summaryrefslogtreecommitdiff
path: root/sc
AgeCommit message (Collapse)Author
2018-02-09CheckLinkFormulaNeedingCheck() for named expressionsEike Rathke
This is a combination of 3 commits. CheckLinkFormulaNeedingCheck() for .ods named expressions This is specifically necessary for named expressions that are used in conditional format formulas, for which RPN is generated at a later stage, not during import. (cherry picked from commit eae9648e99be53ba441d9d8207aac6f3ce211ef2) CheckLinkFormulaNeedingCheck() for .xls named expressions (cherry picked from commit 8512f13c42ae3fbb287a555616fe10ff04295616) CheckLinkFormulaNeedingCheck() for .xlsx named expressions (cherry picked from commit a1f933ee2b9e23a505d937035821e9571cf4119c) Conflicts: sc/source/filter/oox/defnamesbuffer.cxx e03cb5767c34f8157a492a6d6c3b0700d065052d 217c89822ab477a6c383d170ae739e44efd10fa3 Change-Id: I54ab8dc14f81d6b18b0d17f448187d19d8e396fc Reviewed-on: https://gerrit.libreoffice.org/48858 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Caolán McNamara <caolanm@redhat.com> Tested-by: Caolán McNamara <caolanm@redhat.com> (cherry picked from commit 908854a7b281454332af434be9468ec45d420030)
2018-02-09CheckLinkFormulaNeedingCheck() for conditional format expressionsEike Rathke
This is a combination of 4 commits. Prepare CheckLinkFormulaNeedingCheck() to use either RPN or tokenized code Conditional format formulas aren't finally compiled until needed so the check will have to operate on the tokenized expression instead of RPN code. (cherry picked from commit faa0305ba3d0dc698fce4915d4f3a1fb52422380) CheckLinkFormulaNeedingCheck() for .ods conditional format expressions (cherry picked from commit 2930ba2ac5d9423f2848b968edcd8ddc71966186) CheckLinkFormulaNeedingCheck() for .xlsx conditional format expressions (cherry picked from commit fef24d9f999ee54d7936900485d97ff26656f517) CheckLinkFormulaNeedingCheck() for .xls conditional format expressions (cherry picked from commit af2a2a0c72db312902e466c36697b5c198041e82) 45eb1ab5efa0ec9da2663f20427d2474ce300826 31ede1a23223a798141a0891deeabd8cf88fff58 afa112cc591b411d80ead48bf726788d361f6eb3 Change-Id: I68837e9bd33f125ab47b10b1a6fa18175abd1627 Reviewed-on: https://gerrit.libreoffice.org/48719 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Caolán McNamara <caolanm@redhat.com> Tested-by: Caolán McNamara <caolanm@redhat.com> (cherry picked from commit dc44111ad5965bf4179fc654b677e1e445dea2f0)
2018-02-09CheckLinkFormulaNeedingCheck() for .xls and .xlsx formula cellsEike Rathke
This is a combination of 3 commits. Move implementation to CheckLinkFormulaNeedingCheck() for further reuse (cherry picked from commit 55e484c7bcd3ef218e08d3fd93f97bf98fd8cb7f) CheckLinkFormulaNeedingCheck() for .xlsx cell formulas (cherry picked from commit f96dbc3dd9c33202f75e29ef49d962386595995d) CheckLinkFormulaNeedingCheck() for .xls cell formulas (cherry picked from commit 6bc48275558c3f76c4da25eb8af3c48583ac5599) a6dd195f7eb4d43483e87eeca59f651e7bf2dcb8 2587fbc4fec39b6f2c8e733331815a2953dee308 Change-Id: I541d2b6e12a88371c064b901b00e71206ee0c18e Reviewed-on: https://gerrit.libreoffice.org/48143 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Caolán McNamara <caolanm@redhat.com> Tested-by: Caolán McNamara <caolanm@redhat.com> (cherry picked from commit 2bef4debcf7650f3b3922134dff0332d4a95da3f)
2018-02-09handle ocWebservice similarly to ocDdeCaolán McNamara
might have too much in here seeing as we don't need to worry about ocWebservice calling into itself Reviewed-on: https://gerrit.libreoffice.org/47819 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Eike Rathke <erack@redhat.com> (cherry picked from commit e1946d75a1095c2596d7815600454ff01fcd3270) Signed-off-by: Andras Timar <andras.timar@collabora.com> Change-Id: I0145f38cc1c1f9ff514a496f7101d81cde9e7c67
2018-02-09Better handle ScDde formulas with missing dde-link entriesCaolán McNamara
typically each ScDde formula has a matching table:dde-link which results in a ScDdeLink getting inserted during the load. If that dde-link is missing then no ScDdeLink exists and ScDde() will create a new one without cached content. So detect that ScDde is used in the freshing loaded ods and defer fetching new content until the right time. only call GetHasMacroFunc to set SetHasMacroFunc and bHasMacroFunc is not accessed any other way, so this is an oxbow Reviewed-on: https://gerrit.libreoffice.org/47757 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Caolán McNamara <caolanm@redhat.com> Tested-by: Caolán McNamara <caolanm@redhat.com> (cherry picked from commit b0597ba5d745974fce752e1b677451a19350d351) Reviewed-on: https://gerrit.libreoffice.org/47818 Reviewed-by: Eike Rathke <erack@redhat.com> (cherry picked from commit 4ede45eb239b1604bca900c22481b7d13e4c2790) Signed-off-by: Andras Timar <andras.timar@collabora.com> Change-Id: I016b53288076d83dd49e92e245346a5f7f560522
2018-02-09limit WEBSERVICE to http[s] protocolsCaolán McNamara
and like excel... 'For protocols that aren’t supported, such as ftp:// or file://, WEBSERVICE returns the #VALUE! error value.' Change-Id: I0e9c6fd3426fad56a199eafac48de9b0f23914b3
2017-11-09Resolves: tdf#111428 swap ScColumn::mnBlkCountFormulaEike Rathke
This is a combination of 2 commits. (cherry picked from commit 423df1fa929784c14e3a133c06468589fe9269cd) (cherry picked from commit 911e2aff3cc37cb7410292728ffea05fffbfb0b3) CellStoreEvent remembered the original ScColumn::mnBlkCountFormula, hence after inserting a column to the left the quick check of ScColumn::HasFormulaCell() whether there are any formula cells worked on the swapped in count (originally the one to the right) that happens to be empty in the scenario. Things worked correctly by accident if the next column to the right already contained a formula cell. c44fed96c49bea7365bf1200e06788860966795c Change-Id: If993856ceee657736f516a81c293506041a6b7eb Reviewed-on: https://gerrit.libreoffice.org/44210 Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com> Tested-by: Eike Rathke <erack@redhat.com> Reviewed-on: https://gerrit.libreoffice.org/44493
2017-10-04Resolves: tdf#112780 no ResetDirty() after SetHybridEmptyDisplayedAsString()Eike Rathke
Change-Id: Idea5c1bd8e8577eaf90fc47d39f88c79963c422b (cherry picked from commit df778416981ab02d42182e5c2e46dc09ba2e2a3c) Reviewed-on: https://gerrit.libreoffice.org/43048 Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com> Tested-by: Eike Rathke <erack@redhat.com>
2017-10-01Resolves: tdf#100822 use sax::Converter::parseDateTime() for 'd' ISO 8601 dateEike Rathke
Instead of squeezing the string through the locale dependent number parser (which should force '.' Time100SecSeparator for 'T' ISO 8601 but apparently doesn't yet) that is a bottle neck anyway. Change-Id: I70f74e680322a715c9cc658c8be033620d9798e3 (cherry picked from commit 5374c04a89e1720490a3625dfd95406c2d60c0cd) Reviewed-on: https://gerrit.libreoffice.org/42971 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com>
2017-09-29Resolves: tdf#103734 propagate error from matrix to MIN()/MAX()Eike Rathke
(cherry picked from commit 9e694c747954078442d47d3d7bd1d4db283b96ff) Conflicts: sc/source/core/tool/interpr1.cxx Backported. Change-Id: I1ebc5baf4957ef9e3d1477b803cf7fee02754360 Reviewed-on: https://gerrit.libreoffice.org/41886 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com>
2017-09-29Resolves: tdf#112523 do not use anything Ascii in UI contextEike Rathke
Change-Id: I0854f7b1e6e0bf51d8b4c7d846fad9b0ff25e763 (cherry picked from commit 86910087fa41956aad6a22943bfbeb158e205dcd) Reviewed-on: https://gerrit.libreoffice.org/42872 Reviewed-by: Michael Stahl <mstahl@redhat.com> Tested-by: Michael Stahl <mstahl@redhat.com>
2017-09-27tdf#86083 Restore old operation buttons but have them hiddenYousuf Philips
Change-Id: I7d006d3d5e22ade59ddfe46f21fbc777a1c75dac Reviewed-on: https://gerrit.libreoffice.org/42684 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Yousuf Philips <philipz85@hotmail.com> (cherry picked from commit 7dc82cc2e424b6acddb1aa06338f51db20e6a21e) Reviewed-on: https://gerrit.libreoffice.org/42687 Reviewed-by: Eike Rathke <erack@redhat.com>
2017-09-24tdf#112005 xmloff: ODF export: export embedded object text *first*Michael Stahl
... before adding any attributes, so that the attributes don't erroneously get added to the paragraph element. Also fix the problem that if you set "ODF format version" to 1.1 or 1.2, export asserts in SvXMLNamespaceMap::GetQNameByKey() because the "loext" namespace isn't registered, and creates an invalid <p> element. Change-Id: I637f9d3ff746d877ced480e35ef53f4545a06a4b (cherry picked from commit 86f256596c8566e80993e1cf6035bc3534b6f816) Reviewed-on: https://gerrit.libreoffice.org/42569 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Caolán McNamara <caolanm@redhat.com> Tested-by: Caolán McNamara <caolanm@redhat.com>
2017-09-18ofz#3412 SYLK import: check ;X;Y;C;R col/row validity earlyCaolán McNamara
Change-Id: I91fcd2571e528201e01467f3bcdbbff30cdfb50c Reviewed-on: https://gerrit.libreoffice.org/42426 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Eike Rathke <erack@redhat.com>
2017-09-14ofz#3362 SYLK import: check ;X;Y;C;R col/row validity earlyEike Rathke
Change-Id: I37d5ce67f975b6b89c4b8a9baefae2467da2eb84 (cherry picked from commit 34ac0f9a0376b43bcff78a49ccaf4caa34c8c990) Reviewed-on: https://gerrit.libreoffice.org/42278 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Caolán McNamara <caolanm@redhat.com> Tested-by: Caolán McNamara <caolanm@redhat.com>
2017-09-07tdf#111974: Ensure sufficient space for all sheets in FormulaBuffer...Dennis Francis
and not just for the number of sheet fragments. This is important because we changed the internal storage of FormulaBuffer from based on std::map to std::vector in d17a83fa549f828f29e6939b16ba8b568a75f95e Added simple unit test that loads the bugdoc in subsequent_filters-test.cxx Reviewed-on: https://gerrit.libreoffice.org/41792 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Dennis Francis <dennis.francis@collabora.co.uk> (cherry picked from commit 20169423d492b58ab93fb0e8e9d2918c0baebc92) Reviewed-on: https://gerrit.libreoffice.org/41841 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Eike Rathke <erack@redhat.com> Conflicts: sc/qa/unit/subsequent_filters-test.cxx Backported. Change-Id: I7320dd537266e540d585d3fccfbb8665bba457c7 Reviewed-on: https://gerrit.libreoffice.org/42057 Reviewed-by: Eike Rathke <erack@redhat.com> Tested-by: Jenkins <ci@libreoffice.org>
2017-08-22Resolves: tdf#111943 really really limit the match, tdf#108292 follow-upEike Rathke
getRemainingCount() could deliver a wrapped around overflow value if mnIndex was already greater than the end index, which could happen if when/for non-matching larger block sizes were added, and if then a match was found behind those blocks a non-requested/unexpected index was returned, which in turn led to the assert() being hit in ScInterpreter::CalculateLookup(). In non-debug could result in an invalid block position access. This happened with the bug case document of tdf#111943 which in master can be loaded. Also, the start and end index are not dynamic and don't have to be recalculated each time, so make them const; column argument values are unused after. (cherry picked from commit 25b3806ac509006573e669acc33643af3bd77380) Change-Id: Ic294cade4e8e7828bee394e5ade61d7127be6bbb Reviewed-on: https://gerrit.libreoffice.org/41397 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Christian Lohmaier <lohmaier+LibreOffice@googlemail.com>
2017-08-18tdf#109104: respect RFC3986 for newlines in ScEncodeURLJulien Nabet
Quotation of RFC3986: A percent-encoded octet is encoded as a character triplet, consisting of the percent character "%" followed by the two hexadecimal digits representing that octet's numeric value So test the length of the return of OString::number and add "0" if needed ScEncodeURL was added with: https://cgit.freedesktop.org/libreoffice/core/commit/?id=25434372bf56e0ebdb7e7d47ab3c14c68211509f Thank you to Bele (the bugtracker reporter) for code pointer! Change-Id: I8df102eb38b31933c6ebb15bb25c125b423f722b Reviewed-on: https://gerrit.libreoffice.org/41086 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Eike Rathke <erack@redhat.com> (cherry picked from commit dabba2e3368c2e2ae4ab03ddcfc667e13f89841d) Reviewed-on: https://gerrit.libreoffice.org/41279
2017-08-07ofz#2867 null derefCaolán McNamara
Reviewed-on: https://gerrit.libreoffice.org/40792 Reviewed-by: Caolán McNamara <caolanm@redhat.com> Tested-by: Caolán McNamara <caolanm@redhat.com> (cherry picked from commit 42b894f80a6d0c39bb0f7092eb204a15c22c4f38) Change-Id: If856473683685d79d88b024f7fafa2920b403bb7 Reviewed-on: https://gerrit.libreoffice.org/40794 Reviewed-by: Michael Stahl <mstahl@redhat.com> Tested-by: Jenkins <ci@libreoffice.org>
2017-08-07Resolves: ofz#2833 null derefCaolán McNamara
Change-Id: I021a716aa76d430a1d3c6fac2dddec4daa01e563 Reviewed-on: https://gerrit.libreoffice.org/40669 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Michael Stahl <mstahl@redhat.com>
2017-08-02Related: tdf#109856 redo 'insert multiple tabs after' inserts too many tabsCaolán McNamara
Change-Id: I50e8941408343f8fa9acf017d7c195279952916d Reviewed-on: https://gerrit.libreoffice.org/40628 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com>
2017-08-02Resolves: tdf#109856 undo 'insert multiple tabs after' doesn't workCaolán McNamara
Change-Id: I534536358b2a949667c1c258bf27e13763c0de57 Reviewed-on: https://gerrit.libreoffice.org/40622 Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com> Tested-by: Markus Mohrhard <markus.mohrhard@googlemail.com>
2017-07-28Resolves: tdf#106872 only request selected shapesCaolán McNamara
that way we can avoid the super slow code path for filtered rows when we only care about selected shapes Change-Id: I175fa841e406dbbe7075296f2e0a0e79fa115fb7 Reviewed-on: https://gerrit.libreoffice.org/40496 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Eike Rathke <erack@redhat.com> (cherry picked from commit 192d97cdf091af08a492416824918ea447bfb16f) Related: tdf#106872 factor out getting selected shapes Change-Id: I765c482a41e9681a1eb145c1833cc94f35a27db3 (cherry picked from commit 221dae68df80298e81e6e6549636f3528f5c8bc3) Reviewed-on: https://gerrit.libreoffice.org/40519 Reviewed-by: Eike Rathke <erack@redhat.com> Tested-by: Jenkins <ci@libreoffice.org>
2017-07-28ofz: survive missing macro supportCaolán McNamara
Change-Id: Ica9c66fe09f7340f76f62e536527dc63b3735d90 (cherry picked from commit aa529a1957fce324c500753039ae7766b8dcf6a6) Reviewed-on: https://gerrit.libreoffice.org/40523 Reviewed-by: Eike Rathke <erack@redhat.com> Reviewed-by: Caolán McNamara <caolanm@redhat.com> Tested-by: Caolán McNamara <caolanm@redhat.com>
2017-07-27Related: tdf#106872 amount of a11y selection listeners constantly growsCaolán McNamara
so it gets progressively slower Change-Id: Ib53c69231c902d064b939be096e0dbeab2f0fc71 Reviewed-on: https://gerrit.libreoffice.org/40493 Reviewed-by: Eike Rathke <erack@redhat.com> Tested-by: Jenkins <ci@libreoffice.org>
2017-07-21Resolves: tdf#75650 reset boolean number format to General for new formulaEike Rathke
Change-Id: Iec2feefcbb5a92c64f182d93b4055fc009dbe1e5 (cherry picked from commit 1c1783864a356a4394f80dec28db939f4cf572cd) Reviewed-on: https://gerrit.libreoffice.org/40053 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com>
2017-07-21Blind fix tdf#108612 explicitly checking for and against clipboard documentEike Rathke
This is a combination of 2 commits. Attempt to blind fix tdf#108612 explicitly checking for clipboard document It seems that for some yet unknown reason in an (un)certain constellation ScDocument::IsClipboardSource() returns true because ScModule::GetClipDoc() does not return the clipboard document but the actual document instead, so destroying the clipboard document executes ScDocument::ClosingClipboardSource() and (or in earlier versions) ScDocument::ForgetNoteCaptions() on the actual document instead. Check also that ScDocument::bIsClip is set, hopefully curing the symptom, the real cause is something else. (cherry picked from commit d5020f35aec54f0241fa58557dc6caadc149f5a9) Another attempt to blind fix tdf#108612 Oddly in that scenario on Windows there are two clipboard documents alive when destroying one, so the destroyed one erroneously is said to be a clipboard source and then detachs the comments' caption information. There's no legit case where a clipboard document could actually be the real original source of another clipboard document, so bail out if the caller already is one. (cherry picked from commit 52e09f2c03f7cc024b2973c4806283c324fc23df) 0f8d240d8ec13bf91b77176dc6439a541cf14eaa Change-Id: I5ffd0706592d2737808e7004630a56c53115295b Reviewed-on: https://gerrit.libreoffice.org/40251 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com>
2017-07-14tdf#108758 : do not write to undo document when...Dennis Francis
...updating references during a block move, for formula cells that are in the target range of the move operation. The fix is for formula cells that are not grouped. For the grouped case, it was already doing correctly. Added two unit tests in ucalc_formula.cxx for grouped formula and non-grouped formula cases. Reviewed-on: https://gerrit.libreoffice.org/39883 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Eike Rathke <erack@redhat.com> (cherry picked from commit 2caaf9bfe8d3f81517467daf36de0bd8ddd6b543) Conflicts: sc/qa/unit/ucalc.hxx sc/qa/unit/ucalc_formula.cxx Backported, excluding conflicting unit tests. Change-Id: I9f4d988f5e154f56670bd1c0cc366ee6704fb858 Reviewed-on: https://gerrit.libreoffice.org/39922 Reviewed-by: Eike Rathke <erack@redhat.com> Tested-by: Jenkins <ci@libreoffice.org>
2017-07-09tdf#79304 - Handle CrossedOutItems in HTML export.Sean Stangl
Change-Id: I766fedb34737a1a7815693bf496fa08c08f492b1 Reviewed-on: https://gerrit.libreoffice.org/39719 Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com> Tested-by: Markus Mohrhard <markus.mohrhard@googlemail.com> (cherry picked from commit 2f2eba56d1f8ec5cdcadb4852e8856858477c008) Reviewed-on: https://gerrit.libreoffice.org/39722 Tested-by: Jenkins <ci@libreoffice.org>
2017-07-06tdf#108789 quick fix for 5.3 branch onlyJulien Nabet
Change-Id: Ib3a9475a6eec1551a77fb5b5994ec1073f275708 Reviewed-on: https://gerrit.libreoffice.org/39449 Reviewed-by: Lionel Elie Mamane <lionel@mamane.lu> Tested-by: Jenkins <ci@libreoffice.org>
2017-07-03handle SID_ENCRYPTIONDATA for shared password protected docs, tdf#56173Markus Mohrhard
Change-Id: Iadfe880b7be3186410a08568844d8812f176005d Reviewed-on: https://gerrit.libreoffice.org/39332 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com> (cherry picked from commit 669c12d59b6131f44945f166c77ae98df52ebb9e) Reviewed-on: https://gerrit.libreoffice.org/39337 Reviewed-by: Eike Rathke <erack@redhat.com>
2017-07-03always import page scaling, tdf#65848Markus Mohrhard
Change-Id: If5bad2336443280dd33d750b1b2749a06f2b5101 Reviewed-on: https://gerrit.libreoffice.org/39334 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com> (cherry picked from commit fef72f9623bac7f9aa50b4f2606e6ec6282cfe1f) Reviewed-on: https://gerrit.libreoffice.org/39340 Reviewed-by: Eike Rathke <erack@redhat.com>
2017-07-03pBASM can be nullptr in copy docs, e.g. during insert sheet from docMarkus Mohrhard
Found through http://crashreport.libreoffice.org/stats/crash_details/3bd264ba-ec12-476a-9bb4-e832caca3445 Change-Id: I0a4e6dddca61a2538c0fe6bbc51770a9886c464d Reviewed-on: https://gerrit.libreoffice.org/39411 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com> Reviewed-on: https://gerrit.libreoffice.org/39426 Reviewed-by: Eike Rathke <erack@redhat.com>
2017-07-02Resolves: tdf#108671 mark named expression single reference #REF! if deletedEike Rathke
Probably broken already since commit f77c9d5b3cb65d9c2e2417f60bec1113feac50e1 Date: Mon Jul 29 14:47:07 2013 -0400 Change-Id: If90a959d000a1bf5ab0804deb22d5cfad413ea5e (cherry picked from commit 6580e171ad054dca38f55822e67a400191cecdf5) Reviewed-on: https://gerrit.libreoffice.org/39072 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com>
2017-06-24tdf#40713: dBASE, don't lose dbt fileJulien Nabet
if aDbtFile corresponds exactly to aTmpFile, we just have to return Reviewed-on: https://gerrit.libreoffice.org/39189 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Julien Nabet <serval2412@yahoo.fr> (cherry picked from commit 36d91a65ab2db0c4c81e09771f6b44e1905122a0) Change-Id: I1dde8efbe01838dc94f6025d5b81f5aee01932ac Reviewed-on: https://gerrit.libreoffice.org/39196 Reviewed-by: Lionel Elie Mamane <lionel@mamane.lu> Tested-by: Jenkins <ci@libreoffice.org>
2017-06-24Resolves: tdf#108292 WalkAndMatchElements: really limit the matchEike Rathke
... to the columns queried, not just when entering an mdds node. Otherwise it would return an unexpected index, plus bailing out early spares unnecessary comparisons for the rest of a node block. Regression of commit 3fed166279377f7ad702b8911899243b8adff3bf Date: Fri Aug 16 16:29:38 2013 +0200 that started to use commit 7334f8db6f6004d48e2dbf014f27878a7ae21eb1 Date: Fri Aug 16 16:29:27 2013 +0200 with its bad implementation. Just that VLOOKUP on a matrix with a larger block of same typed data as the query *and* a match in an excess column seems to be rare.. Change-Id: Ia4ef3fd56490de82910d5aa13a84be2de851f9b0 (cherry picked from commit d3b77628efc72d857c63c8fb91d7ed2b499ac860) Reviewed-on: https://gerrit.libreoffice.org/39082 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com>
2017-06-20Resolves: tdf#104875 defer filter detection to first load/update of externalEike Rathke
This is a combination of 2 commits. This could never have been meant Since commit 67b374980f340faa2695001984aaf98886a32fad executing ScDocumentLoader::GetFilterName() twice if there was no stored filter, and still executing it if there was one, even if the comment says we shouldn't. (cherry picked from commit 1bc793f8ab8f9b4ce435bd3585f9579df4a11819) Resolves: tdf#104875 defer filter detection to first load/update of external Attempting to access all linked external documents right on load time was specifically nasty for non-local URIs (unmounted smb, http, ...) (cherry picked from commit 416752b9e4bc4605c479d3eff7797be9f0ef2a38) f4a7414dfe36875b505c73b8c963b650eeffe204 Change-Id: Iaf4b88e63b3987fb661df0e9b095f06b29076894 Reviewed-on: https://gerrit.libreoffice.org/38820 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com>
2017-06-15Resolves: tdf#104989 a matrix/array formula cell needs to have dimensionsEike Rathke
This is a combination of 2 commits. Resolves: tdf#104989 a matrix/array formula cell needs to have dimensions Most spreadsheet functions don't need it, but some may. (cherry picked from commit 033d018a3a1d345d61f3c4dc4da693cdd4004dea) Ensure all cases push an error return, tdf#104989 related (not the cause) A formula cell in matrix context could have no dimensions.. (which is the cause but elsewhere). (cherry picked from commit e6abe9c3d06d6776e05bef93b5e6f87b403ea967) 1a04f9808d8f5c802789e3db77052351b24c24c7 Change-Id: I954acaa0a64c4dac8d15acd5b3f404019a7e0bce Reviewed-on: https://gerrit.libreoffice.org/38832 Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com> Tested-by: Jenkins <ci@libreoffice.org>
2017-06-12tdf#106423: dbase file can have no recordsJulien Nabet
Regression from https://cgit.freedesktop.org/libreoffice/core/commit/?id=4e3ff19b33c84557fd20e68960499933b4e52638 tdf#84834 sc: stricter type detection for dBASE files Log test file from tdf#84834 still opens in Writer as expected Change-Id: Ibf5a0f586e3a132455d2363e61e5cc7ea249002d Reviewed-on: https://gerrit.libreoffice.org/38660 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Michael Stahl <mstahl@redhat.com> (cherry picked from commit 57ae8834c8313354477117b65d87c86407f6bf82) Reviewed-on: https://gerrit.libreoffice.org/38692 Reviewed-by: Lionel Elie Mamane <lionel@mamane.lu>
2017-06-03sc::ColRowSpan::mnEnd is inclusiveEike Rathke
Wrong since commit 1e721077b43de84edab2a3ed2f316ddcbec6e3ec Date: Tue Oct 7 00:22:32 2014 +0900 and commit 91502a72c12c559442e8bf77c27a516b49c2a68d Date: Thu Oct 9 14:44:27 2014 +0200 So HasAttribSelection() and HasSelectionMatrixFragment() did not query the rightmost column of each marked range from ScMarkData. Change-Id: Ie3aca26ee1b6c249c1135268dc0ad3ea053ea2b8 (cherry picked from commit 77d1e0c49adea959dd418de9b6cf9901e01be3ae) Reviewed-on: https://gerrit.libreoffice.org/38371 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Caolán McNamara <caolanm@redhat.com> Tested-by: Caolán McNamara <caolanm@redhat.com>
2017-06-01Resolves: tdf#108279 dimension rows are 32-bit and don't wrap at 64kEike Rathke
Change-Id: I7bf7a149fb1db526ef62bea2f52b200cf5053cc2 (cherry picked from commit 259c2409310814ae62c8d0aefef19a71e8d88dbe) Reviewed-on: https://gerrit.libreoffice.org/38331 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Julien Nabet <serval2412@yahoo.fr> Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com>
2017-05-29tdf#107389: Wrong positioning of AutoFilter buttons in tiledrendering modeTamás Zolnai
Must be a regression from this commit: 9113f17e65b1dafbc178fd269e2967c1e2632f0e Where the offset's and origin's meaning were changed, but not all code path was updated accordingly. Change-Id: Ib519303d5ca5ac7f13e74944e8147ea3220d7684 Reviewed-on: https://gerrit.libreoffice.org/38102 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Tamás Zolnai <tamas.zolnai@collabora.com> (cherry picked from commit ba5a9132375bcfedebd7c3d2500006161800da3b) Reviewed-on: https://gerrit.libreoffice.org/38112 Reviewed-by: Eike Rathke <erack@redhat.com>
2017-05-29import stream on the LibreOffice side, tdf#107827Markus Mohrhard
This adds support for all LibreOffice content providers and non-ascii file names. Change-Id: I0a0fd0ab0c78ef039679516e7a3e880929e8f725 Reviewed-on: https://gerrit.libreoffice.org/37578 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com> (cherry picked from commit 5e9178f765b327a8aba17476d135ba3b1f881bf7) Reviewed-on: https://gerrit.libreoffice.org/37582 Reviewed-by: Eike Rathke <erack@redhat.com>
2017-05-29use empty document in orcus gnumeric import, tdf#107828Markus Mohrhard
Change-Id: Ia7ee2de6f9b14a132a29c01f8d863889477c0a8d Reviewed-on: https://gerrit.libreoffice.org/37579 Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com> Tested-by: Markus Mohrhard <markus.mohrhard@googlemail.com> Reviewed-on: https://gerrit.libreoffice.org/37659 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Eike Rathke <erack@redhat.com>
2017-05-29PushError() instead of SetError() so there's a result tokenEike Rathke
Change-Id: I8c2a7e01b04c26e008c8cbea9332f71d47293c75 (cherry picked from commit cd87661b4761ee9f512fc01f95d84f786c97bb66) Reviewed-on: https://gerrit.libreoffice.org/37950 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com>
2017-05-26tdf#108049 VBA exception avoided in ScVbaRange::setStyleJustin Luth
Change-Id: I5fc928373d70829b6e68d9000140a30f04179536 Reviewed-on: https://gerrit.libreoffice.org/37981 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Justin Luth <justin_luth@sil.org> Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com> (cherry picked from commit 4dc3e7d5e9044be66177e467d739bd8be7556783) Reviewed-on: https://gerrit.libreoffice.org/38063 (cherry picked from commit bb9c240046955c432d963610d5dd99edcc6b43e6) Reviewed-on: https://gerrit.libreoffice.org/38083 Tested-by: Justin Luth <justin_luth@sil.org>
2017-05-26tdf#108017 xlsb import: protection settings were reversedJustin Luth
Back in LO4.3, commit 5728a56b20e01d6c9e86a23f20af64a1b68ad604 fdo#70499 selectn of lock/unlocked cells with worksheet protection reversed ...reversed finalizeImport's understanding of maSheetProt.* to match xml import, but then it no longer matched the binary stream import. So, .xlsb imported the opposite of what .xlsx/.xlsm imported. Reviewed-on: https://gerrit.libreoffice.org/37952 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Justin Luth <justin_luth@sil.org> Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com> (cherry picked from commit 0d3fcbf809852aab0d0f88272bbfd2e8507d1c5e) Reviewed-on: https://gerrit.libreoffice.org/38041 Change-Id: I2da3967ef52187e4ef36ce7d01b17a4672f61e1c Reviewed-on: https://gerrit.libreoffice.org/38052 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Justin Luth <justin_luth@sil.org> Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com>
2017-05-18tdf#107920: Crash when using ScCompiler in condformatJulien Nabet
Just call SetGrammar from ScCompiler object with grammar of document See bt in https://bugs.documentfoundation.org/attachment.cgi?id=133382 Change-Id: Iedf4b88f2821bbcf728e4b1c023041fd6f10e5f3 Reviewed-on: https://gerrit.libreoffice.org/37739 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Eike Rathke <erack@redhat.com> (cherry picked from commit 63843a97262ccfa38d838a9cbbc3faadba8290a6) Reviewed-on: https://gerrit.libreoffice.org/37756 Tested-by: Eike Rathke <erack@redhat.com>
2017-05-17tdf#107892: fix negative values case with maxifsJulien Nabet
std::numeric_limits<double>::min() returns positive value for float/double so use std::numeric_limits<double>::lowest() (see http://stackoverflow.com/questions/17070351/why-does-numeric-limitsmin-return-a-negative-value-for-int-but-positive-values) Change-Id: I0afce2d38c6936aeff26923182bcafd2f0008d9e Reviewed-on: https://gerrit.libreoffice.org/37693 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Eike Rathke <erack@redhat.com> (cherry picked from commit 20d3a60b41bda2513723c145d919f584840b1056) Reviewed-on: https://gerrit.libreoffice.org/37730
2017-05-15tdf#82326 calc 32bits unable to open files with a lot of cond formattingfrederic vroman
The code to build and compute formulas is using an fixed size array. The size of this array has been changed from 512 to 8192 by commit 9c1ca6dca3b553c302a635357e33591605343b99 Since then, LO consumes a LOT of memory while opening scalc files which are making extensive use of conditional formatting. This is a silent issue when working with a 64bits versions of LO because the memory is immediately released at the end of the load phase. However, with 32bits versions, some files can not be opened anymore as it can consume the full process memory space (2GB on windows). The parser is creating ScCondtionEntry objects to store conditional formatting expressions which are immediately compiled in the constructor. However the compile method did not return a flattened token array. Change-Id: Id290bc1f5838c246d4c2322b297b13307defaf88 Reviewed-on: https://gerrit.libreoffice.org/37527 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Eike Rathke <erack@redhat.com> (cherry picked from commit f566e2d579f5385b01d9b124e7055721313325c9) Reviewed-on: https://gerrit.libreoffice.org/37644
generated by cgit (git 2.34.1) at 2024-05-29 14:03:01 +0000