Change-Id: I3116dbf41b5672bb90c255b22403c8ce0ff57265

Change-Id: Ibd3598f1d053605dd7dab2fd08896ab4dbd25add

Change-Id: Id441395365d3cb6f598760042d5e54e905783b91

Change-Id: I5219578fbeb40a0fe363f03c64021a83f7e505fb

Change-Id: Id040721f5e41062605e635cb0f1e5b4c724052a3

Change-Id: I885b4c7a60ff51d44138c4f8a56a584c38a9804a
Reviewed-on: https://gerrit.libreoffice.org/6172Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
Tested-by: Stephan Bergmann <sbergman@redhat.com>

Change-Id: I3c87ac7a4696bee732fdb2bb9120ac3d50e9e664

Change-Id: I1f3ef757f76c88243c4be1f1a86042040eb524e8

Minor further changes by tml to match the coding style of surrounding
code mainly.

Change-Id: Ied6087a264f1c6b00763ea36fba9808329afede4
Reviewed-on: https://gerrit.libreoffice.org/5742Tested-by: Tor Lillqvist <tml@collabora.com>
Reviewed-by: Tor Lillqvist <tml@collabora.com>

Change-Id: I98cc2819e5de88cdce235735518b64193610fdc6

Should make the Android tinderboxes happy.

Change-Id: Idf33106514a1d0e34566d76d97d689e7543eda3c

Change-Id: I2894b7ffc186836890ab7b8117e2c7ca43bd21be
Reviewed-on: https://gerrit.libreoffice.org/6087Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>

Change-Id: Ibc5783aeea8bcdc05c14054fa10c43a0c98c2159

Change-Id: I839cbd6b3499f77867f993ece9f764db8f1c18c3

Change-Id: Ie5fc89d24cfff021e362811f3d2eed8608107767

Change-Id: I7527acd4b99b3b2a911532adb85b2a082c6569b7

Change-Id: Ifd7131fd91536fd04ae1c8ab3fc5959b8070ad5c
Reviewed-on: https://gerrit.libreoffice.org/6140Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>

70c35265 and
f89cce87 moved most stuff to instdir,
but forgot about epm and deb-builds. libgetuid.so that is needed to
build the debs is in instdir/ure/lib, but that wasn't added to the list
of include paths.

Change-Id: Iaf3f8cb2f6329dd66fe9f3862fd71f2037813d97
Reviewed-on: https://gerrit.libreoffice.org/6142Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>

Change-Id: I201985753fc53f037807e7870c6786a442717d6e

Change-Id: I55c85c1c44452c7fb6ac40591aea7177d054affe

regression since 55a15541

Change-Id: I9c5853061eef5befba7644ea0ca05b7751addc6c

Change-Id: Iee5c5ae3751d57c39e3dd1bcd1073a0e722bf575

...at least for its (only, it appears) uses at the left of the "Insert -
Hyperlink" dialog and at the left of a Database window.  As aBmpRect and
aTextRect never extend aBoundRect (and aTextRect always leaving a gap left and
right towards aBoundRect, it appears, with the text abbreviated with an ellipsis
if it would be too wide), there should be no need to eventually adjust "the
focus rectangle [to] not touch the text," and for calculating the top of the
focus rect aBoundRect can be used just as well as aBmpRect.  (Though for
calculating the bottom, aBoundRect apparently stretches "too far" compareed to
aTextRect; whatever...  The offsets look somewhat "magic" anyway, but appear to
produce the desired visual results.)

Change-Id: Ic324561294fb6172c0d02a697fc261717e01c87c
Reviewed-on: https://gerrit.libreoffice.org/6165Reviewed-by: Eike Rathke <erack@redhat.com>
Tested-by: Eike Rathke <erack@redhat.com>

(regression from 6b08fe83,
which changed SwUndoReplace::Impl::SetEnd to use the wrongly set end
position in rPam)

Change-Id: I14c6f58bc5c1418c69eb565d42f1829856eed58c

In some (all?) Python unit tests (or subsequentchecks?)
osl_getCommandArgCount() gets called even if osl_setCommandArgs() hasn't been
called, sigh.

Change-Id: I53ba20806da3168182fd2011397fd29e549869dc

Change-Id: I01e2067b4e410809be1278ab709a374d9a0ce314

Stephan says: I don't buy that argument of making things less fragile;
IMO it's that "defensive programming" approach gone wrong, where you
add complexity to mitigate a problem's effects in one place, instead
of fixing the problem itself in the real place. [...] The better
robustification is probably to turn the OSL_TRACE in
osl_getCommandArgCount (sal/osl/unx/process_impl.cxx) into something
stronger like SAL_WARN, assert, or std::abort

Change-Id: I3ba6684b4921eb67c2f4534e29e33056e8a891b3

This was the wrong way to fix the problem. The root cause was a static
initialiser in the OpenCL code, and that code should not even have been
built for iOS.

This reverts commit 261f1439.

Change-Id: I859ebb6e97a935d6e47a5269253165114fd90a1b

Change-Id: Iac0cd17dfa0544ec686120d50a6cfdb70336e042

Change-Id: I9d7c9bf832863c4b4776c7881ab7e44ab5df9607

Change-Id: I343b1d83db8e7538fcb5276a8701b51fe64a2a67

Change-Id: I0c7ed3d5bbad93adeba6175f0a9789fcb8ec3d3f

Change-Id: I32a594464c71215ee7557823aadaa72b8b72b4e2

Change-Id: Id9bd1fd6212eaa99c2e9efff6ea5d1d8c141f627

Change-Id: I249c3a157c7d6924f89e88853f6f63860ed2ff18

Change-Id: I759b22ca0f5bf7e3405572088d79263dd978f43c

Project: help  0168bad31cf97541695bffec867e078d76abf352

Change-Id: Ibff239592b517f8ba2ccd0649184e36f95cf1c2d

on HeaderFooterSettings

(cherry picked from commit b94437b6)

Change-Id: I4e21876f83fc160883c0ffdee5941ee7488e35c8