts that way
but the rest is not underlined, the optimization of laying out
the entire string would be missed. So do the optimization if there
are subsequent calls for adjacent parts of the string (or starting
at the same index and different length).
Change-Id: I3d4a78b0eae42bd3085e96024340e6ed8daa9ad1
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133790
Tested-by: Jenkins
Reviewed-by: Luboš Luňák <l.lunak@collabora.com>
|
|
The SalLayoutGlyphsCache optimization of laying out an entire string
if it looks like a caller will call it repeatedly for parts of the string
conflicts with the SAL_ABORT_ON_NON_APPLICATION_FONT_USE checks
causing PrintFontManager::Substitute() abort if font fallback happens.
Simply disable the optimization, the output should be the same.
Change-Id: I0cce6a0c2a1a2ce93b95df54d18fedbdc779f34f
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133789
Tested-by: Jenkins
Reviewed-by: Luboš Luňák <l.lunak@collabora.com>
|
|
To avoid this bactrace : https://crashreport.libreoffice.org/stats/crash_details/4d1984f3-3352-49fa-8569-ebf6994ed216
Change-Id: I45ae8b56191c546c551ccaf48d5ab27a20b8e0e6
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133662
Tested-by: Jenkins
Reviewed-by: Arnaud Versini <arnaud.versini@libreoffice.org>
|
|
It's enough to store just SalLayoutGlyphs, not whole SalLayout.
Change-Id: I31d0faccf08142326255c527b875ba8128f9bb38
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133764
Tested-by: Jenkins
Reviewed-by: Luboš Luňák <l.lunak@collabora.com>
|
|
If the glyphs have already been used, the AdjustLayout() call in
OutputDevice::ImplLayout might have altered them, since
MultiSalLayout::ImplAdjustMultiLayout() drops glyphs that need
falllback from the base layout. And then then
GenericSalLayout::LayoutText() would not know to call
SetNeedFallback().
Change-Id: I2f79d26c8b861f20d7d52abaa0d917aaeefb37a0
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133758
Tested-by: Jenkins
Reviewed-by: Luboš Luňák <l.lunak@collabora.com>
|
|
Change-Id: I9ca24a135494a49449aec63615b2d945a93e8b27
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133735
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
|
|
This allows to center the icon in the respective entry,
not having whitespace in the bottom
Change-Id: Ib148df6911f020f8d4efca4f6a80a65b7f95945f
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133720
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
|
|
Change-Id: I32ded64d7e9a86a6226dfbc4a3b9262bc7572ad2
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133761
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
|
|
Change-Id: I8975120de36575e1af5297db4f40ba3a39398bb6
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133748
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
|
|
...where a signed and an unsigned value are compared, and the signed value has
just been proven to be non-negative here
Change-Id: I297d0f80b2b3e68ea7362de64c871d647ac07371
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133739
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
|
|
Change-Id: I2e283197024c89681dc123e77c05d593ccec4552
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133738
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
|
|
Change-Id: I042b8dcadbf7581de325c161763fe35aecde5ca2
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133694
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
|
|
It often happens that the entire text will be laid out in parts,
so if the first call is a prefix and another one follows that,
lay out the entire string and then use the glyph subset optimization.
Doing this only for the second segment instead immediately for
the prefix is more efficient, as sometimes there is only the prefix
call and there's no call for the rest of the string. This also
avoids failures in CppunitTest_sw_layoutwriter, as a number of tests
such as testAbi11870 lay out only a prefix and doing a layout
for the entire string leads to font fallback, on which
CppunitTest_sw_layoutwriter aborts in PrintFontManager::Substitute().
Change-Id: I76554868ec7e8a79dd09709a247ad1d839291c06
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133495
Tested-by: Jenkins
Reviewed-by: Luboš Luňák <l.lunak@collabora.com>
|
|
Change-Id: Ib9b588cd10154049a5493c1be16de72955ea6077
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133736
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
|
|
It didn't quite make sense before that maxsize == 0 meant no cleanup,
and now with items optionally being count as having size more than 1
it makes even less sense to limit the size to max_size() of the
containers. This comes from 16a338e173083954a9932a3a4005f17230,
so instead make that special caller pass a large size, which in
practice is the same.
Change-Id: Id875862126200ba889211f6e4079ae5921f27650
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133690
Tested-by: Jenkins
Reviewed-by: Luboš Luňák <l.lunak@collabora.com>
|
|
> vcl/source/helper/strhelper.cxx: In function ‘rtl::OString psp::WhitespaceToSpace(std::string_view)’:
> vcl/source/helper/strhelper.cxx:365:9: error: ‘pBuffer[-1]’ may be used uninitialized [-Werror=maybe-uninitialized]
> 365 | if( *pLeap == ' ' )
> | ^~~~~~
that hits at least with GCC 13 trunk now (and which actually is a false positive
as far as I can tell)
Change-Id: I69b96e4cf68f0ed20ac7d73014c164db792983a6
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133704
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
|
|
'Cop~y' translated into Chinese is '复制(~Y)'
But the mnemonics '(Y)' is useless on macOS.
In English, 'Cop~y' with the '~' removed is the correct result.
But in CJK should remove along with '(' and ')',
not just only for mbMenuBar. Use the method MnemonicGenerator::EraseAllMnemonicChars
already defined in vcl/mnemonic.hxx
Change-Id: Ic37ce566923e3dea47535d187b1b3dba2088e960
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133572
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
|
|
Use spacing between top, bottom, icon and text (the case of sum
of items heights greater than element height is not yet handled).
Draw the background even when there's no text: that allows to
center the icon in the item, and still have proper highlight of
selected item.
Change-Id: I5d5dd14060efd15beb0e69df859ecf1efe8a0287
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133612
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
|
|
export SAL_DISABLE_CAIRO_DIFFERENCE=1 to experiment with this
Change-Id: If0e5fee8c71ae3d2181b1e0b90bf9ecd6536ffc0
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133671
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
|
|
Text layout code already filters out unsuitable characters,
the null character just wasn't included there because it's normally
not expected to be present in text, only something broken like
ofz34898-1.doc causes it.
This basically reverts commit 3d7ca1bd1c4cc9d468ae214ecb497f71bad340f8.
Change-Id: Ic29674d9507340c2a43098a88c0320d4253a0bf8
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133689
Tested-by: Jenkins
Reviewed-by: Luboš Luňák <l.lunak@collabora.com>
|
|
This is not as important as SalLayoutGlyphsCache, as these should
be smaller and less needed, but still, make sure to limit the memory
the cache may use.
Change-Id: I4051331f8c5254cb5723772bac4dd1bceb9a2a41
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133674
Tested-by: Jenkins
Reviewed-by: Luboš Luňák <l.lunak@collabora.com>
|
|
With just limit on the number of cached SalLayoutGlyphs instances
the actual memory used could vary wildly depending on how long
the text is.
Change-Id: Ibcf6918e562e81276d21876c532838996e275bd6
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133673
Tested-by: Jenkins
Reviewed-by: Luboš Luňák <l.lunak@collabora.com>
|
|
Change-Id: Ife277f789270f80a40cd711a9235e41bbb3384ad
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133686
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
|
|
Change-Id: I412961bd59855dcb4204dfb917b30539dac0eac2
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/132756
Tested-by: Jenkins CollaboraOffice <jenkinscollaboraoffice@gmail.com>
Tested-by: Szymon Kłos <szymon.klos@collabora.com>
Reviewed-by: Szymon Kłos <szymon.klos@collabora.com>
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133653
Tested-by: Jenkins
|
|
mostly boilerplate code
jsdialog changes:
- added force parameter to sendAction
- added support for key press/release and command events
- moved ActionDataMap to jsdialog namespace for sharing
formulabar changes:
- added calls to send jsdialog messages with formula
- added cursor moving support - on command event
Change-Id: I714715133901941ba0758655e2d5907a3bae79f2
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133010
Reviewed-by: Mert Tumer <mert.tumer@collabora.com>
Tested-by: Szymon Kłos <szymon.klos@collabora.com>
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133652
Tested-by: Jenkins
Reviewed-by: Szymon Kłos <szymon.klos@collabora.com>
|
|
Change-Id: I4d94b60531505f2d5a493b4c51f4c9110d3a4517
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133633
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
|
|
if (the default) of --with-fonts is enabled then additionally restrict
fontconfig to those "application" fonts so the linux ci can be used to
safeguard that font and glyph fallback isn't in operation during the
layout tests
Change-Id: Ic7fc705827c8177b8367cf41f55cdd8ae002c1f5
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133523
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
|
|
try an alternative fix
git show -w is your friend here
Change-Id: Ie9644f73431243ea0e9c6ef7f5c7ca4dadc7ed9e
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133638
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
|
|
zero size font is odd, but that's as it was already
Change-Id: Ib430f2ff2a53c21dc7d15d11377762ab336a61e0
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133660
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
|
|
found by examining uses of OUString::copy() for likely places
Change-Id: I6ff20e7b273ad6005410b82719183c1122f8c018
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133617
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
|
|
Change-Id: I64c63418ca51a09312083dab2432bcca0ea262a3
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133647
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
|
|
It seems now the cache key contains all the necessary info and
results can be shared even between different OutputDevice instances,
as long as they are "the same".
Change-Id: I367322f469ab6713299f178a711a63065b245f9a
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/132902
Tested-by: Jenkins
Reviewed-by: Luboš Luňák <l.lunak@collabora.com>
|
|
which is more obvious, from the perspective of the caller, and lets us
avoid creating a new String if nothing needs to be stripped
Change-Id: I66a980eaf4aa818251bec49bdb16c2dddb0745e7
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133657
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
|
|
to find places where string_view is pointing into a temporary String
Change-Id: Ib530b36f441e95d83d8f687d40a97516a0806721
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133656
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
|
|
This reverts commit bd1d0967920655ef85a2352259332f99fef6876e.
Change-Id: I4f4c0aa7d779487b9f3b1da46e7b2e1070f2dba3
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133637
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
|
|
- do not block painting
- use welded wrappers to send JSON
- don't send tunneled dialog
Change-Id: I54c3cd02ab63bad4a50a3623a32f13b0c94a3595
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/132757
Reviewed-by: Mert Tumer <mert.tumer@collabora.com>
Tested-by: Szymon Kłos <szymon.klos@collabora.com>
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133651
Tested-by: Jenkins
Reviewed-by: Szymon Kłos <szymon.klos@collabora.com>
|
|
Change-Id: Ib7a9d0250e855d4c80c7ce2d66df02eac0ab2e69
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133618
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
|
|
MNEMONIC_RANGE_2_START and MNEMONIC_RANGE_2_END defined in 'core/include/vcl/mnemonic.hxx' are 'a' and 'z'
This variable c is UpperCase, should be converted to lowercase, and thencompare with MNEMONIC_RANGE_2_START and MNEMONIC_RANGE_2_END.
Change-Id: I721e293e88bf1b8260b7ea2dc39ae6a7fd4f3d79
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133587
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
|
|
Change-Id: I4462f7cf4740fa4d1b129d76a0775f4250f41bbd
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133555
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
|