Age | Commit message (Collapse) | Author | Files | Lines |
|
Change-Id: Ib4c0df8d7637bff7d7a45d24482fff4342169fae
|
|
multiple documents sharing the same doc pool so can't unset its
secondary here.
Here the assumption is that SharePooledResources has always been called on Clip
or Undo docs (otherwise persumably need a flag which SharePooledResources
toggles on)
Change-Id: Ibcdbe5f698102dc683c290d3bd581c31e214765f
|
|
Change-Id: I8b977766e694ab449d612322138f5baa23fef29b
|
|
Change-Id: I83fe85a50e6e98da02a7c46a9e0c07a043244d56
|
|
Change-Id: I86ec7c73090b93cf080ced2bd5e24d98e2b3ac0e
|
|
Doing this ahead of time eliminates the need to update script types of
all replicated cells after the paste. This makes significant performance
impact.
Change-Id: Ic4f7d7b2fe152bd5640ddb1dae01cc2ed757657e
|
|
Change-Id: Icdb68af83a0b208f9a427ee61ae18020c8835a41
|
|
Change-Id: Id2957f1cb68fc4b2fef9966abeb6745b548bcb75
|
|
Change-Id: I4e6b362f31d93964567fe9ca394cd96568981746
|
|
Iterating over a range and attempting to get a broadcaster for each cell
position is a performance bottle neck. Take advantage of the column's
existing maBroadcasters structure instead.
Change-Id: I5467a64ee3c0b5f430be1f0c4b940d3f71874827
|
|
Change-Id: I9cebe8f7979fd0b502fb23e7048b24a5b9ceb5d4
|
|
Change-Id: I9a426f40f8cdce1f1ebe207dd9ada8154d8fde46
Reviewed-on: https://gerrit.libreoffice.org/12195
Reviewed-by: Samuel Mehrbrodt <s.mehrbrodt@gmail.com>
Tested-by: Samuel Mehrbrodt <s.mehrbrodt@gmail.com>
merge fixes
Change-Id: Ibca73c7099739790e1659a24841f168a2c06b71e
|
|
Change-Id: I5846940cc45781f4a0264047107dbb568604d543
|
|
Change-Id: I5846940cc45781f4a0264047107dbb568604d543
|
|
Change-Id: I9a426f40f8cdce1f1ebe207dd9ada8154d8fde46
Reviewed-on: https://gerrit.libreoffice.org/12195
Reviewed-by: Samuel Mehrbrodt <s.mehrbrodt@gmail.com>
Tested-by: Samuel Mehrbrodt <s.mehrbrodt@gmail.com>
|
|
Change-Id: I9a426f40f8cdce1f1ebe207dd9ada8154d8fde46
Reviewed-on: https://gerrit.libreoffice.org/12195
Reviewed-by: Samuel Mehrbrodt <s.mehrbrodt@gmail.com>
Tested-by: Samuel Mehrbrodt <s.mehrbrodt@gmail.com>
|
|
Change-Id: I5846940cc45781f4a0264047107dbb568604d543
|
|
No need to iterate through all cell positions and attempt a broadcast if
no one is listening. Instead, broadcast only the cell broadcasters and
use AreaBroadcast for the area listeners.
Change-Id: I1e666e8ff19ac0715f73d73f54da2e4c8d523173
|
|
FormulaGroupInterpreterSoftware is known to be broken, says moggi. We should
not use it as a fallback to OpenCL.
Not sure whether it makes sense, but let's keep it in the code for now. Make
using it conditional on setting the environment variable
SC_ALLOW_BROKEN_SOFTWARE_INTERPRETER (to any value). Only a developer who
wants to work on it should set that.
sc::FormulaGroupInterpreter::getStatic() can now return NULL, adapt callers
accordingly.
This might fix fdo#87119, but I am not able to check, as the document from
that bug causes an unrelated assertion for me.
Change-Id: I24454f46332014cbcbf696252059b9743f3c84d6
|
|
Change-Id: I5fd2c7635f204bda982f1df58b4c19fe9b12464a
|
|
It could happen that during a SetDirty/Notify cycle a formula cell is
appended to the formula track but not tracked yet so doesn't end up in
the formula tree. If it was deleted then without removing it from the
track the cell pointer shortly after was moved into the tree, possibly
setting pFormulaTree (and/or pEOFormulaTree) to that cell if it was the
last cell, and if immediately after that a new ScFormulaCell was
allocated at exactly the same memory location it virtually ended up as a
successor of itself in the formula tree ... leading to a crash if pCode
was accessed in a subsequent RemoveFromFormulaTree because the cell was
assumed to be already in the tree.
Change-Id: I9d1885a26b85c2331a084b5f89a2d7373cf2df0f
|
|
Change-Id: I61b6912da303b9d71eed142535f3aa0cac1681eb
|
|
Not only are multiple boolean parameters ugly, but prepare also for the
new broadcasters-only mode yet to be implemented.
Change-Id: Ie6383826e76a771b88e7b4b29e5de9a58c598ad5
|
|
on the basis that it doesn't look obviously wrong to the original author after
a once-over
Change-Id: If333a17a2a9954d33ebefc6846e40e0e1e29c857
|
|
Change-Id: I82c771f0ca7d2593a6f34dd84689cf055f9cf3f3
|
|
The SdrPage that contains the shapes for the note captions is deleted
before the notes. Prevent that case by deleting the notes before
deleting the SdrPage.
Change-Id: I13727cea7df0b033fb386a6a8e20dcadec743335
|
|
Undo Sort did not select/mark the original range if it included header
row/column.
Change-Id: Ie4c800e7a44fd910c871731673093400c6400e4a
|
|
Change-Id: I1847b2900b513cc9b7b47767a15fb7e11d0abf26
|
|
That special case only works if the whole source range contains formula
cells. For now we just limit it to the special case of a single source
cell.
Change-Id: I71ccfde06edd97e5a9c78da22053583d05aac411
|
|
Change-Id: Id652ff18023965cb2bcc7c10520f95a51e8ac15d
|
|
Change-Id: I107201ab736b548f010d813540e0a9cec7402e79
|
|
... as a result of the introduction of range-based area listeners.
With this change, the insertRangeData() function for ucalc no longer needs
the additional bGroupListening flag. All tests pass with group listening
enabled at all times.
Change-Id: I9b9fb9443c727ff62badbd60ec9cd94094eb5a45
|
|
... now that also empty cells are to be broadcasted.
Set dirty and broadcast only the effective data range as determined by Sort.
This is more a workaround, a cleaner solution would be to refactor the
SetDirty() algorithm to iterate only through broadcasters and use
AreaBroadcast() to notify area listeners.
However, this can also be easily backported to 4-3.
Change-Id: I6d68ca0088cec6a8328a3e93364ac928ef69babe
|
|
Change-Id: Ieaeea85149e4b6d263b112205bb17c155e2d12be
|
|
Change-Id: I982e03a12dd80be0787f22dce4495065775e7de0
|
|
Change-Id: Ib256ed35a9ecb59a07ed31edec9bd8abb31a4610
|
|
Empty columns were being unnecessarily included in the sorting
range which caused extra allocation. This happens when you select
the entire sheet to sort. There was already code to skip empty
trailing cells in rows but not to skip empty columns.
Change-Id: Ieac22647bb13165cbf66cb1a93f567e5169fff6a
Reviewed-on: https://gerrit.libreoffice.org/13294
Reviewed-by: Eike Rathke <erack@redhat.com>
Tested-by: Eike Rathke <erack@redhat.com>
|
|
Change-Id: I48c2421598152b64d40237eb0e917a680a27b643
|
|
Change-Id: Ia03dc71a3d4e4380258e7f8d2ebc3c9f32d83493
|
|
One for IsUpdateRefs() true, the other for IsUpdateRefs() false.
Change-Id: I2b18e2b8a9770fd9499dd9fc0d26397e67c1e8ae
|
|
This code block is sufficiently self-contained.
Change-Id: I0eb2d3ee8175af60394ff4c408a9c245f77044d8
|
|
When I said 'Union' originally, I meant 'Intersection'. Let's fix that.
Change-Id: Iaef5035bb1dbb676c63bd17b3650266605104181
|
|
Change-Id: Ie370161e51ff83cb605bc57d317ff945404e5611
|
|
This reverts commit b8a7c39cf26f6d417efc34acdf6f4c2a6762052d.
Bah, this was a leftover of an idea I didn't follow..
|
|
Change-Id: I3c89e649c289cd50035929d4025d0c18968b36f9
|
|
Change-Id: I378cd0e1eaa0f41799e4e18d1daae813540da77d
|
|
Change-Id: I62c3714c8036fe8d4b70ee188546c7d69ece1ea5
|
|
... and force the caller to make a decision. In fact some needed to
include empty cells.
Change-Id: I82b6ef72f7ecda6543b3cafc70f475519870b471
|
|
Change-Id: Iba030c1a4fd3dad34bdd61377b73fa76a062b5f8
|
|
Change-Id: I826db178ceea4a409ad9e14ff73b5edaf136e435
|