summaryrefslogtreecommitdiff
path: root/sc
AgeCommit message (Collapse)Author
2022-02-07Resolves: tdf#74577 tdf#92960 Create missing offsets in very outer tableEike Rathke
... if local table is outer table. Change-Id: I5996609264f4ad9efa8d76a8d0ab730bf1dded24 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/129540 Reviewed-by: Eike Rathke <erack@redhat.com> Tested-by: Jenkins (cherry picked from commit 8a9b870a40c41c7cae6e43452b593804fa17f831) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/129550 Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
2022-02-03Avoid calling AVX code in library static initializationThorsten Behrens
Despite 26072b8db7ba53f00c83197cb064229a76001989 and 12c6b1ef6a824b09778163ec83fc44bb196e65db, Linux builds using RH devtoolset-7 gcc7 still hit occasional unit test crashes with illegal instructions. Backtrace is always: > #0 __static_initialization_and_destruction_0 (__initialize_p=1, __priority=65535) at core/sc/source/core/tool/arraysumAVX.cxx:20 > #1 _GLOBAL__sub_I_arraysumAVX.cxx(void) () at core/sc/source/core/tool/arraysumAVX.cxx:112 > So lets make sure initializer is run only on function entry. Change-Id: I69f5cb725e4d72fb346166b21870612473f974b8 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/129246 Tested-by: Jenkins Reviewed-by: Thorsten Behrens <thorsten.behrens@allotropia.de> (cherry picked from commit d140817428cdbb519efa496f578bf6c054c94361) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/129251 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2022-02-03tdf#146742 - Move the initialization code of a method to SbMethod::CallAndreas Heinisch
Change-Id: I516bacdcd22ed094a0e9b6e125e941f32fe88c82 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/129238 Tested-by: Jenkins Reviewed-by: Andreas Heinisch <andreas.heinisch@yahoo.de> (cherry picked from commit 6a0f00d3b9d1a74637c92ec6eff1ba5fedc82f3d) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/129341 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2022-01-31Resolves: tdf#146722 Force row label for single text with numeric to the rightEike Rathke
Note this *might* break other uses of the non-deterministic automatic labels where exactly this constellation previously lead to a column label. For which expecting a column label though makes less sense.. Change-Id: Id85df7485f9b71620d8cbc404a981852d1567a54 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/129187 Reviewed-by: Eike Rathke <erack@redhat.com> Tested-by: Jenkins (cherry picked from commit 61cb699e88b3680becf78dc7b8bd514b2b7318a4) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/129079 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2022-01-30Resolves: tdf#147016 ensure focus returns to document after popup closesCaolán McNamara
Change-Id: I67e59afc08bc5983d473cb9dceb4fcefedeec117 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/129095 Tested-by: Jenkins Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
2022-01-26warning is a dialog not a popupCaolán McNamara
with the wrong setting it appears misplaced on a multi-monitor setup Change-Id: I43faeb72f6678a653feef82e0f0deaf87c039e9c Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128958 Reviewed-by: Eike Rathke <erack@redhat.com> Tested-by: Jenkins
2022-01-26tdf#146970 use a textview to show the proposed replacement formulaCaolán McNamara
so it can scroll and keep the text area to some finite size no strings changes or new translations, so this is a backportable change. Change-Id: Ibc36812c6aa7fff8e035c06d0dad885909f057ff Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128933 Tested-by: Jenkins Reviewed-by: Caolán McNamara <caolanm@redhat.com> (cherry picked from commit 36121e87b7e2a2defbfe75dfdadb2dde4fff6b4f) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128948 Reviewed-by: Eike Rathke <erack@redhat.com>
2022-01-25ScInputHandler::UpdateFormulaMode() also for multi-line formulasEike Rathke
Additionally sprinkle some TODO: MLFORMULA comments on places where multi-line formulas should be supported as well but code currently only works for one paragraph. Change-Id: If2630a6493910984e0abb81ff2ed42ed51394681 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128787 Tested-by: Jenkins Reviewed-by: Eike Rathke <erack@redhat.com> (cherry picked from commit 8204d322379cbc054fb834911d88fb0fe3b040c6) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128822 Reviewed-by: Michael Weghorn <m.weghorn@posteo.de>
2022-01-25tdf#146571 Ignore focus changes after closing Manage changes dialogCaolán McNamara
Change-Id: Ib588310410a265acdd129a8e51adf56bb5973bd9 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128869 Tested-by: Jenkins Reviewed-by: Caolán McNamara <caolanm@redhat.com> Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128910 Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
2022-01-24sc: fix VBA Copy-Paste using same and separate documentTomaž Vajngerl
This fixes 2 issues with VBA copy-paste: - VBA command Range(..).Copy issue where the range wasn't selected when copying and the copied cells were from the previous selection. The Copy command now does the same selection as the Cut command. - VBA PasteSpecial issue where the wrong view was used to get the clip document. - VBA Workbooks.Add issue where the new workbook wasn't activated after it was created, which causes an issue when running tests, but not when running in LO application. The Add command does now the same as the Workbooks.Open command. All the issues are supported by new test cases. Change-Id: I36ec45c01f18f7f76e4f95a25a28402a6ee0e2e2 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128720 Tested-by: Jenkins Reviewed-by: Tomaž Vajngerl <quikee@gmail.com> (cherry picked from commit 0c6ee963e1f089cb73e0c68a28af29d0f8d9e0df) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128678 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2022-01-22tdf#146367 reintroduce approxAdd() similar handling with last summandEike Rathke
... to tie result to 0 eliminating opposite sign small equalish fractions. This does not "fix" all interim values of the bug's sample document, which repeatedly adds single values in sequence using SUM() and should use the + operator instead to benefit from approxAdd() for each, but maintains the better accuracy of Kahan summation in the not near equal zero cases. And the dreaded SUM(0.1;0.2;-0.3) also works again resulting in 0 with all permutations. Change-Id: I67188a6abbdb98212f070166ad02319c3d477e28 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128451 Reviewed-by: Eike Rathke <erack@redhat.com> Tested-by: Jenkins (cherry picked from commit 730b8aba72356bb8ba0066a5517b1224a4f1e232) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128432 Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
2022-01-18tdf#136362 Remove unnecessary blank area and option in some image toolbarJeff Huang
Only in Calc, Draw and Impress. Change-Id: I944aad875fedd1dd19657c49b059e33fd6e9d3f9 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/127984 Tested-by: Jenkins Reviewed-by: Heiko Tietze <heiko.tietze@documentfoundation.org> (cherry picked from commit 0596ef354ee8d40976c6be044a646757240f7921) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128443 Reviewed-by: Po-Yen Huang <jeff@mail.ossii.com.tw> Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
2022-01-16tdf#136481 show the focus rect for the case nothing is yet selectedCaolán McNamara
but the cursor is at pos 0 Change-Id: I7c94480bffdced23fe89d4e3ba0e729742247130 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128414 Tested-by: Jenkins Reviewed-by: Caolán McNamara <caolanm@redhat.com> (cherry picked from commit 4bdf28b404f90236d93ce207671f755a71f28496) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128429 Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
2022-01-12tdf#146454 - Check for multiline string in setDataArrayAndreas Heinisch
Check for a multiline string in the setDataArray function. If that is the case, set a string or edit cell. Otherwise, just set a plain string in order to prevent performance issues due to the adjustments of the row height. Regression from b9e2c5d2a36fbd189c20448cadf4212edf02914d Change-Id: I9216b198b878fda5da29596d25504db853acd756 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/127660 Tested-by: Jenkins Reviewed-by: Andreas Heinisch <andreas.heinisch@yahoo.de> (cherry picked from commit 2d11710df6d81c03007d136667a443e09cc70eed) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128332 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2022-01-11tdf#145059: sc: always write dxf node for color filter for XLSXVasily Melenchuk
If there is no dxf reference to color filter (for example if selected color is not in range of available colors for current range) Excel is not able to show color filter correctly: it is not marked as used filter, no ability to reset, etc. Change-Id: I65378ddd6f8f8233cda147ff9dcd28f455a58225 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/127745 Tested-by: Jenkins Reviewed-by: Thorsten Behrens <thorsten.behrens@allotropia.de> (cherry picked from commit f0ad6ec2a23a36ade407db8cda36655ba2f144c1) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128159
2022-01-07tdf#146617: iterators may be invalidated during handler callMike Kaganski
... when the handler decides to remove itself from listeners. Change-Id: I1d5fe802f50d284bf70d747edaff0a19852c5a7d Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128054 Tested-by: Jenkins Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com> (cherry picked from commit e9dce4e8cbea04d6c6636a45f5a172e986714409) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/128074 Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
2022-01-07Resolves: tdf#146592 "Reset" to the current document propertiesCaolán McNamara
not those seen at the initial load of the TabPage Change-Id: Ieba4c278a0a9b7fef3030d8462cb845c27dba9b7 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/127959 Tested-by: Jenkins Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
2022-01-04For now, disable tests giving dubious results on AArch64Stephan Bergmann
when built with -ffp-contract enabled (-ffp-contract=on default on Clang 14 trunk since <https://github.com/llvm/llvm-project/commit/f04e387055e495e3e14570087d68e93593cf2918> "Making the code compliant to the documentation about Floating Point", cf. 19559ebbee160d1625d06feec7e6566772dad231 "Allow for a presumably more precise result of BESSELY(80,9)"; and -ffp-contract=fast default when building with optimizations on GCC) on both Linux (with 6cb20e0b298f41fe88984aebfe5454f936a0ae3a "Disable CppunitTset_sc_*_functions_test for linux_aarch64 for now, too" reverted) and macOS. Change-Id: Ie867d999173410ec9868bd14c0ee04bbba371920 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126727 Tested-by: Jenkins Reviewed-by: Stephan Bergmann <sbergman@redhat.com> (cherry picked from commit 72b25619960cdaa829c8ea10e27d9c0b20a2c26f) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/127784 Reviewed-by: Eike Rathke <erack@redhat.com>
2022-01-03CppunitTest_sc_subsequent_filters_test2: check format instead of stringXisco Fauli
Otherwise, this might fail. See 660e538ee77109bc3e7dff4491164d696e503abc < Disable broken part of ScFiltersTest2::testTdf126116 > Change-Id: I40534b70fc19ee163d9e6cb0c081fae40b81fdb9 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/127774 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2021-12-26Resolves: tdf#146377 Propagate condition error of IF(condition)Eike Rathke
Previsouly a given ELSE-path was executed as the condition is never TRUE on error, which wasn't significant unless the ELSE was an error evaluating function like IFERROR() or ISERROR() or other IS...() functions. This changes behaviour of an undocumented side effect, so *might* break existing expressions that (unwillingly?) relied on it. Gnumeric, Excel and GoogleSheets all agree on propagating the error value as implemented now. Change-Id: I0905a363be7a3b02925f1c41b4e6c8d36f12df93 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/127515 Reviewed-by: Eike Rathke <erack@redhat.com> Tested-by: Jenkins (cherry picked from commit 1c14b9efb0677dea65ff220222fbb8d5c2aa6973) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/127481
2021-12-25Related: tdf#146377 Let AND(), OR(), XOR() propagate the current error, if anyEike Rathke
... instead of forcing Err:504 Change-Id: I7e91671ffc61bf82d9095378656b41981040f55e Reviewed-on: https://gerrit.libreoffice.org/c/core/+/127492 Reviewed-by: Eike Rathke <erack@redhat.com> Tested-by: Jenkins (cherry picked from commit 4674ff347cd2e71445503efdc0c110d094daa80b) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/127476
2021-12-23tdf#143582: Avoid error on clearing leftover return value of a methodMike Kaganski
SbxValue::Clear happens to not error out on clearing string and some other types of objects, but for the rest, it calls SbxValue::Put, and the latter errors out if not CanWrite(). The original test implemented in commit 24d24debef4cda7de702c4b470a3707f1aae3ea3 only checked string return value, so happened to miss this problem. See similar code in SbiRuntime::FindElement. Change-Id: I7c17137cc9e7ee3133ee86a9f701559df66e53b6 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/127248 Tested-by: Jenkins Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com> (cherry picked from commit 822998f1dc202aba7a558767f641687aef0c1148) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/127257 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2021-12-23CppunitTest_sc_subsequent_filters_test2: fix test depending on current yearXisco Fauli
While at it, move the file from ods to fods for better tracking in git Change-Id: I16df221e3a7341dc7c53de6130158c3887c472a6 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/127361 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org> (cherry picked from commit f4ab9cf69c3b1e9f17fc3873f160ddebd5f812fb) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/127370 Reviewed-by: Eike Rathke <erack@redhat.com>
2021-12-23tdf#145057 Fix saving color filter when multiple data ranges in sheetSamuel Mehrbrodt
When mutliple data ranges existed in one sheet, only the first one was considered when exporting color filters. Consider all of them, as any could hold a color filter. Change-Id: I13ae2018057eef7ef24fc8298c814a93df24f74b Reviewed-on: https://gerrit.libreoffice.org/c/core/+/127328 Tested-by: Jenkins Reviewed-by: Samuel Mehrbrodt <samuel.mehrbrodt@allotropia.de> (cherry picked from commit b85e99950dc4584160512cffec303827c02f2d15) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/127259 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2021-12-20tdf#103331: Make styles attributes optional.Kohei Yoshida
Apply only those that are actually specified. Change-Id: Ib2e090fefe4dbfe3d4fca2b953bcf51d97d9ddec Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126901 Tested-by: Jenkins Reviewed-by: Kohei Yoshida <kohei@libreoffice.org> (cherry picked from commit 6b2da3ae3ea7f47dff3c807c151f88a9e1ae9964) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/127167 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2021-12-20tdf#146019 Remove confirmation dialog about applying images EXIF rotationGabor Kelemen
Change-Id: I53553f91eeb5bd56bbad19b80421177a84625d96 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126616 Tested-by: Jenkins Reviewed-by: Heiko Tietze <heiko.tietze@documentfoundation.org> Signed-off-by: Xisco Fauli <xiscofauli@libreoffice.org> Reviewed-on: https://gerrit.libreoffice.org/c/core/+/127151
2021-12-19ofz#42605 Null-dereferenceCaolán McNamara
Change-Id: Ief752f20aa3e7672e4ed7f6cd1809ee56a096c43 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/127051 Tested-by: Jenkins Reviewed-by: Caolán McNamara <caolanm@redhat.com>
2021-12-17tdf#58125 sc: don't show comment notifier for hidden columnsJustin Luth
This fixes a LO 3.6-ish regression. Things have changed a lot since then. I assume that pCell was empty when it was hidden. The test for pCell was removed in commit c06dbbe7594c2a0b5a5b19f8e183d9c421e6e094, which was in the range that bibisect suggested. Change-Id: I0af137358335a808de901111a71f5c113fabcf24 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/127001 Reviewed-by: Justin Luth <jluth@mail.com> Tested-by: Jenkins (cherry picked from commit b58dbb862296b63f3f2395c59d30fba40fa5f59f) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126975 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2021-12-16Update liborcus to 0.17.2.Kohei Yoshida
Change-Id: I76c0d57da63c1e35f80b13071793dbbb27cb218a Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126655 Tested-by: Jenkins Reviewed-by: Kohei Yoshida <kohei@libreoffice.org> (cherry picked from commit aadbac5467bb6ab768f87ed6ec003c55159d54aa) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126886
2021-12-16Resolves: tdf#126565 RefEditModified() preset current sheet for address-onlyEike Rathke
Change-Id: Ifa662a9d55fae10e3e1e3e115c9c5eb10972cbbe Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126899 Reviewed-by: Eike Rathke <erack@redhat.com> Tested-by: Jenkins (cherry picked from commit ec4f8cd2172bf7b232961f31ebd22d7e13af261e) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126882
2021-12-14Related: tdf#132466 Relative address parsing needs current positionEike Rathke
For the bug doc example the #C target lead to always column 1 (A,0) being selected. Instead, the column of the current cell cursor position should be selected as C is the short notation for C[0] in R1C1 address syntax. This does not "fix" the alleged bug but would yield the desired behaviour if there was no sheet C where the intention was to jump to. Change-Id: Ibaed0250b9fb2cbc7f15a4b22404c48b39d41ff8 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126496 Reviewed-by: Eike Rathke <erack@redhat.com> Tested-by: Jenkins (cherry picked from commit 155024040de888270bf90bdd200799cea7d73e4e) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126825
2021-12-14make sure text operations are not queried by (numeric) valueLuboš Luňák
E.g. SC_CONTAINS is, according to isPartialTextMatchOp(), a text-only operation, so query it as such and not as a numeric value. This fixes/allows e.g. substring queries on dates. Change-Id: I6c612d9934193828b7a7eabed92f2bfeb385e5a0 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126767 Tested-by: Jenkins Reviewed-by: Eike Rathke <erack@redhat.com> (cherry picked from commit 0d1971a8dc1f7ce24f67abcab4d6af9cf2b7b823) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126816
2021-12-14tdf#124412 - Set actual condition for an inactive date entryAndreas Heinisch
In the conditional format screen, set the actual condition for an inactive date entry instead of "Date is today" in all cases. Change-Id: I3d3f05c35a9830a1e82d68521e80e88c99b43a90 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126059 Tested-by: Jenkins Reviewed-by: Andreas Heinisch <andreas.heinisch@yahoo.de> (cherry picked from commit dfdf33a34729de7f256ed396be93fb96b4844404) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126238 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2021-12-13related tdf#92010 sc autoformula: ensure proper sorting orderJustin Luth
When typing "=a", autoformula was suggesting ACCRINT,ACCRINTM,ACOS instead of starting with ABS. [This has been true since these 3 suggestions started in LO 5.2.] Prior to this patch, the very first item in the sorted list (ABS) was placed at the end of the suggestion vector. That is because the loop immediately increments. Since the given initialization value is end(), it set begin() as the starting loop value and then immediately incremented to item 2. Item 1 was finally evaluated last, putting ABS after ZTEST. The backwards loop handled this properly, so do the same thing for the forward loop. Change-Id: I539c749ea43140a1480d74471787bc886dda671e Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126723 Tested-by: Jenkins Reviewed-by: Justin Luth <jluth@mail.com> Reviewed-by: Eike Rathke <erack@redhat.com> (cherry picked from commit b2fb1631ab5b3c1d9d09aefa0d3a81307e7ffa35) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126750
2021-12-13tdf#92010 sc autoFormula: prefer MRU formulasJustin Luth
Prior to 4.3, only MRU formulas were auto-completed. Then it changed to suggest all formulas - alphabetically for the most part. That 4.3 commit 5b0b7553241bb5150b12bbf7625b4b0b36970272 completely removed all reference to MRU. But it makes sense to prefer an MRU over a strictly alphabetical match. This patch depends on LO 7.4 patch "new ScTypedStrData: typically missed argument in CTOR" Change-Id: Id5d860d1401693f43833719977d1c1e4c386385c Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126499 Tested-by: Jenkins Reviewed-by: Justin Luth <jluth@mail.com> Reviewed-by: Eike Rathke <erack@redhat.com> (cherry picked from commit b93a6964e0466c1b67d2c233040357a8f6c75214) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126749
2021-12-13Remove dubious code in ui fileXisco Fauli
Introduced in f773e7aa8c96e070085ece95889f02590ed65f89 < Resolves tdf#144227 - Default command to switch UI > Change-Id: I1975d6d61a2fd225e87dffe65708a1f5e531f106 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126735 Reviewed-by: Heiko Tietze <heiko.tietze@documentfoundation.org> Tested-by: Xisco Fauli <xiscofauli@libreoffice.org> (cherry picked from commit 29bbed22e404700f9538e244df6e02681da1fc1b) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126747 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2021-12-13lok: don't crash accessing an invalid ObjectContact cacheDennis Francis
Fix description: Avoid storing a reference to the object-contact (sdr::contact::ObjectContact) of the page-window (SdrPageWindow) related to ScDrawView in the "proxy" object-contact. In the current setup there is no way to invalidate the proxy object when the original object-contact and its page-window are destroyed (in this case it seems during a sheet-switch). Instead query the real object contact just in time when the grid offsets are requested from the proxy object-contact. Performance: Behaviour of offset computation is not affected. It is still cached in the "real" object contact and it is only computed after it gets invalidated (because of change in zoom) Crasher reproduction(LOK): 1. Create a spreadsheet with two sheets - one with a table of texts and shapes and other may be empty. 2. In first sheet select a single row(via header) and press Ctrl+X. 3. Go to the empty sheet, and paste (Ctrl+V) and immediately save using Ctrl+S. ---Relevant part of backtrace----------------- <signal handler called> 0x00007ff96781cc70 in main_arena () from /lib64/libc.so.6 0x00007ff964f2f47b in sdr::contact::ViewObjectContact::getGridOffset (this=0x6f01f20) at /opt/libreoffice/co-2021/svx/source/sdr/contact/viewobjectcontact.cxx:456 0x00007ff95cffac5a in (anonymous namespace)::ScLOKProxyObjectContact::calculateGridOffsetForViewOjectContact (this=0x6fb00a0, rTarget=..., rClient=...) at /opt/libreoffice/co-2021/sc/source/ui/view/gridwin4.cxx:1315 0x00007ff964f2f493 in sdr::contact::ViewObjectContact::getGridOffset (this=this@entry=0x5e0c5e0) at /opt/libreoffice/co-2021/svx/source/sdr/contact/viewobjectcontact.cxx:459 0x00007ff964f30732 in sdr::contact::ViewObjectContact::getPrimitive2DSequence (this=0x5e0c5e0, rDisplayInfo=...) at /opt/libreoffice/co-2021/svx/source/sdr/contact/viewobjectcontact.cxx:364 0x00007ff964f30a82 in sdr::contact::ViewObjectContact::getObjectRange (this=0x5e0c5e0) at /opt/libreoffice/co-2021/svx/source/sdr/contact/viewobjectcontact.cxx:198 0x00007ff964f30d00 in sdr::contact::ViewObjectContact::ActionChanged (this=0x5e0c5e0) at /opt/libreoffice/co-2021/svx/source/sdr/contact/viewobjectcontact.cxx:220 0x00007ff964f20294 in sdr::contact::ViewContact::ActionChildInserted (this=0x5dd83a0, rChild=...) at /opt/libreoffice/co-2021/svx/source/sdr/contact/viewcontact.cxx:180 0x00007ff96506628a in SdrObjList::impChildInserted (rChild=...) at /opt/libreoffice/co-2021/svx/source/svdraw/svdpage.cxx:288 0x00007ff95c808bcd in ScDocument::CreateAllNoteCaptions (this=<optimized out>) at /opt/libreoffice/co-2021/sc/source/core/data/document.cxx:6614 ... 0x00007ff95cbedcc5 in ScXMLImportWrapper::Export (this=this@entry=0x7fff5432f110, bStylesOnly=bStylesOnly@entry=false) at /opt/libreoffice/co-2021/sc/source/filter/xml/xmlwrap.cxx:730 0x00007ff95ccfd896 in ScDocShell::SaveXML (this=0x5c4c330, pSaveMedium=<optimized out>, xStor=...) at /opt/libreoffice/co-2021/sc/source/ui/docshell/docsh.cxx:556 0x00007ff95cd009c7 in ScDocShell::SaveAs (this=0x5c4c330, rMedium=...) at /opt/libreoffice/co-2021/sc/source/ui/docshell/docsh.cxx:1801 ... 0x00007ff95d081af4 in ScTabViewShell::ExecuteSave (this=0x5e9b100, rReq=...) at /opt/libreoffice/co-2021/sc/source/ui/inc/viewdata.hxx:354 ----------------------------------------- Change-Id: I00eac440546624bc448dcd30499957dea7c1de87 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126471 Tested-by: Jenkins Reviewed-by: Dennis Francis <dennis.francis@collabora.com> (cherry picked from commit ff175d858e897964d81cf3b7edaa6ccde32e098b) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126519
2021-12-13tdf#121699 sc: wrap LOKit stuff with isActiveJustin Luth
This is a partial revert of LO 6.1 commit 009d2756b1678477ec23d5647bd5004c4bff3a62. It is one thing to change code that you don't really know why it is setup that way, but it is another thing to ignore the resulting regression reports for 3 years. Not impressed. It look me only 1 hour to work through this, and this is completely outside of my normal line of programming. Change-Id: I8671677f38cc1332c6c5a76847801e2680097048 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126500 Tested-by: Jenkins Reviewed-by: Henry Castro <hcastro@collabora.com> Reviewed-by: Justin Luth <jluth@mail.com> (cherry picked from commit a99aea61c3199dc8188d4d873a10201082dff2e2) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126686 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2021-12-11tdf#128895 sc xmlimport: create enough dynamic cols if propsJustin Luth
Since LO 6.3, only 64 columns are created by default, where previously it was a fixed 1024. A common user practice is to hide all columns not used, but this collapsed property was lost because only part of the columns were actually created and thus exported. In this example, import specifies 1017 hidden columns (H-AMJ), but since only 64 columns are created, export only specified 57. So ensure that on import, any column with defined properties is created - even if they don't contain any content. Change-Id: If928880baf5585613715a1f4361a9059584d1ad2 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126540 Tested-by: Jenkins Reviewed-by: Justin Luth <jluth@mail.com> Reviewed-by: Bartosz Kosiorek <gang65@poczta.onet.pl> (cherry picked from commit 297ab561c6754f89326a1e8ce1751233669578d7) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126668
2021-12-10tdf#133353, Make 2nd argument optional for DCOUNT and DCOUNTA.Winfried Donkers
The functions accept empty second arguments, but the function wizard and hints stated that it was required. Change-Id: I74fcfcc31492ed776085d1bc6ee6a9ff22a87818 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126620 Tested-by: Eike Rathke <erack@redhat.com> Reviewed-by: Eike Rathke <erack@redhat.com> (cherry picked from commit 885465bc0ae6029d8adad57015fb0ff6c26a6ba6) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126594 Tested-by: Jenkins Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2021-12-10new ScTypedStrData: typically missed argument in CTOR, tdf#142910 follow-upJustin Luth
Pretty much any attempted use of eType was completely wrong and lost. Regression from commit f6b143a57d9bd8f5d7b29febcb4e01ee1eb2ff1d CommitDate: Wed Jul 7 17:44:46 2021 +0200 tdf#142910 sc filter: fix "greater than" or "smaller than" etc Most calls to this are missing the "rounded number" argument, so the enumator is actually accepted as the double fRVal, and the StringValue eType was left as the default value (Standard), instead of the intended enumerator. 0.0 looks too much like 0, 0 to even notice in casual code reading. This had rendered the type mostly irrelevant. Change-Id: If4fa69d4b3077981244a2c3a785f80b77f9f9501 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126453 Tested-by: Eike Rathke <erack@redhat.com> Reviewed-by: Eike Rathke <erack@redhat.com> (cherry picked from commit ac1ea5944a236efef75af7c85d0ffe10d374e7fa) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126592 Tested-by: Jenkins
2021-12-09tdf#145671 Don't open URL twice in cell edit modeSamuel Mehrbrodt
Change-Id: I8a324d6a037c8a5c0efc0b2825657513c2f9841c Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126159 Tested-by: Jenkins Reviewed-by: Samuel Mehrbrodt <samuel.mehrbrodt@allotropia.de> (cherry picked from commit b9362e87a3646a6693b81fb0d40e28ce7813a251) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126441 Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
2021-12-09tdf#144107: fix incorrect conditioning on bGrid...Dennis Francis
introduced in 3092322bb83998397b09d3f34b6d04b5e5c1da50 The intention was to avoid the default line drawing in the original else branches to be conditional on bMergeCover. The previous patch used 'bGrid' as a surrogate for the original 'else' case - which is wrong. Change-Id: I7c6b49fa9c21064312273b1051ffc7f81de3fef3 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126076 Tested-by: Jenkins Reviewed-by: Dennis Francis <dennis.francis@collabora.com> (cherry picked from commit 307c1f13821386687b8cf3c560ff2570b89ac3a9) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126134
2021-12-08Related: tdf#144410 excessive "resize" triggeredCaolán McNamara
so changing cursor pos becomes hit or miss in the dropdown arrow shown case Change-Id: I6e08a022041667fe8368b2cb09289a598957eb87 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126511 Tested-by: Jenkins Reviewed-by: Caolán McNamara <caolanm@redhat.com>
2021-12-07signed-integer-overflowStephan Bergmann
...after 8406139062d9ffe1daed32aefe4e261c6c55d63e "process broadcasts for adjacent cells together (tdf#119083)" during CppunitTest_sc_bugfix_test, > sc/source/core/tool/grouparealistener.cxx:340:72: runtime error: signed integer overflow: 2147483647 + 1 cannot be represented in type 'int' Change-Id: Ica29fc4c204cd6fd5c52c61271ebfcc9691f7f6f Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126458 Tested-by: Jenkins Reviewed-by: Stephan Bergmann <sbergman@redhat.com> (cherry picked from commit 70fc728ceef4dcf45dd97dc9051ddb73ff734169) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126437
2021-12-07signed-integer-overflowStephan Bergmann
...after 8406139062d9ffe1daed32aefe4e261c6c55d63e "process broadcasts for adjacent cells together (tdf#119083)" during e.g. CppunitTest_sc_ucalc_copypaste, > sc/source/core/data/documen7.cxx:143:59: runtime error: signed integer overflow: 2147483647 + 1 cannot be represented in type 'int' Change-Id: If848747f6b004219475b39b48c6a3de086bd1ae7 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126457 Tested-by: Jenkins Reviewed-by: Stephan Bergmann <sbergman@redhat.com> (cherry picked from commit 8bc85cd8a6a9a879c143d8defacc45967a0ac99e) Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126436
2021-12-06process broadcasts for adjacent cells together (tdf#119083)Luboš Luňák
The problem in tdf#119083 is that it sets up listening for the range of cells used by VLOOKUP, and when adding a new column changes all those cells, it results in repeated SfxHintId::ScDataChanged broadcasts to the cells with the VLOOKUP formula. This commit makes ScHint include a row count, making it possible to group adjacent rows for SfxHintId::ScDataChanged together and send notifications just once for the range. Change-Id: Ib3439de58a2b1e5e8f01b037a62608e38b8e9125 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126395 Tested-by: Jenkins Reviewed-by: Luboš Luňák <l.lunak@collabora.com> Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126448
2021-12-06make it clearer that ScHint is modifiedLuboš Luňák
It's somewhat confusing that an accessor is provided to give a reference to internal data and then the object is modified indirectly using the reference. It appears to be only for performance reasons, so I thought that inlining the ctor and ctor could help the compiler to optimize this, but apparently it can't move this outside of the loop, so at least make it clearer. Change-Id: I72cf15d1446daa559ac4079b9478e53694d7d198 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126394 Tested-by: Jenkins Reviewed-by: Luboš Luňák <l.lunak@collabora.com> Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126447
2021-12-06try to broadcast in bulk (tdf#119083)Luboš Luňák
This somewhat improves the situation, but still a lot of time is spent repeatedly making the same cells dirty. Change-Id: I3444ca6dac4b3a40c6ac40a0b70cf5baeefa2f09 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126391 Tested-by: Jenkins Reviewed-by: Luboš Luňák <l.lunak@collabora.com> Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126426
2021-12-05fix caching of ScQueryItem values for multiple entries (tdf#146037)Luboš Luňák
The code assumed that there would be only one ScQueryEntry. Change-Id: Idf4cc6bdbbf1edad7d13eb9a9643bb67e199dd01 Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126368 Tested-by: Luboš Luňák <l.lunak@collabora.com> Reviewed-by: Luboš Luňák <l.lunak@collabora.com> Reviewed-on: https://gerrit.libreoffice.org/c/core/+/126300 Tested-by: Jenkins Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
generated by cgit (git 2.34.1) at 2025-03-14 06:19:01 +0000