Age | Commit message (Collapse) | Author |
|
...now also covering variables with internal linkage that don't need a redundant
"static". (Unlike with functions, with variables there are also cases that are
not in an unnamed namespace, hence the rename of the plugin.)
All the relevant changes across the code base have been done in the preceding
"Upcoming improved loplugin:staticanonymous -> redundantstatic" commits.
Ideally the changes would have been done with a rewriting plugin, but it can be
quite tedious in general to identify the correct occurrence of "static" that
must be removed, consider e.g.
struct { int init() { static int n; return n++; } int x = init(); } static
const a[10] = {};
However, it turned out that in all cases across the code base, the relevant
"static" was either at the start of the declaration or came after an initial
"const". So I temporarily changed the plugin with
> --- a/compilerplugins/clang/redundantstatic.cxx
> +++ b/compilerplugins/clang/redundantstatic.cxx
> @@ -59,7 +59,7 @@ class RedundantStatic
> }
> report(
> DiagnosticsEngine::Warning, "redundant 'static' keyword in unnamed namespace",
> - decl->getLocation())
> + decl->getBeginLoc())
> << decl->getSourceRange();
> return true;
> }
> @@ -73,7 +73,7 @@ class RedundantStatic
> DiagnosticsEngine::Warning,
> "non-inline variable of non-volatile const-qualified type is redundantly marked as"
> " 'static'",
> - decl->getLocation())
> + decl->getBeginLoc())
> << decl->getSourceRange();
> return true;
> }
to report the diagnostics at the start of the declarations (instead of at a more
natural place which is typically somewhere in the middle of the declaration),
compiled LO from within Emacs and then ran a function
> (defun doit ()
> (interactive)
> (while t
> (next-error)
> (with-current-buffer (window-buffer)
> (when (re-search-forward
> "\\=\\(\\<static\\>\\s *\\|\\(\\<const\\>\\)\\s +\\<static\\>\\)"
> nil t)
> (replace-match "\\2")))))
to do all the replacements. (Plus solenv/clang-format/reformat-formatted-files
where necessary.)
Change-Id: Ie7efc8e0593a407c390a6a7a08c81e547410f18a
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/97779
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
|
|
Change-Id: Ic1454d63a17bd3ec1d70cd5eef5895aee2d919ec
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/97642
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
|
|
...so drop the former. But keep the relevant externvar tests by moving them
into compilerplugins/clang/test/external.cxx. (Which revealed one difference
between the two plugins, regarding certain extern "C" variables in unnamed
namespaces, where Clang (and for that matter also e.g. GCC, it appears)
deliberately deviates from the Standard and considers them to have external
linkage. Add clarifying comments that loplugin:external keeps considering these
as having internal linkage, following the Standard.)
Change-Id: I344fcd0135fdaf6bf08a4b396af2ed2299389a7d
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/97639
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
|
|
Change-Id: I8f21c12f7ee10e1b9ba883a8ff01bb5252429f09
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/97353
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
|
|
Change-Id: I6490a4d6151ae163053f12c872477f2bbfcb3d95
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/97630
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
|
|
...introduced with a7f2239e649ea1b9ef611d758841584ab51abed0 "external: bundle
box2d"
Change-Id: Ia1a46dc517073d3cf7f221d3f7275741e94995a1
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/97629
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
|
|
...instead of via the XJob hack introduced in
a6862a26d6cd17f6b4e4f6577bcd778bf952e65b "tdf#134106: Get rid of
XDesktopInternal again"
Change-Id: Ic533e8faba1a3da65ead74d78f0ab9c4aab1ca6a
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/97397
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
|
|
(lexicographically, by Unicode code point values; to make it easier to make
further modifications)
Change-Id: I867889b37d7395c510dc8cd1310454a4f4ac9982
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/97362
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
|
|
Don't use $host_os="darwin" for both macOS and iOS depending on
$host_cpu. Soon macOS will run on either x86_64 or arm64. Instead, use "darwin"
(or "macos") for macOS and "ios" for iOS.
Some other early changes for arm64-apple-macos, too.
Change-Id: Id89987d854ceba2cd87c6222db2081ccdec0c73e
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/96976
Tested-by: Jenkins
Reviewed-by: Tor Lillqvist <tml@collabora.com>
|
|
Change-Id: I6fa692bb3e4a16400ee2ae847a1f97201493f53a
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/97287
Tested-by: Jenkins
Reviewed-by: Tomaž Vajngerl <quikee@gmail.com>
|
|
Change-Id: I58ef867586bdd2c8c9f4eb249f10610860c9568a
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/97285
Tested-by: Jenkins
Reviewed-by: Tomaž Vajngerl <quikee@gmail.com>
|
|
...by addressing the follow-up TODO mentioned in the commit message of
7a3736f908c0ae207567603c61ce0f617339bac0 "New loplugin:elidestringvar"
(extending it not only to uses with a constant sal_Unicode, but also to uses
with OUStringLiteral).
(All necessary changes have been made in preceding "Upcoming improved
loplugin:elidestringvar" commits.)
Change-Id: Ib0000ef9c4a1dad52124dfd039dd936cf7e3ba3f
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/97226
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
|
|
Change-Id: I1696105ead648c86076f05f50f6286e9a3a932d4
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/97218
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
|
|
Change-Id: I6938049072f458f45d47a1a5ae8901767d95e83b
|
|
Change-Id: I007edc0e339bd9cfd8caf95ecc954628d450664f
|
|
Change-Id: I6266dedb17cd7c3b730fc69804695536fef37cfc
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/96751
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
|
|
Change-Id: I22bb548f7f0c1ef0de157f99915fbaf6473c284b
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/96861
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
|
|
sharing a single widget between multiple notebook pages isn't
going to work with native notebooks, so replace with a row
of toggle buttons
Change-Id: Ic24632f6d94d9238423f8b5e61d5e945c98e4a3b
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/96891
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
|
|
Change-Id: I78241701de0756f5fe9bc9b793dadccd4ae119d7
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/96750
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
|
|
Change-Id: Ia990b30d66f8fb71b4bb49c68e0855de643d1a34
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/96818
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
|
|
...for consistency with other options like --enable-assert-always-abort
Change-Id: I3734485e03203aaee065e1435dfd33f980b1b5eb
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/96736
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
|
|
Change-Id: Ib9723c9793244069407ceaa4935a11da08db3795
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/95998
Tested-by: Jenkins
Reviewed-by: Szymon Kłos <szymon.klos@collabora.com>
|
|
Change-Id: I60e61b7444d90a8fe7e99154a3cf3a526fb87f09
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/96518
Tested-by: Michael Stahl <michael.stahl@cib.de>
Reviewed-by: Michael Stahl <michael.stahl@cib.de>
|
|
Fixes CVE-2018-3081 CVE-2020-2574 CVE-2020-2752 CVE-2020-2922 CVE-2020-13249
Remove obsolete patches:
* mariadb-msvc.patch.1
* mariadb-swap.patch
* mariadb-inline.patch.1
* mariadb-CONC-104.patch.1
Don't build anything from plugins/ in the hope that it's not needed.
Change-Id: I1c8633866b7108a8bb22dae0e0dd5f4a44bf5150
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/96466
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
|
|
Change-Id: Ie8fa026becb1899e466fb0e7dbb987290788aaf2
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/96207
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
|
|
rework the "menu" to be a treeview using hover selection instead of
a custom set of widgetry, and drop the newly unused custom a11y code
Change-Id: Ie7d9b7875ce00843b3f262882816cebb472bf681
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/95223
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
|
|
192721b6c7698d1db2d9d404c6f30b2eb9224796 removed the factory
Change-Id: I4397d656ff9e6fa56539d949f7609061309edd7d
|
|
Change-Id: I175ff33704bb1f709a59d53d3e11aa54f45d9b8b
|
|
This reverts commit 756c027198adc5212d46d72e0628c90698652af5.
Reason for revert: It never did work (was a test), and it cannot work because cores are all stored in /cores under mac, which means we cannot tell which cores belongs to which unit tests, or even if it belongs to the current build on jenkins
Change-Id: I866173a41077dbb40d61fb81953d7a92f7d0789e
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/95888
Tested-by: Noel Grandin <noel.grandin@collabora.co.uk>
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
|
|
See tdf#74608 for motivation.
Change-Id: Ibb3bc2afba00ff962e3ed4f14a5e5a3c735a1a29
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/95963
Tested-by: Jenkins
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
|
|
Change-Id: I695d3c9c520def62e666a9b9f295a9e1235f4e83
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/95620
Tested-by: Caolán McNamara <caolanm@redhat.com>
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
|
|
See tdf#74608 for motivation.
Change-Id: I24f64c9ebc2c86a574f3a76f1e739f61bc458dc7
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/95536
Tested-by: Jenkins
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
|
|
...to avoid confusion, when the log reads e.g.
[...]
> It looks like /home/tdf/lode/jenkins/workspace/lo_tb_master_linux_dbg/instdir/program/soffice.bin generated /home/tdf/lode/jenkins/workspace/lo_tb_master_linux_dbg/workdir/UITest/calc_demo/done.core/core.10572
[...]
> Core was generated by `/home/tdf/lode/jenkins/workspace/lo_tb_master_linux_dbg/instdir/program/python.'.
[...]
(<https://ci.libreoffice.org/job/lo_tb_master_linux_dbg/29831/>).
Change-Id: I59f544c6739329620064315c2a54c21e29865b68
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/95480
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
|
|
Change-Id: I4f07e1e7b35bccc217d78662a45ac5eaf5865ee7
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/95394
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
|
|
The original design was planned in tdf#38246 and documented as:
https://wiki.documentfoundation.org/Design/Whiteboards/Comments_Ruler_Control
This patch includes the following changes:
* use the correct highlight colors
* use a down-pointing arrow for expanse with a stable position,
like some tree view implementations
* left-align the label in RTL for better us- / read-ability
* dynamic sized arrow for HiDPI:
* draw arrow using a polygon
* calculate arrow size and padding based on font size
* drop almost all constant values
* fix the instable, mouse-over tooltip -> just show it always
While at it, I found clang-format just added some minimal,
additional changes, so I removed the file from the blacklist
and included these.
Change-Id: I6e0bade387639eae6bdaaf7960b2fe44c73adf65
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/75421
Tested-by: Jenkins
Reviewed-by: Heiko Tietze <heiko.tietze@documentfoundation.org>
Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de>
|
|
Fixes in various languages, also word transliteration
support in Old Hungarian (renamed to hu_Hung.sor).
Change-Id: I2eed60142273b8c353304a5b714f4e4bde5af1b7
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/95302
Tested-by: László Németh <nemeth@numbertext.org>
Reviewed-by: László Németh <nemeth@numbertext.org>
|
|
another small chunk of libs shared by calc/write/impress,
doing this in stages in case anything goes wrong.
Change-Id: Ie0628e9abd6a78eb3882dcd2030cb5dbcc5490f3
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/95098
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
|
|
See tdf#74608 for motivation.
Change-Id: I068b45302d200ccbde082a748a110324bbbbbaec
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/95065
Tested-by: Jenkins
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
|
|
Change-Id: Id3dcceeddb35e3712d52ddf940eab82dab6a6da9
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/94746
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
|
|
Add a java module named "org.libreoffice.unoloader" for this jar.
Require this module from org.libreoffice.uno
(libreoffice.jar has unoloader.jar in its classpath,
so add the same dependency to the modules)
Change-Id: I7471d65ac7a0d2c6a11c002027a21f0c441dd1f2
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/94419
Tested-by: Jenkins
Reviewed-by: Christian Lohmaier <lohmaier+LibreOffice@googlemail.com>
|
|
Add a java module named "org.libreoffice.uno" for this jar.
This needs to be compiled with Java 9.
But since we want to keep b/c with Java 8,
the rest of the jar has to be compiled with Java 8.
This bumps the *build* requirement to Java 9 while keeping
the *runtime* requirement at Java 8.
The gbuild JavaClassSet class is changed to invoke javac twice,
where the 2nd invocation compiles with --release 9 and a --patch-modules
argument so that it finds the results of the first invocation and also
the javamaker generated files in CustomTargets.
Change-Id: I888f5dbe097cc37136e68db5919939877c981862
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/91105
Tested-by: Jenkins
Reviewed-by: Christian Lohmaier <lohmaier+LibreOffice@googlemail.com>
|
|
This reverts commit bd39ec554f90badccfb26c843fd0d43237f1b622.
Reason for revert:
For -ffunction-sections and -fdata-sections, <https://gcc.gnu.org/onlinedocs/gcc-10.1.0/gcc/Optimize-Options.html#Optimize-Options> warns: "Only use these options when there are significant benefits from doing so. When you specify these options, the assembler and linker create larger object and executable files and are also slower. These options affect code generation. They prevent optimizations by the compiler and assembler using relative locations inside a translation unit since the locations are unknown until link time. An example of such an optimization is relaxing calls to short call instructions."
Change-Id: I84225ca94de9afff23bdd2913c77095d5d26f7ee
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/94950
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
|
|
strips out 3+ megabytes of duplicate functions (which originate in
the generated UNO header files)
Change-Id: I69ba5f478b2b062e2b05516bbaa2d18fa8312aa6
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/94868
Tested-by: Noel Grandin <noel.grandin@collabora.co.uk>
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
|
|
...after ecbaf980625a9e7b06abe91c7c70e78f6ad469a7 "-I$(dir $(3)) in
gb_CObject__command_pattern is no longer needed" and
f218f8f6c8c587d2d78679f935093329f145fede "-I$(dir $(3)) in
gb_CObject__command_pattern is no longer needed" had already dropped it from
plain gb_CObject__command_pattern. (0349c738da5970d9f0fc10d7cf4d7b766ce10e13
"support for compiler rewriters" had created gb_*Object__tool_command at a time
when plain gb_*Object__command still contained those -I arguments, so had
presumably just copied them over without a specific additional need to have them
in the tool_command variants; and ecbaf980625a9e7b06abe91c7c70e78f6ad469a7 and
f218f8f6c8c587d2d78679f935093329f145fede then just failed to also remove them
from the tool_command variants.)
Presence of the -I arguments in the tool_command variants caused e.g. `make
COMPILER_PLUGIN_TOOL=fakebool` to fail with
> sw/source/filter/ww8/ww8par.cxx:97:10: error: replace <...> include form with "..." for inclusion of a source file next to the current source file, sw/source/filter/ww8/../../core/inc/DocumentRedlineManager.hxx [loplugin:includeform]
> #include <../../core/inc/DocumentRedlineManager.hxx>
> ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Change-Id: I7ef513507375dcee5d88db53cef58433f7f68e36
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/94867
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
|
|
The option, available since Clang10, emits debuginfo for classes
with explicit ctors only together with the ctor, which avoids
redundant debuginfo for such classes, noticeably saving both build
time and disk space. https://reviews.llvm.org/D72427
I've been using this for quite some time, without problems.
Change-Id: I24e5339f983cb8d8990185436dd63e00f0d3ce60
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/94756
Tested-by: Jenkins
Reviewed-by: Luboš Luňák <l.lunak@collabora.com>
|
|
This reverts LO 6.3.4 commit 5d1709a7c4184eb31cfc4c2d3acadff3a4a68189,
which tdf#133334 shows is wrong. How this made it past QA
is a mystery to me. There should be lots of examples.
Change-Id: I17be6e4bab44057f4535d4728825e12d068b65d2
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/94782
Tested-by: Jenkins
Reviewed-by: Justin Luth <justin_luth@sil.org>
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
|
|
To permit pluggable crypto services, abstract existing
implementation behind an XPackageEncryption API.
Previous code already had two halfway-polymorphic classes (agile and
standard 2007 engine), so we're not adding much additional layers.
As MS crypto always uses OLE storage to wrap content into one single
file, current implementation passes all substorage names down into
XPackageEncryption APi, so different downstream implementations (e.g.
for MS RMS, or Azure AIP) are possible.
Because OleStorage classes are internal to LibO core, access is provided
via XInput/XOutput stream API function.
Change-Id: Icc32a4e0ce215090c3b739f1dcaa0654b36b7f08
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/84436
Tested-by: Jenkins
Reviewed-by: Thorsten Behrens <Thorsten.Behrens@CIB.de>
|
|
Files which could become clang-format conformant with
under 5-percent lines of change relative to the total
count of lines in the file are found by using bin/find-clang-format.py,
and fixed with /opt/lo/bin/clang-format -i <path-of-the-file>
There will be follow-up patches to fix all 'under-5-percent' files.
Change-Id: I2d37d9c94116d2823f9a21024bdb7ff1a6207d5d
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/94744
Tested-by: Jenkins
Reviewed-by: Muhammet Kara <muhammet.kara@collabora.com>
|
|
The --enable-canonical-installation-tree-structure option is gone. We
always use a "canonical" app bundle structure (with MacOS, Frameworks,
Library, and Resources) on macOS since
6492c8576e0393f7ee548cd938e84c80e8d37127 in 2014.
Change-Id: I6cbd3c5141aba643bc05bfa7d4de1e7d73a14112
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/94761
Tested-by: Jenkins
Reviewed-by: Tor Lillqvist <tml@collabora.com>
|
|
Change-Id: I48afc12427e25424b275c1f0d0484ec300948b3d
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/94743
Tested-by: Jenkins
Reviewed-by: Muhammet Kara <muhammet.kara@collabora.com>
|