Change-Id: Ida97950286bc69154d10cbc9b5ac699e7ad608ad

Change-Id: I7783d96e40de7a8df794f143d2c7830bf3314218

Change-Id: I715d49722ebe4658c108cb9b7bd6fa3934d973f8

Since commit 5bff4b01 the
SvxAutoCorrCfg::pAutoCorrect is cleared by one terminate() listener but
then another terminate() listener calls Commit() on all modified
utl::ConfigItem and these two have a Commit() that does not clear the
modified flag so they are always modified.

Sadly there's no non-virtual Commit() wrapper that
calls ClearModified() on the base class...

Change-Id: I9ae220d78bb109c7bf0fdc544754a0686b357115

Thanks to Eike for finding this:

The anchors ^ and $ now anchor at the selection boundary because the
only text the regex matcher gets passed is the selected text. This in
two paragraphs

 aaa bbb aaa bbb
 aaa bbb aaa bbb

when the selection spans from the second aaa to the third bbb, for
"^aaa" finds the second aaa, where previously it found the third aaa at
the real paragraph start.

This may not be expected by the user, because the behavior of ^ is
described as "Match at the beginning of a line" (or paragraph in our
case), which the previous implementation did.

(regression from 806ced87)

Unfortunately it's not obvious how to implement the same in
searchBackward().

Change-Id: I07f7a8476b672d9511fa74ca473c32eea427698f

Change-Id: I9ea963d09c2975b8e40856e3ba2d3896be9a1876

Change-Id: I4a43ea664e9db2bb44e0e7f089a71abf6f572b1e

Commit 8e2293e0 changed
STREAM_SEEK_TO_END logic, ensure that in a dbgutil build we catch
unnoticed corner cases with assert() instead of just another fly-by
warning.

Change-Id: If6c4fb4d2467f7dc7b8177caf7353d0b0e56ed0d

Remove the unnecessary ones, but make sure that e.g. the sofficeapp library is
built (which was not covered by depending on services.rdb).

Change-Id: I1adfa4aaaf60787c944d71cc289cdafba8351c1b

Change-Id: I543a561fc7fa1213c8e319f227b761f6328a6eaa

Change-Id: I1cff71d15f4d1b5ee47cc734b3fc2168d7f99201

only used for upstream packages/when epm is used (any auto-dependency
creation is disabled for those). Add two that might be missing in
barebone installations

Change-Id: I97c3af8cc8622c63ac831701e509097711822098

Change-Id: I47b1d8e7b119e50406b50f580ee06f9f1b5f8401

This also drops an unnecessary lock.

Change-Id: Ia5eb459adf3cb585aa2484866670949120315780
Reviewed-on: https://gerrit.libreoffice.org/14765Tested-by: Michael Stahl <mstahl@redhat.com>
Reviewed-by: Michael Stahl <mstahl@redhat.com>

We didn't actually check this correctly at all, but gladly overwrote the
allocated part of the output PDF, thus obviously rendering it invalid.

The parameter passed to PORT_NewArea is a default chunk size, not a maximum
anything, so it was misleading, even if not wrong as such, to pass
MAX_SIGNATURE_CONTENT_LENGTH to it. Use 10000 instead.

No need to do the overflow check twice in the Win32 case.

Change-Id: Ifa796dbb74b32e857f7184c1e8ada97ba124b020

This task waits for user io.

Change-Id: Id5b3725b316f1bbd73e94d5368bc0d5c7d057189

Change-Id: I4eb896813d6cec49bf5f1c8ac2ee8fafc06fbf52

this solves the bug of icons in the standard toolbar
under gtk3 of getting rendered with a black background
on first render

Change-Id: I73bae4cd0b9f38c9e9caef8019d773d42b5e8f10

all three addressed places hit frequently now during "make check", likely
related to recent changes to timer/idle code?

Change-Id: I75963c707807c29db7865b11f9ed532b779765d8

Change-Id: Id657299322ddefb095a2f192df818a873498e411
Reviewed-on: https://gerrit.libreoffice.org/14825Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Noel Grandin <noelgrandin@gmail.com>

Change-Id: I5d9734f302ff58637f69c1574bc39ca85ff992e7

Regression from b1534edf (tdf#89592
Using initializer_lists to populate Sequence<PropertyValue>,
2015-03-07).

Additionally, avoid makeAny() on the Any, the original code didn't do
that.

Thanks to lgodard/sberg for noticing these.

Change-Id: Icb3a6039885c582ba9f789c5c4768b865b593aee

Change-Id: I85993df91bd14966f9175c356966fbd8eb1a927d

since 44337a1f "Unit tests for PDF import"

Change-Id: I6c7746814af669dd0eb45b1013577cacc109a014

Change-Id: Idef9dd55eb1719eaf592bc4a86440cbd5aa4fb32
Reviewed-on: https://gerrit.libreoffice.org/14781Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>
Tested-by: Miklos Vajna <vmiklos@collabora.co.uk>

Change-Id: I2b21a488a7f597603b65fac0ff3c6b05c3660994

"... OpenCL on CPU"

I can't reproduce the problem any more. It was likely fixed by
fa4ce83f.

This reverts commit 219f2dc8.

Change-Id: Ib6d3a42137bcf21b0edec2173cf483ade6104b0b

Change-Id: I2d56b0ddf0b7e076ea570879b0ab33390306f3b2

Change-Id: If7351b2fb37b68ea2838cf442ca440a01587d66a

Change-Id: I4d8d0ee853e93c8620295e0598de358e66b4f730

after commit 924b2923
"V801: Decreased performance"

Change-Id: I32af1195f6e7e1fc4b94487bfbfab669bcad3951

Change-Id: I9e8d5863c31187b373ff93d5a6e992a1281220b4

Change-Id: Ic45e42d5ce7e0509cd2afc8be018f78cd6083aae

The TokenPool::operator[] is used to initialise and take into use an object
from the pool. Which is a fascinating thing as such and probably not entirely
in good style. Anyway, the objects in the pool are of type ScTokenArray, a
class derived from FormulaTokenArray. The operator[] called the
FormulaTokenArray::Clear() function to initialise the object. This left the
fields added in ScTokenArray uninitialised, having whatever value the previous
use of the object had set. Which of course is bad.

In practice, this showed up in the handling of formulas in the .xls input
filter. If an earlier (or the first?) formula had happened to be one for which
we don't want to use OpenCL, the meVectorState of its ScTokenArray object in
the pool had been set to FormulaVectorDisabled. When the same pool object was
later re-used for another formula, it kept that same meVectorState, even if
there was no reason to. Thus formula groups that should have been OpenCL
accelerated weren't. This can have a significant impact on performance of
document loading and recalculation for large documents.

I added a function to ScTokenArray to clear (initialise) such an object, both
the FormulaTokenArray part and the ScTokenArray-specific part, and call that
instead. This fixes the issue.

I named the added function ClearScTokenArray() to make it clear that it is a
separate function. Sure, possibly Clear() should be made into a virtual of
FormulaTokenArry and overridden in ScTokenArray, and the overriding Clear()
would first call the base class's Clear(). But I can't be sure that there
aren't other calls of FormulaTokenArray::Clear() that *must* mean the base
class one. Better safe than sorry.

And of course, I did *not* want to name the function in ScTokenArray also
"Clear()", like in the base class, without it being virtual. That is horrible
style in my opinion, even if there certainly is precedence for such even in
the very same classes, i.e. the Clone() function...

Change-Id: I0e0e13e5ca705603005a1e0a46866f095cd2ac4d

Change-Id: I07400d6dead66ec437436b5ea8b49491f8048335

Change-Id: I1e141632172f53ddbd2f5f434206646c9a1e9cf0

introduced in commit d22519f6
"V801: Decreased performance"

Change-Id: Ie8e1e00db91b0d1874abc6a3e6399ff30484f993

Change-Id: I84d997997100af4e21eafc5b458ff98d99c7dbbd

Change-Id: I6a888b58668ccb64c5a9c3270d06c3b9bcf8ce64

Change-Id: Ia076670e7356c378c6e8feb5970033638d2ba0b7