Age | Commit message (Collapse) | Author |
|
We don't have there Select(bool) since
c3a89936abc7967ebe8916ca83cb38f6837f9aa1.
(cherry picked from commit 4604a6a7cd66907fc29e01aacf2146f17fa19d34)
(cherry picked from commit bc3c7f2557a1871fb1e41fc9cac4426a6b208901)
Conflicts:
sc/source/ui/cctrl/tbinsert.cxx
sc/source/ui/inc/tbinsert.hxx
Change-Id: Id496725d82289d1be004bdc36195877c12f336b0
Reviewed-on: https://gerrit.libreoffice.org/16773
Reviewed-by: Eike Rathke <erack@redhat.com>
Tested-by: Eike Rathke <erack@redhat.com>
|
|
Change-Id: Ic3ba292e28fe12d7dcc2c2e67aeea48a4c8aaac2
(cherry picked from commit b161552bd9f7d6b6de9752e4f0e7d6f65bbcf42e)
(cherry picked from commit ea228fdffd17b87e398216625213a691fcb34825)
Reviewed-on: https://gerrit.libreoffice.org/16714
Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>
Tested-by: Miklos Vajna <vmiklos@collabora.co.uk>
|
|
IMDeleteSurrounding is used by input methods to take multiple
keystrokes and convert them into a special character. Then the
composing keystrokes are removed.
Before this fix, the cursor placement was not adjusted,
so the special character was inserted in the wrong position
and if 3+ keystrokes were involved, the wrong characters were deleted.
Also fixes "editLine should have focus on accessibleText init."
The first time an accessibleEdit is created, it didnt recognize any
focused text when editing in the "Input Line".
Change-Id: Ib2608c46bba1c84bb49aa29619a004c125a3bf98
Reviewed-on: https://gerrit.libreoffice.org/16637
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>
|
|
Change-Id: I70218bc41df0d56ab42d652aa7ac51733dc06f4b
(cherry picked from commit 71dba79c1c84d693a59c53a8965caef0ddd1c2cc)
Reviewed-on: https://gerrit.libreoffice.org/16496
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>
|
|
... that take relative/absolute addressing and sheet 3D flag into
account to be fed to reference extension via range operator.
(cherry picked from commit fb6dd2a73074b9695bd8ddf7ba40f1819b03024e)
properly inherit relative and 3D flags when extending, tdf#83365 related
Commit 194e9f9bae28bdf22a9ed4779c1656ee693f3302 oversimplified things.
(cherry picked from commit e503addfbbe45efe1a3c06392c66fb79c3c72d07)
Change-Id: Iabe13ae384577e2d71ca87af6482ddccbf7ad0ac
ea6a84c4a7be49af036690afbb1512ae2c1045a2
Reviewed-on: https://gerrit.libreoffice.org/16352
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>
|
|
Try to resync shared group top and length.
Change-Id: I31bd0db7c1dceb880a22274edc4c3f20ce253095
(cherry picked from commit d8541c2a62121894bf87c91f1f89aea1ea30d680)
Reviewed-on: https://gerrit.libreoffice.org/16533
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>
|
|
Change-Id: Iac8574329c8c8e0bc0ac956993ccdd085372a6cc
(cherry picked from commit 5e83f49e5d62587f427ad416f5188ce81506c05b)
Reviewed-on: https://gerrit.libreoffice.org/16598
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>
|
|
No trailing characters must be present in 1:1 or A:A full row/column
references, e.g. 2:2,2 is not valid. The original cause is the lexical
analyzer that accepted 2,2 as the second part after the range operator
because ',' is also the group separator here, which needs further
changes.
Change-Id: Iff354469f7dcb4e0b6fa645c39cc1f8ebad4a6ea
(cherry picked from commit c48af2b55a60492565b987ee65d98edc6bd94533)
Reviewed-on: https://gerrit.libreoffice.org/16490
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>
|
|
(cherry picked from commit 77cc71476bae2b3655102e2c29d36af40a393201)
Conflicts:
sw/source/core/doc/DocumentLinksAdministrationManager.cxx
sw/source/filter/xml/xmlimp.cxx
Change-Id: Ida1257337c6e0916f2228fe053d9c9f085183af6
Reviewed-on: https://gerrit.libreoffice.org/16422
Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>
Tested-by: Miklos Vajna <vmiklos@collabora.co.uk>
|
|
Change-Id: I541577e0b99b0144a755e5755adc890c0ca8d204
(cherry picked from commit ffc1ffed11dc63a69fc2db04f12b3ea266b580fe)
Reviewed-on: https://gerrit.libreoffice.org/16326
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>
|
|
Change-Id: I624e64745fd3874be3e1bd3df6bac18dfb17aebb
(cherry picked from commit c1fc84ac140d519e0bfa7a607e36771682b08eed)
Reviewed-on: https://gerrit.libreoffice.org/16258
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>
|
|
anchored to cell but changes position after reopening
Also included tdf#68797 "FILEOPEN lost position of lines
anchored to cell". It was marked as duplicate but the
step to reproduce are different.
Conflicts:
sc/qa/unit/subsequent_filters-test.cxx
Conflicts:
sc/qa/unit/subsequent_export-test.cxx
Reviewed-on: https://gerrit.libreoffice.org/15523
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Andras Timar <andras.timar@collabora.com>
(cherry picked from commit 487880b6882ec01c1b4679eae60bec484272a86b)
Conflicts:
sc/qa/unit/subsequent_export-test.cxx
Change-Id: Ia1c4010f118749256077a0ecad6ca16b867d22f7
(cherry picked from commit b1f1c77687e301c99e81631d33298e44bdb5ca84)
Reviewed-on: https://gerrit.libreoffice.org/16280
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>
|
|
... not in cell areas of all sheets.
(cherry picked from commit 80ec99db4325a439a8a3f1d420d0a80f8bf9c439)
Conflicts:
sc/source/core/data/document.cxx
Change-Id: I8d19c15c6d8d89652ac5695214bc4dd930783087
Reviewed-on: https://gerrit.libreoffice.org/16303
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>
|
|
regression from
commit 3d6521280929ecacc53b7c358d29d0b5d31b3462
CommitDate: Thu Jul 31 22:14:25 2014 +0200
fix memory leak around function descriptions
Found by Lsan.
There are two implementations of getCategory, one (sc) returns a new one each
time (hence the leak fix) and the other (reportdesign) returns a pointer to one
that belongs to the manger (hence the crash).
The code in formula really looks to me to expect that the getCategory return a
pointer that "someone else" needs to look after, i.e. the reportdesign variant
is the more correct so revert 3d6521280929ecacc53b7c358d29d0b5d31b3462 and to
fix the leak make the sc own the ScFunctionCategories and just cache them like
the reportdesign one does
Change-Id: Ifd986301a54b4d20449e864697655cd973e0c4df
(cherry picked from commit 7c3abee29c742593206b755b20a718c46f0780fa)
(cherry picked from commit 73107eb3375f1671f549f0467be2812df9223848)
Reviewed-on: https://gerrit.libreoffice.org/16232
Reviewed-by: Eike Rathke <erack@redhat.com>
Tested-by: Eike Rathke <erack@redhat.com>
|
|
i.e. setting progress triggers ScTabView::DoResize and an InterpretVisible and
InterpretDirtyCells which resets the mpFormulaGroupCxt that the current rCxt
points to, which is bad, so disable progress for the duration of the GetResult
loop
Change-Id: I8e88cee4dd2308ef61dee934d300a38978833703
(cherry picked from commit e3c1a394a00cef416a81b89b6d5c204891abb286)
Reviewed-on: https://gerrit.libreoffice.org/15926
Reviewed-by: Eike Rathke <erack@redhat.com>
Tested-by: Eike Rathke <erack@redhat.com>
|
|
correct 3D reference update on sheet deletion, tdf#90001 follow-up
be1b3daeeb64a44e577510561d02147b87db72c2 missed a few cases
* a deleted reference start was shifted by two sheets instead of one if
the formula was on a sheet before the deleted one
* a deleted reference end was not correct if the formula was on a sheet
behind the deleted one and used relative sheet reference
* #REF! wasn't set when the entire referenced range was deleted
(cherry picked from commit 20e4ccc6ba366cd70c9218e016ccff4025e3d816)
b9c7492a651a8429fa1c411c8e447593e366e09d
Change-Id: I501ad421c283ec6a1f2629aa424f4d4c32228516
Reviewed-on: https://gerrit.libreoffice.org/15043
Reviewed-by: Eike Rathke <erack@redhat.com>
Tested-by: Eike Rathke <erack@redhat.com>
|
|
IDF_CONTENTS includes IDF_NOTE, which should not be used here
Change-Id: I904fbdbfcb97865a5682d9073e0ce2d5a2c694fb
Reviewed-on: https://gerrit.libreoffice.org/16085
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Michael Meeks <michael.meeks@collabora.com>
Tested-by: Michael Meeks <michael.meeks@collabora.com>
(cherry picked from commit a977937927729eaff61bd670fdd45d3508b19371)
Reviewed-on: https://gerrit.libreoffice.org/16091
|
|
overwriting the original one, and when the temp doc goes away then its drawing
layer goes away autoamtically settings the secondary pools of the other pools
it is a secondary of back to 0
(see also tdf#87245)
Change-Id: Ie6dd8d615f81cf4d4776eaeac1cc2507b2abaa37
(cherry picked from commit 8ae25bbc092254f680e2976873f15bdbc26615ef)
Reviewed-on: https://gerrit.libreoffice.org/16075
Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>
Tested-by: Miklos Vajna <vmiklos@collabora.co.uk>
|
|
otherwise pressing return in the object goes to the
sheet and attempts to paste over the current cell
reproduce with insert->comment in A1, insert some text and
click in e.g. A5, ctrl c, right click on A1 and show comment,
double click on comment to edit, go to end of text and
press enter, either "bad things happen" if a11y is enabled
, or "nothing happens" instead of a new line appearing in
the object
Change-Id: Ie44851cf3394652e1a60eecb59a999889094b3c5
(cherry picked from commit 7b75dc8ccca61988392779d4c61eef15f1211c16)
Reviewed-on: https://gerrit.libreoffice.org/15942
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
Tested-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
|
|
Listeners are set up from references in RPN, so check those for bounds
to catch also references resulting from named expressions, database
ranges, tables, ... and references in the token code array that are not
referenced in RPN.
(cherry picked from commit 4baf76ddb39580678cf14019900be78bb9071d7b)
Windows MSVC: cannot specify explicit initializer for arrays
(cherry picked from commit 883ebe0283dc6bdf62f08191dede2a249f777f63)
c42897ba6bb520c931f63e56d0f453ed14cfaa3d
Change-Id: I54770b45818f4c0541a39815278d3271a77b345d
Reviewed-on: https://gerrit.libreoffice.org/16154
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>
|
|
(cherry picked from commit 2f6a06856ad8df0c11a112d1e457b408e9a7af1d)
tdf#90694 reset group area listeners on correct top
Follow-up on 2f6a06856ad8df0c11a112d1e457b408e9a7af1d, rTop is top of
created group, listening needs to be ended on original (previous) group.
(cherry picked from commit 5d6448a5131488eac3a6af16bff06140663c3db4)
1f975a0a6c85e649aa07ce5c6e97bc6e917ff9ce
Change-Id: Ib3e85e1f7e12447fc2998711663ea1e9d1b322e5
Reviewed-on: https://gerrit.libreoffice.org/16104
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>
|
|
Add a test to check if name already exists
Change-Id: Ie23de04c0b747f58ab61a6e498bf212bf38b346a
Reviewed-on: https://gerrit.libreoffice.org/16000
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>
(cherry picked from commit 3807946b5515e943169da907866c93b52a2b3f19)
Reviewed-on: https://gerrit.libreoffice.org/16056
|
|
This fix is a SCCOL variant of:
commit c66d1b9fcc2244b4fd8940f17ebf4e772f09c84e
Author: Kohei Yoshida <kyoshida@novell.com>
Date: Thu May 5 00:23:25 2011 -0400
fdo#36406: Let's not use invalidated iterators.
std::set::erase(iterator) call invalidates the iterator of the erased
element. We better not use it after the erase() call. Since the number
of manual breaks should not be high enough to cause a performance
bottleneck under normal usage, a safer linear copy should just be fine.
Now, I'm not sure if this indeed is the cause of the erroneous amount
of manual breaks in the test document, but I didn't see any other likely
spot.
Change-Id: I0c2e8f738949776aa7f8ea5528e2c0eeb9351c16
Reviewed-on: https://gerrit.libreoffice.org/16063
Reviewed-by: Eike Rathke <erack@redhat.com>
Tested-by: Eike Rathke <erack@redhat.com>
(cherry picked from commit 1ba1cbe48c40d649c33661abb32369c39a23fff4)
Reviewed-on: https://gerrit.libreoffice.org/16081
|
|
the cell content is now drawn over the cell note rectangle
Change-Id: I6874ecd879b01ebc27b5eb912d472af704387a1f
Reviewed-on: https://gerrit.libreoffice.org/15823
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
Tested-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
|
|
... in case the header is the only row.
Change-Id: I5e6046007a8d668f9834e108aaf8af0072629fc8
(cherry picked from commit 46fa99f61aff88f1697959a9d3c41a5c3c3c05e9)
Reviewed-on: https://gerrit.libreoffice.org/15631
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>
|
|
This adds ScDocument& to all RefUpdate...Context; another approach could
had been to add an UpdatedDBData similar to UpdatedRangeNames and gather
those in the ScDBData::Update...() methods, but as long as
ScDBData::IsModified() works that isn't necessary.
(cherry picked from commit 848dc76e76c1c4a9040be4a0088c4d4527af6f40)
Conflicts:
sc/source/core/tool/token.cxx
assume modified if an index has no corresponding ScDBData (anymore)
(cherry picked from commit 4a53021f0662e08b56fd076f78ac182634fcec4c)
16cbc7d772f39c8778b8ba21a629ced6b0dbedeb
Change-Id: Iae3ec6c8b8356cbd8acb2445489e91e7d6922fb3
Reviewed-on: https://gerrit.libreoffice.org/15654
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>
|
|
See source code comment.
(cherry picked from commit cff5ee864b2d87d74079697425d7895dbf1b2ba4)
more SAL_INFO sc.core.grouparealistener
(cherry picked from commit b03563571fb922636635ea72f2dbda18d736ff89)
Conflicts:
sc/source/core/tool/grouparealistener.cxx
a3c75e626edee3ad28ad5b7eb80f729ae8aaa83f
Change-Id: I3ab7ab6aec1d782de0733064fea031c895f28965
Reviewed-on: https://gerrit.libreoffice.org/15331
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>
|
|
Change-Id: I5ab279a7d7af6829735b47ef5228f56dc77580c9
Reviewed-on: https://gerrit.libreoffice.org/15448
Reviewed-by: Eike Rathke <erack@redhat.com>
Tested-by: Eike Rathke <erack@redhat.com>
|
|
Change-Id: Ic68ddc56d9cf9dd410c1f073ba83bfeea8220bf0
(cherry picked from commit cee6eb254fae1012bc6ffe1b2e440fe58fedac1b)
Reviewed-on: https://gerrit.libreoffice.org/15318
Reviewed-by: Eike Rathke <erack@redhat.com>
Tested-by: Eike Rathke <erack@redhat.com>
|
|
mnLastIdx doesn't get reset when the color list is reset, the code already
acknowledges that mnLastIdx is the last set index but that it may not be a
valid position, so only set rnIndex if mnLastIdx is valid and the data at that
pos is the desired data.
Change-Id: I57ce32fd1a13bfe9049ed94be34229c8c2808b8e
(cherry picked from commit 6c3b1361a97f5219f90bf2f09f3c3f2b70b8e30e)
Reviewed-on: https://gerrit.libreoffice.org/15320
Reviewed-by: Eike Rathke <erack@redhat.com>
Tested-by: Eike Rathke <erack@redhat.com>
|
|
Change-Id: Iaacf5636749077efc6f91f0eb0bac477cfcf4553
Reviewed-on: https://gerrit.libreoffice.org/15129
Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com>
Tested-by: Markus Mohrhard <markus.mohrhard@googlemail.com>
|
|
Another UniString to OUString conversion fallout.
Change-Id: I5e62b049da3e7f8b5a892ea6aae7110a33564a46
(cherry picked from commit 9f52efa43b2d65c0b6c92790cc2859e752bf975f)
Reviewed-on: https://gerrit.libreoffice.org/15286
Tested-by: Markus Mohrhard <markus.mohrhard@googlemail.com>
Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com>
|
|
Change-Id: Ic6d6e7d7d8d42af0fc739d964a1190d40f9dba2e
(cherry picked from commit b56b5d449fc74809fddc9b9bbd1d9f8d244a0d10)
Reviewed-on: https://gerrit.libreoffice.org/15190
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
Tested-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
|
|
Conflicts:
sc/source/filter/oox/stylesbuffer.cxx
Change-Id: I8489b0341709d44bc62e5ddef49fe232998f3c80
Reviewed-on: https://gerrit.libreoffice.org/15032
Tested-by: Katarina Behrens <Katarina.Behrens@cib.de>
Reviewed-by: Katarina Behrens <Katarina.Behrens@cib.de>
|
|
This implements search criteria "" and "=" to match empty cells in
spreadsheet functions SUMIF, AVERAGEIF, COUNTIF, SUMIFS, AVERAGEIFS and
COUNTIFS.
Change-Id: I1b4a4c14bac7b974428bf64afb549707a0d75a90
(cherry picked from commit 01b615687fe0f39c65e0e8290db434db2f1ef8ac)
Reviewed-on: https://gerrit.libreoffice.org/15091
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>
|
|
Change-Id: Iae73c46d6ae5b141c3f5680b2623ee06205c22d2
(cherry picked from commit 3cae8ee10f297e42adf9f3b33809b4e7e3af2840)
Reviewed-on: https://gerrit.libreoffice.org/15111
Tested-by: Markus Mohrhard <markus.mohrhard@googlemail.com>
Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com>
|
|
Change-Id: I5a980f0b1ca47b18ce2e479e46971d7c6690c437
Reviewed-on: https://gerrit.libreoffice.org/15125
Reviewed-by: Eike Rathke <erack@redhat.com>
Tested-by: Eike Rathke <erack@redhat.com>
Reviewed-on: https://gerrit.libreoffice.org/15126
|
|
Change-Id: Ib97509609bd3e6629e3efd0c633535564f1c64d6
(cherry picked from commit f3989e4d3d87f07a484d7c404bc2bfc678faa7f0)
Reviewed-on: https://gerrit.libreoffice.org/15096
Tested-by: David Tardon <dtardon@redhat.com>
Reviewed-by: David Tardon <dtardon@redhat.com>
|
|
... instead of setting an error string with an unwanted side effect that
may lead to wrong results instead of error.
Change-Id: I42ade52e86520535c879e9bd68156873d706f33c
(cherry picked from commit 836d05d32e36aafc00de59ca51878f47f7ce816a)
Reviewed-on: https://gerrit.libreoffice.org/15072
Tested-by: David Tardon <dtardon@redhat.com>
Reviewed-by: David Tardon <dtardon@redhat.com>
|
|
Reviewed-on: https://gerrit.libreoffice.org/14920
Reviewed-by: Eike Rathke <erack@redhat.com>
Tested-by: Eike Rathke <erack@redhat.com>
(cherry picked from commit c79bdd062f657d1be98aa815d9b882d144f35e04)
only handle formula group if useful, tdf#90042 related
Same as in ScDocument::SetString() as changed with
c79bdd062f657d1be98aa815d9b882d144f35e04, do it similar in
ScDocument::SetValue()
(cherry picked from commit 02bea75dd6bc91759e987fafb5dccec6ce92b0c2)
7c666b176062c81a8a74cb49f345408c1060d973
Change-Id: I497a5bae4b08b46357415501ddbafc70585adfb0
Reviewed-on: https://gerrit.libreoffice.org/15036
Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com>
Tested-by: Markus Mohrhard <markus.mohrhard@googlemail.com>
|
|
for pivot table cache records.
Regression from commit d17a83fa549f828f29e6939. On loading pivot
table cache records, a dummy sheet is created to store them. Formula
buffers are now vectors, not maps and need to be resized in such case.
+ add a const version of getDocImport method, need to use it in
the new getAllSheetCount method
Change-Id: I5d57a822da5fe5a97029b83406e8ff877a5dbd7c
Reviewed-on: https://gerrit.libreoffice.org/15030
Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com>
Tested-by: Miklos Vajna <vmiklos@collabora.co.uk>
|
|
Change-Id: I2a8d4cce979ce0ff1fd260bfa796817a0dbb7ce2
Reviewed-on: https://gerrit.libreoffice.org/14512
Reviewed-by: Eike Rathke <erack@redhat.com>
Tested-by: Eike Rathke <erack@redhat.com>
(cherry picked from commit 38c22ed2b8b8f3530b8a0fe5a23a329e8e511c76)
Reviewed-on: https://gerrit.libreoffice.org/14989
Reviewed-by: Jean-Baptiste Faure <jbfaure@libreoffice.org>
|
|
Change-Id: Iba362cc2ee624271dcc078d7e7ded1ef6ff44ae4
Reviewed-on: https://gerrit.libreoffice.org/14933
Tested-by: David Tardon <dtardon@redhat.com>
Reviewed-by: David Tardon <dtardon@redhat.com>
|
|
Regression of 8850633fe4cdac306c35caab1a0d14c13d9df864 String to
OUString and xub_StrLen to sal_Int32 conversion, it worked by accident
only anyway..
Change-Id: Id6fea7075949b98e58294d9fdaae4313f69f9b55
(cherry picked from commit 9e2e9453cbca9399e1c670b354cfc3c23d70307f)
Reviewed-on: https://gerrit.libreoffice.org/14921
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>
|
|
That allows us to potentially change the import depending on the
producer of the document.
This becomes necessary to handle MSO 2007 chart drawingml streams
correctly.
Change-Id: I9be8b019fae69cd206203591982a89648965692f
Reviewed-on: https://gerrit.libreoffice.org/14876
Tested-by: David Tardon <dtardon@redhat.com>
Reviewed-by: David Tardon <dtardon@redhat.com>
|
|
Change-Id: I611531e86363d6c2b4ccce1aa0f1a940693bc8d8
Reviewed-on: https://gerrit.libreoffice.org/14658
Reviewed-by: Eike Rathke <erack@redhat.com>
Tested-by: Eike Rathke <erack@redhat.com>
|
|
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
Reviewed-on: https://gerrit.libreoffice.org/14824
Reviewed-by: Michael Meeks <michael.meeks@collabora.com>
Tested-by: Markus Mohrhard <markus.mohrhard@googlemail.com>
Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com>
|
|
Change-Id: I07400d6dead66ec437436b5ea8b49491f8048335
Reviewed-on: https://gerrit.libreoffice.org/14823
Tested-by: Markus Mohrhard <markus.mohrhard@googlemail.com>
Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com>
|
|
This reverts commit 104a1e641554be2e789758ae67c0e24620df8035.
A number sequence has to ignore all string and empty elements by
definition, see tdf#89872 for an example regression, further affected
are all functions that use ScInterpreter::GetSortArray() which in turn
uses ScInterpreter::GetNumberSequenceArray()
(cherry picked from commit 18de9b1de1aa404b3ca12e2018e5e5e220dd9786)
|
|
Use symbolic (not numeric) opcodes in the registry for the OpenCL use
subset. The opcode numeric values are not stable, and have in fact changed in
master recently.
Even if they were stable, storing them symbolically would be a good
idea. Makes it easier to edit them manually directly in the registry files.
(Something end-users are of course not expected to do, but admins or hackers
may want to do.)
I figured out how to do the mapping from symbolic names to enum values and
back without an ScDocument, turned out it was not complicated after all.
Change-Id: I062dbfa633668a645c6c308acacf0826b6459a93
Reviewed-on: https://gerrit.libreoffice.org/14775
Reviewed-by: Michael Meeks <michael.meeks@collabora.com>
Tested-by: Michael Meeks <michael.meeks@collabora.com>
|