android: Fix type of a C string to be const char[].

With the wrong type, we were measusing the sizeof() wrongly, leading to
a hard to catch crash at start that appeared only from time to time.

Improve the concatenation too when at that.

Change-Id: I4a4ab2909124281aac99118d66c62d669294d5f7
Reviewed-on: https://gerrit.libreoffice.org/74375
Reviewed-by: Michael Meeks <michael.meeks@collabora.com>
Tested-by: Michael Meeks <michael.meeks@collabora.com>

sfx2 character map dialog: avoid double insert of glyph

The problem was that both the special character widget in
SvxCharView::InsertCharToDoc() and the special character dialog in
SvxCharacterMap::insertCharToDoc() reacted to KEY_RETURN, resulting in a
double insertion.

master has this resolved in commit
374599f8c26713905a310673d2b429083321186a (weld SvxCharacterMap dialog,
2018-03-22) already, do the minimal fix on this branch by reacting to
KEY_RETURN only in the dialog.

This is consistent with how the non-recent/favorite characters are
handled, where also the dialog does the insertion, not the widget.

Leaving the KEY_SPACE part unchanged, since that doesn't close the
dialog, so no double insertion happens in that case.

Change-Id: I9f75eb1d1b46cbbfe8dd4e160d245b5744acc8e0
Reviewed-on: https://gerrit.libreoffice.org/74360
Reviewed-by: Michael Meeks <michael.meeks@collabora.com>
Tested-by: Michael Meeks <michael.meeks@collabora.com>

[cp] android: Default to colibre.

Not ideal to use an ifdef, but I don't want to affect anything other
than Android; and there is no need to up-stream this either, colibre is
the default in master already.

Change-Id: I3b1a29b69cfda89c8615586a5afc78fd1eca8c3e
Reviewed-on: https://gerrit.libreoffice.org/74266
Reviewed-by: Andras Timar <andras.timar@collabora.com>
Tested-by: Andras Timar <andras.timar@collabora.com>

android: Add the sidebar-related constructors to liblo-native-code.so.

Change-Id: I25acb8965d190fc443d75357352e4d02c9b03cc2
Reviewed-on: https://gerrit.libreoffice.org/74265
Reviewed-by: Andras Timar <andras.timar@collabora.com>
Tested-by: Andras Timar <andras.timar@collabora.com>

android: Define the icon theme in distro-configs and switch to colibre.

No need to reinvent the wheel and hardcode it in configure.ac.

Change-Id: Idb08ea0e5ce228bb0758dbdf023f3aee44da76eb
Reviewed-on: https://gerrit.libreoffice.org/74247
Reviewed-by: Andras Timar <andras.timar@collabora.com>
Tested-by: Andras Timar <andras.timar@collabora.com>

android: Fix non-debug build.

Change-Id: If8bb944417da9a0aa6343d7dc87e7a1931689f1f
Reviewed-on: https://gerrit.libreoffice.org/74147
Reviewed-by: Andras Timar <andras.timar@collabora.com>
Tested-by: Andras Timar <andras.timar@collabora.com>

tdf#124907 vcl: react to pan gesture for listbox and other widgets

Reviewed-on: https://gerrit.libreoffice.org/74087
Tested-by: Jenkins
Reviewed-by: Tomaž Vajngerl <quikee@gmail.com>
(cherry picked from commit 08f485c6e92bd56c8c26171316b842cd213a8a09)

Change-Id: I4579f8edd2bf24506b1d55a7291cd86e276f30a4
Reviewed-on: https://gerrit.libreoffice.org/74164
Reviewed-by: Andras Timar <andras.timar@collabora.com>
Tested-by: Andras Timar <andras.timar@collabora.com>

lok: send message when in place editing

Change-Id: I96d22cabeda1eb851116d556a5302394a0dd7f93
Reviewed-on: https://gerrit.libreoffice.org/73162
Reviewed-by: Andras Timar <andras.timar@collabora.com>
Tested-by: Andras Timar <andras.timar@collabora.com>

lok: remove ext text event re-entrancy hazard.

Change-Id: I7566c158330bab77589d422c61c64210727ab835
Reviewed-on: https://gerrit.libreoffice.org/73625
Tested-by: Jenkins
Reviewed-by: Michael Meeks <michael.meeks@collabora.com>
Reviewed-on: https://gerrit.libreoffice.org/73756
Reviewed-by: Andras Timar <andras.timar@collabora.com>
Tested-by: Andras Timar <andras.timar@collabora.com>

Introduce --enable-android-editing

...to select the experimental ...Editing... Android build variant.  (Ignored
for non-Android builds, but using libo_FUZZ_ARG_ENABLE anyway, just in case.)

Change-Id: I670925ff358039e38edc29db69f48a78d484f133
Reviewed-on: https://gerrit.libreoffice.org/65077
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
Reviewed-on: https://gerrit.libreoffice.org/74148
Reviewed-by: Jan Holesovsky <kendy@collabora.com>
Tested-by: Jan Holesovsky <kendy@collabora.com>

Compile also the palette_neon_intrinsics file for ARM

Change-Id: I26deeefc8cf335f53aa55c8523cd08687f27b63b

android: Unregistering in runLoop() is important even on Android.

Otherwise we get crashes in online Kit when attempting to shutdown
reasonably cleanly and the poll kicks in.

Change-Id: I9a1744b42fec9f8ffddc6094773bda943caedf2a
Reviewed-on: https://gerrit.libreoffice.org/74056
Reviewed-by: Jan Holesovsky <kendy@collabora.com>
Tested-by: Jan Holesovsky <kendy@collabora.com>

Add some more SAL_INFOs

Change-Id: I5b50df5ea6b99164d150f99ae4fd36cddde22a60

explictly exclude LibreLogo from XScript usage

Change-Id: I567647f0e2f8b82e4ef2995c673abe82f4564228
Reviewed-on: https://gerrit.libreoffice.org/73659
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>

sanitize LibreLogo calls

Change-Id: Ie4d9858e5b4b3e55ab08416fb9338d2df34ee5e1
Reviewed-on: https://gerrit.libreoffice.org/73627
Tested-by: Jenkins
Reviewed-by: László Németh <nemeth@numbertext.org>

It is not a good idea to allocate a UNO object statically

I should have known that. But sometimes you subconsciously just want
things to be easy for change. But no. This fixes a problem that showed
up in CppunitTest_services.

Change-Id: Iaefcbe4a07716f4d1b2c269627a99d43e3738af3

Veto process exit while an OLE client is connected

Change-Id: Iad9fc1742ae371a8a162edbc16998e9cb6895919

Fix thinko in sw unit test.

Change-Id: I6c72ce5d416151e7864e87e3b9d801a218525a16

lok: re-factor getTextSelection.

Change-Id: I2c27c213ee980e19d6020e9599b2b72115e7f28e

tdf#124947 fix hit area of the radio buttons, size in ios theme

The hit area of the radio buttons was not calculated correctly, so
sometimes when you hit the radio button at the "wrong" place,
nothing happened. This fixes the hit area to correctly cover the
radio button and the text.

Another issue was that the ios theme size of the radio button was
not defined correctly (32px instead of 26px), which increased the
error when calculating the hit are even more. The height of the
radio button should be the same as defined in the definition.xml,
and not bigger (or smaller).

Change-Id: I4b03f36ca9d9c82bd6dc442bd6e06af938c62bdd
Reviewed-on: https://gerrit.libreoffice.org/73592
Tested-by: Jenkins
Reviewed-by: Tomaž Vajngerl <quikee@gmail.com>
Reviewed-on: https://gerrit.libreoffice.org/73593
Tested-by: Tor Lillqvist <tml@collabora.com>
Reviewed-by: Tor Lillqvist <tml@collabora.com>

Adapt gcc3_linux_arm __cxa_exception to NDK 18 libc++abi

...which adds a member, and will be relevant when switching armeabi-v7a to
libc++/libc++abi/libunwind (coming soon).

Change-Id: I1f98f50ca299a583e73e0dbd6c8c94d973329f3a
Reviewed-on: https://gerrit.libreoffice.org/64966
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>

adjust path to generated apk to new defaults

current android-gradle plugin creates the apk in dimension/flavor/variant
(whatever you want to call it) specific directory

Change-Id: I5514fe53c5c8c556d194a0f06ad6167b0b1cc98e

android: bump gradle, gradle-plugin, support libary & target versions

Change-Id: I971f02c624a19ec58539738f5e736a986b5ba0d6

icu: don't skip building the data just because tools were disabled

building for e.g. android disables the tests and requests data in a
static library, but icu completely skips building the data directory in
case --disable-tools was specified:
icu/source/Makefile.in
    @TOOLS_TRUE@DATASUBDIR = data
will become
    #DATASUBDIR = data
and then
    SUBDIRS =  stubdata common i18n $(LAYOUTEX) $(ICUIO) $(TOOLS) $(DATASUBDIR) $(EXTRA) $(SAMPLE) $(TEST)
will not have the data dir and make will ignore it.  Add it back by
specifying it when invoking make and all is fine.

Change-Id: I0af693f22938ebabdc189a97f1cfc3f8b1c042ee
Reviewed-on: https://gerrit.libreoffice.org/45107
Reviewed-by: Christian Lohmaier <lohmaier+LibreOffice@googlemail.com>
Tested-by: Jenkins <ci@libreoffice.org>

oss-fuzz build failure

Change-Id: I27e822ac2432eb2746ed9181cee158012e54252e

fix oss-fuzz build

Change-Id: Ia32a80547cd3e3fd21791d8ef2cf717700c790a8

javaunohelper: blind fix for Android build broken by 2d22f8bb15c1

Change-Id: Ie56197b3f4a90b8f9510a1c29d0388f5f70e1d16
Reviewed-on: https://gerrit.libreoffice.org/72408
Reviewed-by: Michael Stahl <Michael.Stahl@cib.de>
Tested-by: Michael Stahl <Michael.Stahl@cib.de>

xmlsecurity: fix the Android build

Change-Id: Ib04060ee4dc2a4ece2858548690e8ef182bf00b1

Fix Android build after commit ac419786b3244d909901db053841862abf5e7a2f

... giving this error:

   In file included from
   /home/android/lo/master-android-arm/filter/source/xmlfilteradaptor/XmlFilterAdaptor.cxx:50:
   /home/android/lo/master-android-arm/include/comphelper/scopeguard.hxx:36:14:
   error: function 'comphelper::<deduction guide for ScopeGuard><(lambda at
   /home/android/lo/master-android-arm/filter/source/xmlfilteradaptor/XmlFilterAdaptor.cxx:146:34)>'
   has internal linkage but is not defined [-Werror,-Wundefined-internal]
       explicit ScopeGuard( Func && func ) : m_func( std::move(func) ) {}
                ^
   /home/android/lo/master-android-arm/filter/source/xmlfilteradaptor/XmlFilterAdaptor.cxx:146:28:
   note: used here
       comphelper::ScopeGuard guard([&]() {
                              ^

See https://lists.freedesktop.org/archives/libreoffice/2019-April/082599.html

Change-Id: If719d183d2ce15fa4877cd8bbf457d79097765be
Reviewed-on: https://gerrit.libreoffice.org/71510
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
Tested-by: Mike Kaganski <mike.kaganski@collabora.com>

update references about selective debuginfo in android/README

The second case is very likely mistaken though.

Change-Id: I6bfc55800a6b26ce0ae6d3645e4179dc5f683205

android: Make the tiled rendering from the loolwsd actually work.

Not to break the 'old' Android app, introduce a bool that can indicate
if we are using the LOK from the 'old' (LOK-via-JNI-based) or from the
'new' (loolwsd-based) app.

Change-Id: I38bd665cc1d5bc88018574171443ecabc46763df
Reviewed-on: https://gerrit.libreoffice.org/70678
Reviewed-by: Jan Holesovsky <kendy@collabora.com>
Tested-by: Jan Holesovsky <kendy@collabora.com>

graphite: Force optimization on Android.

The Pass.cpp triggers an infinite loop in clang when not optimizing (ie.
with the debugging buildls), at least with this version:

  > ~/Android/Sdk/ndk-bundle/toolchains/llvm/prebuilt/linux-x86_64/bin/clang++ --version
  Android (5058415 based on r339409) clang version 8.0.2 (https://android.googlesource.com/toolchain/clang 40173bab62ec746213857d083c0e8b0abb568790) (https://android.googlesource.com/toolchain/llvm 7a6618d69e7e8111e1d49dc9e7813767c5ca756a) (based on LLVM 8.0.2svn)
  Target: x86_64-unknown-linux-gnu
  Thread model: posix
  InstalledDir: /local/home/kendy/Android/Sdk/ndk-bundle/toolchains/llvm/prebuilt/linux-x86_64/bin

Change-Id: Ib4984f8838fc733be6da4e24eb10915d3efbf79e

android: fix ArrayIndexOutOfBounds on selecting 'Drawings' in DefaultFileFilter setting

Change-Id: I5ed924d9d79db169b11a0561b89da9b57eb078d9
Reviewed-on: https://gerrit.libreoffice.org/66971
Reviewed-by: Jan Holesovsky <kendy@collabora.com>
Tested-by: Jan Holesovsky <kendy@collabora.com>

sal, android: fix -Werror,-Wreturn-type

Change-Id: Ic37967dedb2378106db2f3f437781c8ad20cad56

android: Set the program path sensibly when we've got nullptr.

Change-Id: I4415459ca5d758d06d866faeca85e1cf154d0567
Reviewed-on: https://gerrit.libreoffice.org/67877
Tested-by: Jenkins
Reviewed-by: Jan Holesovsky <kendy@collabora.com>

android: Set the FONTCONFIG_FILE envvar to the fonts.conf (if exists).

Change-Id: Ic9fd97a2ff8a6d96ffcc7ad300ef30201d786528
Reviewed-on: https://gerrit.libreoffice.org/67876
Tested-by: Jenkins
Reviewed-by: Jan Holesovsky <kendy@collabora.com>

android: Separate the Android-specific setup from the actual LOK init.

Change-Id: I433376dfea0a43c63827ba15308a614f3466fb71
Reviewed-on: https://gerrit.libreoffice.org/67875
Tested-by: Jenkins
Reviewed-by: Jan Holesovsky <kendy@collabora.com>

android: Expose setting of the JavaVM.

Needed when we use the liblo-native-code as a library.

Change-Id: Ia4d1d72f2203a96d1196d015b521201919bbb165
Reviewed-on: https://gerrit.libreoffice.org/67965
Tested-by: Jenkins
Reviewed-by: Jan Holesovsky <kendy@collabora.com>

android: Generate liblo-native-code.so so that it can be linked against.

And used for LibreOfficeKit directly, without a JNI wrapper.

Change-Id: If4231f53b579417b6c9c90b39541193bf2e8f27c
Reviewed-on: https://gerrit.libreoffice.org/67874
Reviewed-by: Jan Holesovsky <kendy@collabora.com>
Tested-by: Jan Holesovsky <kendy@collabora.com>

Export RTTI symbols from liblo-native-code.so, for binary UNO bridge

This will become important when switching armeabi-v7a to
libc++/libc++abi/libunwind (coming soon) which uses address instead of string
comparison when checking for type equality, so that exceptions thrown from the
binary UNO bridge will need to use the exact same RTTI objects as referenced
from the compiled catch clauses.

Change-Id: If8bcb39212b5f5e154aee215cb5f471fe2dc4a7b
Reviewed-on: https://gerrit.libreoffice.org/64965
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>

The 2nd parameter to lo_initialize() is a URL so use such naming consistently

Change-Id: Id4e7f8556d6910ab43b0f7b8ba09e7be8b1ec1d1

Switch Android armeabi-v7a to libc++/libc++abi/libunwind too

It had been left out in 4082a18406c18af7b4fcef7bd501c3679c3be56b "android: use
unified headers and llvm-c++ STL (x86) with NDK 16" because "arm unfortunately
crashes with llvm-c++, so keep with gnustl for now/fix that later".

Making armeabi-v7a work with libc++ etc. required a number of changes, listed
below, in this commit and in preceding ones.  At least 32-bit x86 already worked
with libc++ etc. prior to these changes in view mode, though it crashed in the
experimental editing mode (enabled with strippedUIEditing in
android/soruce/Makefile) as soon as one types in something,  But it is not
entirely clear to me why 32-bit x86 view mode didn't also fail similar to how I
saw armeabi-v7a fail.  (On 32-bit x86, these changes appear to neither improve
nor worsen the current state, view mode still appears to work fine while editing
still crashes upon typing anything.  With these changes, editing mode on
armeabi-v7a appears to work fine.  But I tested armeabi-v7a only with a real
device and 32-bit x86 only with an emulator, in case that might make a
difference.)

* Preceding <https://gerrit.libreoffice.org/#/c/64964/> "Move NSSLIBS to a more
  sensible place on the linker command line" plus this change's addition of
  -lunwind to the liblo-native-code.so linker command line make sure that
  liblo-native-code.so uses _Unwind_* functions from libunwind.a, instead of
  erroneously picking up the ones from libgcc.a that happen to be included in
  NSSLIB's nspr4 (-lgcc is automatically added to the end of the linker command
  line by the invoking compiler, that's how libgcc.a's _Unwind_* end up in
  NSSLIB's nspr4; it is neither clear to me why NSSLIB's nspr4, being a pure C
  library, uses _Unwind_* functions, nor why exception handling in
  liblo-native-code.so fails when using _Unwind_* functions from libgcc.a
  instead of from libunwind on armeabi-v7a, nor why that would work on 32-bit
  x86, but that's what I observed: ModuleManager::identify
  (framework/source/services/modulemanager.cxx) throws a
  css::lang::IllegalArgumentException, which calls __cxa_throw ->
  _Unwind_RaiseException, which ultimately lead to odd misbehavior and
  std::abort during stack unwinding when using _Unwind_RaiseException from
  libgcc.a instead of from libunwind).  (There is no libunwind.* in
  android-ndk-r16b for 32-bit x86 at least, so is presumably using _Unwind_*
  functions from libgcc.a.  It doesn't appear to make a difference if it
  indirectly uses those _Unwind_* functions from NSSLIB's nspr4, or directly
  from libgcc.a included in liblo-native-code.so if the

    $(if $(filter armeabi-v7a,$(ANDROID_APP_ABI)),-lunwind)

  had a ",-lgcc" else branch.)

* Preceding <https://gerrit.libreoffice.org/#/c/64965/> "Export RTTI symbols
  from liblo-native-code.so, for binary UNO bridge" makes sure that excpetions
  thrown from the binary UNO bridge can be caught by compiled catch clauses.
  Not sure why the corresponding state of
  bridges/source/cpp_uno/gcc3_linux_intel shouldn't have run into the same
  issue.

* Preceding <https://gerrit.libreoffice.org/#/c/64966/> "Adapt gcc3_linux_arm
  __cxa_exception to NDK 18 libc++abi" makes sure that our version of
  __cxa_exception matches the version from libc++abi.  This is clearly not
  relevant for 32-bit x86.  (The comment there android-ndk-r18b, but the
  additional member is already present in
  android-ndk-r16b/sources/cxx-stl/llvm-libc++abi/src/cxa_exception.hpp, too.)

The remainder of this change just drops old armeabi-v7a--specific workarounds
that are no longer needed/no longer work.

Change-Id: Ief4c2d562c5032abe6c3b94ca3b3394be6fcd4d3
Reviewed-on: https://gerrit.libreoffice.org/64973
Tested-by: Stephan Bergmann <sbergman@redhat.com>
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>

Move NSSLIBS to a more sensible place on the linker command line

511ae02c6457e69cb6daab871acd9c3e7d64e2e3 "Android: Enable HAVE_FEATURE_NSS and
package the NSS libraries with apk" had added them to WHOLELIBS probably just
because that already had the $(addprefix -l,...), even though --whole-archive
doesn't make any sense for shared libraries.  Better place them later on the
linker command line (after all our own archives and compiler support libraries),
so that switching armeabi-v7a to libc++/libc++abi/libunwind (coming soon) will
be able to override erroneously picking _Unwind_* symbols from NSSLIBS's nspr4
instead of libunwind.

Change-Id: Ie0c0b7a55da3eabe1bb427232d698b2a4af63e78
Reviewed-on: https://gerrit.libreoffice.org/64964
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>

android: use unified headers and llvm-c++ STL (x86) with NDK 16

gnustl (and others) are to be removed in future versions of the ndk
also bump gradle and build-tools to current versions along with it

arm unfortunately crashes with llvm-c++, so keep with gnustl for now/fix
that later

Change-Id: Ic794c3293b599b77ec48096bf3283a99c09cbb79
Reviewed-on: https://gerrit.libreoffice.org/45163
Reviewed-by: Christian Lohmaier <lohmaier+LibreOffice@googlemail.com>
Tested-by: Christian Lohmaier <lohmaier+LibreOffice@googlemail.com>

android: Add a missing dependency.

Useful when you switch sdk to a different location.

Change-Id: I95301cc9adf50a0bbd918cc7562b8871093625b8
Reviewed-on: https://gerrit.libreoffice.org/67873
Tested-by: Jenkins
Reviewed-by: Jan Holesovsky <kendy@collabora.com>

Don't build dconf on Android and iOS.

Without that modification, it finds dconf in linux system and then compilation fails when it tries to build configmgr/source/dconf.cxx because dconf headers are not found in Android NDK.

Change-Id: I25ab7f1ce66ed491f08a526e462e00957135b0c2
Reviewed-on: https://gerrit.libreoffice.org/65987
Tested-by: Jenkins
Reviewed-by: Michael Stahl <Michael.Stahl@cib.de>

Fix getRTTI for Android x86

First, make sure existing compiler-generated RTTI from liblo-native-code.so is
found (so that catching bridge-synthesized exceptions in native code works with
libc++abi, which checks type equivalence by address instead of string
comparsion), by using the dlsym(RTLD_DEFAULT,...) mechanism as in the ANDROID
gcc3_linux_arm bridge case.  And second, if that should fail, synthesize the
type_info even if the included cxxabi.h doesn't provide the relevant type
declarations, by using copies from the ABI specification, as also done on other
platforms.

Instead of always having getRTTI fail and raiseException throw a non-synthesized
css::uno::RuntimeException("no rtti for type ...").  Which explains the mystery
discussed in the commit message of 312eeeee42cb4a1e356943e17305555e41afc4ef
"Switch Android armeabi-v7a to libc++/libc++abi/libunwind too", why the observed
misbehavior on x86 was so different from that on armeabi-v7a.

Change-Id: I9308654c5c2b88b4d27e0e8e9edda1849133a161
Reviewed-on: https://gerrit.libreoffice.org/65070
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>

Make android/mobile-config.py compatible with Python 3

Change-Id: I079c4efd28e7e0d10ca6edf242ddd85b9294db08
Reviewed-on: https://gerrit.libreoffice.org/64962
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>

android: harfbuzz refuses to compile using mmap.

Not ideal not to use mmap on a small device though, but disable for now.

Change-Id: Id595cbc87c93679c1ce186f91ad36a961aee4132
Reviewed-on: https://gerrit.libreoffice.org/63342
Tested-by: Jenkins
Reviewed-by: Michael Meeks <michael.meeks@collabora.com>

android: remove redundant header causing build failure.

Change-Id: I857224b551ff6e5959ffe31e872ec4e4993bc174
Reviewed-on: https://gerrit.libreoffice.org/63343
Tested-by: Jenkins
Reviewed-by: Michael Meeks <michael.meeks@collabora.com>

tdf#119082 Exception wrong thread on Android Viewer

Change-Id: I58a8d104b24c7cf2d021e2dce700e25592c8a605
Signed-off-by: Mert Tumer <merttumer@outlook.com>
Reviewed-on: https://gerrit.libreoffice.org/58563
Tested-by: Jenkins
Reviewed-by: Tomaž Vajngerl <quikee@gmail.com>

tdf#118585 Fix exception when closing Android Viewer

Change-Id: I9d87168dce4b5c6e503d41366aa35d5dadf89dae
Reviewed-on: https://gerrit.libreoffice.org/57081
Tested-by: Jenkins
Reviewed-by: Tomaž Vajngerl <quikee@gmail.com>

android: simplify boolean statement

Change-Id: I92cdccfe7b97d9baf29b1155990a38b26579e365

android: suppress deprecation warning for Html.toHtml(String)

deprecated since 24/Nougat, but we still want to support devices running
Marshmallow and older..

Change-Id: Iedcf8a56028eff44134b548dc07a89573cb2210d

tdf#117777 fix show hidden files/folders option on Android

Change-Id: Ic80ca7f8ebb93f78a58cc0ad778db90deb0c51ba
Signed-off-by: Mert Tümer <merttumer7@gmail.com>
Reviewed-on: https://gerrit.libreoffice.org/54771
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Thorsten Behrens <Thorsten.Behrens@CIB.de>

android janitor... convert chained ifs to switch statement

Change-Id: I9929bd077a33349a6e57c3a427df7e77b3007c88

pdfium: don't need two copies of freetype on android either

Change-Id: I581575e2b13482d16aca284bfa502ed164b8c9d0

tdf#96795: Use locale-sensitive sorting in Android viewer

Change-Id: I98edec9f57f1bf0576e16a1ba3c0e023d6d0b6ed
Signed-off-by: Mert Tümer <merttumer7@gmail.com>
Reviewed-on: https://gerrit.libreoffice.org/52494
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Thorsten Behrens <Thorsten.Behrens@CIB.de>

add hint on how to use lldb pretty-printers form within android studio

Change-Id: Ia028dd64c04d1de95e42ea2b89ee2bca689a2390

libmspub: fix build for android-arm with debug enabled

Change-Id: I2fff80f2e4a0bd1946c9b49d46871245f93bbb3e

tdf#116152 fix unsupported url exception on Android Viewer

Change-Id: I39233369e754919aeb0dc46856a3746f33e89e9b
Signed-off-by: Mert Tümer <merttumer7@gmail.com>
Reviewed-on: https://gerrit.libreoffice.org/51675
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Thorsten Behrens <Thorsten.Behrens@CIB.de>

-Werror,-Wunused-parameter (Android)

Change-Id: Ib15218f1f3bc2867ca0e8daa50a2d4edad4a9d87

sal android: fix -Werror,-Wimplicit-function-declaration

Change-Id: If19dbb654d473e8785dc69f96775c78cc95a7bd6

sw android: fix -Werror,-Wsign-compare, take two

This time not breaking other platforms. Thanks to Stephan Bergmann for
suggesting this alternative way.

Change-Id: I2535e57de1aafd17960a1d1054fc58a47933bb58
Reviewed-on: https://gerrit.libreoffice.org/49441
Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>
Tested-by: Jenkins <ci@libreoffice.org>

tdf#115388: Android: Use HTTP Secure

Change-Id: Ia2e198c20b904ba178fac2969dfa80f3f1291be1
Reviewed-on: https://gerrit.libreoffice.org/49123
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>

fix android red and blue channel swap on jpegs w/ 24bit RGB surface

that was introduced with 54596087e57ea533253e19eea594d9b6c06e8d26
without adjusting for Android accordingly

Change-Id: Ie6a2001ab13ecec959da9b57bb49e1bac76e8691

android: fix some layout issues with new-document buttons

functionality to create new documents was added for tdf#106325 with
commit 1503769fe15c122ff2bb8f6f7e7b4ab72656ddc2
However the commit had some issues, most notably clipping of the shadow
effect and alignment issues to to scaling down (via the animation
effect, took me quite a while until I looked at that) with having a
centered pivot. I think it was due to those alignment issues that the
label was set to a fixed with - which is a problem for future
translation.
This commit:
* changes the animation to not scale down the initial views
* changes the FABs' setting to use compatPadding - this not only ensures
  enough room for the shadow, but also allows for consistent look across
  android versions. Side effect is that it needs more height, so
* the effect was adjusted accordingly, to shrink to/expand from the
  lower-right corner of the parent view, instead of "mid-air"
* the labels were given a proper background drawable with rounded
  corners and some padding for a (much more9 polished look
* formatting of the labels was split out to a style, for ease of
  tweaking the looks
* add tools:visibility statement to see the elements in layout editor
* fixed the strings for the labels (and ID) to some less awkward ones

Change-Id: Ifd2d90ec8d23e40603edff19c9af9e0b32090c9f
Reviewed-on: https://gerrit.libreoffice.org/42033
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Christian Lohmaier <lohmaier+LibreOffice@googlemail.com>

android: make compat header self-contained

Change-Id: I0362725646521c9e2126faeaaf35a1c1dbf775af

i18npool: try to fix Android build, no std::trunc

Change-Id: Idf609a4012499e76acc5e7ae76565937f2a58fbd

sal: try to fix Android build, no std::round

Change-Id: I58bc7c9099cd2a3a90f92efbd179d0f3abce02a6

tdf#103083 option to show/hide hidden files/folders on Android

The patch was sent for the ULAKBIM/PARDUS project.
Signed-off-by: Mert Tümer <merttumer7@gmail.com>
Change-Id: Iadbb688a1a976279b23d246001818ac1c21b3d47
Reviewed-on: https://gerrit.libreoffice.org/47637
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Thorsten Behrens <Thorsten.Behrens@CIB.de>

update android rgba cairo patch to apply again

Change-Id: Iaee918439f8472bab7fab2cf545ac2e420d614ab

oox: fix Android build

Change-Id: Iad0d51311ba0e74e1fbc33ee3fb09c8c9a89b0c0

oox: try to fix Android build, no std::lround there

Change-Id: I47e1df5d27e788ef5aa48c3a09daf9888c33f920

have "make clean" also remove android's/gradle's output dirs

esp. removing the jniLib one is important, as otherwise it is easy to
package both x86 as well as arm libs into the apk, making it huuuge :-)

Change-Id: I6364b82800a449682b618f0bb98a47825dcb0742

disable SwUiWriterTest2::testTdf122942 on Windows

Change-Id: Ie885224f257916debec85ab4a42435ff606b0acd

[cp] HACK: sleep before sign

On Windows, with 16x parallel build, build-nocheck target fails,
because it tries to sign a file that is already in use (gengal).

Change-Id: Ic582334f2c02f5e38f49d048fc0425025ffeee41

Bump version to 6.0-32

Change-Id: Ic1eeb36584144785f1ec5c502140ea0405020cac

tdf#118375, tdf#70838 correct position of rotated shape in doc

Word relates the position to the unrotated shape in legacy doc
format. Writer uses the rotated shape. The patch corrects the
difference on import and export.

Reviewed-on: https://gerrit.libreoffice.org/70152
Tested-by: Jenkins
Reviewed-by: Regina Henschel <rb.henschel@t-online.de>
(cherry picked from commit 421e6fc3cd2e6fe37afbef341e2d0ad7b8edde37)

Change-Id: I25537123656e62d6ffae5118ee8d621a4b5c5be0

Register spsupp*.dll during installation

This registers SharePoint integration libraries using regsvr.exe.
Both 32-bit and 64-bit libraries are registered; registration of
SharePoint.OpenDocuments is unconditional (subject to ActiveX
feature install state), substituting respective MSO component.

This only works for 32-bit MSI, because in case of 64-bit MSI, only
64-bit SharePoint DLL is included; this makes the library ~useless
in case of 64-bit MSI, because IE needs 32-bit component absent there.
Proper solution will be created for master branch.

Change-Id: Ic5bf9e7090acfaa17d47cb72b9ebfd483242f656
Reviewed-on: https://gerrit.libreoffice.org/72801
Reviewed-by: Aron Budea <aron.budea@collabora.com>
Tested-by: Aron Budea <aron.budea@collabora.com>

tdf#122221: Bin apparently unnecessary early return

Removing these few lines of code fixes the problem on iOS, as far as I
can see. Let's hope it doesn't introduce any new issue in the iOS app.
In this branch, I didn't have the clout to remove them for macOS, too.

Change-Id: I2380d010ba223a698acfe916fca4580a1502be98

Resolves: tdf#125279 do not double _x005F_ escapement

Reading OOXML _x005F_ escaped content may not get unescaped, so
when writing back to OOXML do not attempt to escape it again, i.e.
write _x005F_xHHHH_ as is and not as _x005F_x005F_xHHHH_.

This is more a workaround, the proper fix would be to unescape
_x005F_ content upon read. But then the entire "invalid XML
character" escapement and handling control characters rat tail
would come into play.

Change-Id: I3d31dc84a362753c23a8c89f7a5d7bfd06e4367b
Reviewed-on: https://gerrit.libreoffice.org/73187
Tested-by: Jenkins
Reviewed-by: Eike Rathke <erack@redhat.com>
(cherry picked from commit f677885fec59f252f36673ee4d8c0b4863625a4d)

More loplugin:cstylecast: scripting

auto-rewrite with <https://gerrit.libreoffice.org/#/c/47798/> "Enable
loplugin:cstylecast for some more cases" plus
solenv/clang-format/reformat-formatted-files

Change-Id: I52e58aa38c600faf677e932cb3caad07547ccbb7
(cherry picked from commit 02416cc1ce1f5d64265d509a0c1b149d11b5a3cb)

tdf#120145 ww8import: ignoreCols if section is inserted

Otherwise, the column setting is duplicated both in the section
and in the page style.

Change-Id: I14383c646e709a3653f1054f0d4170a2963529c1
Reviewed-on: https://gerrit.libreoffice.org/66151
Tested-by: Jenkins
Reviewed-by: Justin Luth <justin_luth@sil.org>
(cherry picked from commit 84fefd7c295fc05499ca222dff50c2fe4e0fb27e)
Reviewed-on: https://gerrit.libreoffice.org/66217
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
(cherry picked from commit 0e863f5529318e07f46568150e489cb0bef9b616)

Add css_comp_Impress_oox_PowerPointExport

Change-Id: I17d95cce0dd69e2ba8b4ebc7045544fc5659a249

Fix "external" graphic filter code for the DISABLE_DYNLOADING case

The DISABLE_DYNLOADING case in ImpFilterLibCacheEntry::GetImportFunction()
had apparently been broken from the start, but nobody never noticed as
the common formats (PNG, JPEG, GIF etc) use a more "internal"code
path. The problem was noticed now with a TIFF image in a presentation.

The fix for the import case seems to work, but I could not test the
fix for the export case.

Change-Id: Icbd33c7cb654136a562f18b06d5b0896e701bf3d

mysqlc: Add support for mysql type INT24

Which is mapped to sal_Int32.

Change-Id: Ibf12e92a20034440fa990ed0c6f1196f4ca3f40f
Reviewed-on: https://gerrit.libreoffice.org/73218
Reviewed-by: Andras Timar <andras.timar@collabora.com>
Tested-by: Andras Timar <andras.timar@collabora.com>

lok: Fix status of InsertPage, DeletePage and DuplicatePage.

Status true / false means checked, status enabled / disabled means the
visibility (enablement) of the uno command.

Change-Id: Ib2671984f59ffd5e6f1823a65ea42b33bdfe67d6
Reviewed-on: https://gerrit.libreoffice.org/73384
Reviewed-by: Szymon Kłos <szymon.klos@collabora.com>
Tested-by: Szymon Kłos <szymon.klos@collabora.com>

[cp] sw HTML import: change config default to always ignore comments

Change-Id: Iae6191fbcf1d02e9e5dd29613eaf5c36646afccc
Reviewed-on: https://gerrit.libreoffice.org/73084
Reviewed-by: Michael Meeks <michael.meeks@collabora.com>
Tested-by: Michael Meeks <michael.meeks@collabora.com>

sw HTML import: add an option to ignore comments by default

No function change, default behavior is still mapping them to postits.

(cherry picked from commit 256107622c252afdce35d1c77fe70eb4e4bc57ce)

Conflicts:
sw/source/filter/html/swhtml.cxx

Change-Id: I55041452909e9f0b83d6c5d113af3067b3c0afdb
Reviewed-on: https://gerrit.libreoffice.org/73083
Reviewed-by: Michael Meeks <michael.meeks@collabora.com>
Tested-by: Miklos Vajna <vmiklos@collabora.com>

Revert "[cp] sw HTML import: ignore comments unconditionally"

This reverts commit f6d2125bf3511701d261b83bbfc50785b707d026. No need to
hardcode this, we'll have a proper option instead.

Change-Id: I9778c19715d52bfa066585791b2002e9837e1d14
Reviewed-on: https://gerrit.libreoffice.org/73082
Reviewed-by: Michael Meeks <michael.meeks@collabora.com>
Tested-by: Michael Meeks <michael.meeks@collabora.com>

Add distro config for iOS simulator to make building easier

Change-Id: Iff70622732792224b3c0a080f95d69f86c167ae6
Reviewed-on: https://gerrit.libreoffice.org/73369
Reviewed-by: Tor Lillqvist <tml@collabora.com>
Tested-by: Tor Lillqvist <tml@collabora.com>

LOK: Factor out iOS painting code

Change-Id: Ica706842003fe60ba864e013614eb535580446bb
Reviewed-on: https://gerrit.libreoffice.org/73365
Reviewed-by: Tor Lillqvist <tml@collabora.com>
Tested-by: Tor Lillqvist <tml@collabora.com>

tdf#123697 we need to create a bitmap context with alpha

Change-Id: Ib934314e56163983e70a0b04eb667e6b8527c783
Reviewed-on: https://gerrit.libreoffice.org/73364
Reviewed-by: Tor Lillqvist <tml@collabora.com>
Tested-by: Tor Lillqvist <tml@collabora.com>

tdf#124364: Add some missing components from chart2, svtools, and xmloff

Change-Id: Ie9efd1c5b4f5f6ffa66215df27a78027ba0dea39

Add the Visio import filter constructor

Change-Id: I6f1ebf18e0565862703bcd4291f2705c8a605de0

Add the librevenge library

Change-Id: I786001d66e9b40df5c30e19ff2381c9ed6c7db94

-Werror,-Wself-assign

...ever since 2bd76c3928d6250c287fb8910e8843d035100e5e "INTEGRATION: CWS aw033",
and looks like it wants to use the

  aRetval = rCandidate.getB2DPoint(nIndex);

already assigned to it ahead of the if/else branches.

Change-Id: Iac242281312fe28b9a17a6a2a930e9597f77b7aa
Reviewed-on: https://gerrit.libreoffice.org/52811
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
(cherry picked from commit 4bf1bf9dd4598f34e43c7c8a69f3665175c3cae5)

tdf#122765: Legend interferes with pie chart after pptx import, second part

Revert of:
2903c5f59d257b6ab94f32df64513f2edf94fb0b

Reverting this commit does not lead to the original bug
to come back, so we can revert it without a problem.

Change-Id: I244a6b9451c47e1094db8a77c71b6696e0c464cc
Reviewed-on: https://gerrit.libreoffice.org/73139
Reviewed-by: Tamás Zolnai <tamas.zolnai@collabora.com>
Tested-by: Tamás Zolnai <tamas.zolnai@collabora.com>
(cherry picked from commit 63d17d01da0f9e3a7554e4327244ee4457b7188d)
Reviewed-on: https://gerrit.libreoffice.org/73160
Reviewed-by: Andras Timar <andras.timar@collabora.com>
Tested-by: Andras Timar <andras.timar@collabora.com>

tdf#122765: Legend interferes with pie chart after pptx import

Revert the commit caused this regression:
0fc41c53dfbd21e526fb0ad68a6651693c4a2ecd

The original issue does not come back with
reverting this commit.

Reviewed-on: https://gerrit.libreoffice.org/72679
Tested-by: Jenkins
Reviewed-by: Tamás Zolnai <tamas.zolnai@collabora.com>
(cherry picked from commit 10609749126ca76eaf12904d4cce9cc5a16d8405)

Change-Id: I666c4f92e3b70b416ec6da7a704298d207451649
Reviewed-on: https://gerrit.libreoffice.org/72767
Reviewed-by: Andras Timar <andras.timar@collabora.com>
Tested-by: Andras Timar <andras.timar@collabora.com>