summaryrefslogtreecommitdiff
path: root/sc
AgeCommit message (Collapse)Author
2015-02-13add unittest for tdf#89330Jan Kantert
Change-Id: I4c5c65733700e7e7245e96f85714221acf23bcfb Reviewed-on: https://gerrit.libreoffice.org/14473 Reviewed-by: Eike Rathke <erack@redhat.com> Tested-by: Eike Rathke <erack@redhat.com>
2015-02-13tdf#89330 preserve file name case in an unresolvable external named rangeJan Kantert
Change-Id: Ib8728a4a9c793b162de07a0cef66e242879f2aa1 Reviewed-on: https://gerrit.libreoffice.org/14474 Reviewed-by: Eike Rathke <erack@redhat.com> Tested-by: Eike Rathke <erack@redhat.com>
2015-02-13Compute n-th root of negative numbers where n is odd. (tdf#69293)Supreme Aryal
Compute expressions like (-8)^(1/3) correctly. This makes calculations compatible with Excel. Exponents must reduce to the form 1/n. Change-Id: I007c818f584323f80f2f6b1000d931f19a4590ad
2015-02-13Translated german commments in sc/source/core/data/table1-3.cxxAlbert Thuswaldner
Change-Id: Ic28cb4f9371caf39e783f39fd8a117260b962bfe Reviewed-on: https://gerrit.libreoffice.org/14451 Reviewed-by: Caolán McNamara <caolanm@redhat.com> Tested-by: Caolán McNamara <caolanm@redhat.com>
2015-02-12tdf#89281 fix performance regression of XLS import - cleanupLászló Németh
Change-Id: I6d7f279732d9992d584aab96c3a747d6e6130147
2015-02-12fdo#39468 translated german commentsMichael Jaumann
Change-Id: Iec457e270790cb6f58b31e5002f9b2ce8acc1618 Reviewed-on: https://gerrit.libreoffice.org/14444 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Eike Rathke <erack@redhat.com>
2015-02-12apparently calling match on a string with an index past the end is now fatalCaolán McNamara
assert on loading ooo55345-1.ppt etc Change-Id: Ie76720e511674ddd883f7e397442a84838c1a1f4
2015-02-12coverity#1267653 Logically dead codeCaolán McNamara
Change-Id: I371e719cc7e1ba2faa53535f25eca1d9074342bb
2015-02-12coverity#1242895 Untrusted loop boundCaolán McNamara
Change-Id: I73b8d2c6f9ef81c246b541aa34c4b61109964440
2015-02-12Translated german comments in sc/source/core/data/ #2Albert Thuswaldner
Change-Id: Ia8a0235b8acb5c76c195e280a75256f6b82684bb Reviewed-on: https://gerrit.libreoffice.org/14372 Reviewed-by: Caolán McNamara <caolanm@redhat.com> Tested-by: Caolán McNamara <caolanm@redhat.com>
2015-02-12Translated german comments in sc/source/core/data/Albert Thuswaldner
Change-Id: I82ca987d5d74d8137b9ed02e085f390191344bb0 Reviewed-on: https://gerrit.libreoffice.org/14371 Reviewed-by: Caolán McNamara <caolanm@redhat.com> Tested-by: Caolán McNamara <caolanm@redhat.com>
2015-02-11Resolves: tdf#39316 add matrix empty cells to ScInterpreter::QueryMatrixType()Eike Rathke
Change-Id: Ifa5d59e90afcfff66f2e8683fac2a9090ed615da
2015-02-11add ScMatrix::IsEmptyResult() for tdf#39316 fix preparationEike Rathke
Change-Id: I9f2a3e6fa9f4ef43df672661afef996269b90a7a
2015-02-11tdf#89319 remove redundant file seekings in XLS exportLászló Németh
Normally StartRecord() is called with closed records and with end position of the stream set by EndRecord(), so the mrStrm.Seek( STREAM_SEEK_TO_END ) in the InitRecord() is a redundant call. The patch removes this call, and sets it only for the non redundant cases: when there is an unclosed record in StartRecord() or a continue record in StartContinue(). Change-Id: Iecbcaf01cbfe6094fa73d5ed41dba5f01417edb3
2015-02-11fdo#39468 translated german commentsMichael Jaumann
Change-Id: If229c1ae9a737b492948c78b619aca90b4a33e83 Reviewed-on: https://gerrit.libreoffice.org/14417 Tested-by: Michael Stahl <mstahl@redhat.com> Reviewed-by: Michael Stahl <mstahl@redhat.com>
2015-02-11tdf#88576 check that two parameters are given for IFERROR() and IFNA()Eike Rathke
Otherwise the functions failed only for the error case resulting in #NULL! error. Change-Id: Ieb987637698ab418fc0a60cd873e23878c9f497b
2015-02-11tdf#88576 fix handling of empty arguments in IF(), IFERROR() and IFNA()Winfried Donkers
Change-Id: I8acca26cf7398768a9e25f97f3a9e61754ab2179 Reviewed-on: https://gerrit.libreoffice.org/14415 Reviewed-by: Eike Rathke <erack@redhat.com> Tested-by: Eike Rathke <erack@redhat.com>
2015-02-11eliminate false failure warningEike Rathke
Change-Id: I86a1110c50ace958a68a6971aec34ec632b91c93
2015-02-10tdf#79239: make the dialog closable with 'X' button againKatarina Behrens
that one of window manager, top-right corner mostly. We need to reimplement Close() method for this class - I don't quite get why this has to be the case, but all classes derived from ScAnyRefDlg do it as well. Change-Id: I3e94b7ee09f9b3581d054818d36ea4fb0fd55f78
2015-02-10ExtCfRule::maModel and CondFormatRule::mpDataBar are 2 distinct thingsKatarina Behrens
so do it right on 2nd attempt Change-Id: I2c51943ec831591a47afc16599e2e7246407b31a
2015-02-10mbNeg means: a different colour for negative values is definedKatarina Behrens
It is totally orthogonal to whether there's a gradient fill or not Change-Id: I32ab465d9116c3dc31cde3611cd19d3ec10ac5d6
2015-02-10tdf#68117 Partially fix boolean labels in XLSXUrsache Vladimir
Change-Id: I731a58c7749e157f6b40c60808687ce629683742 Reviewed-on: https://gerrit.libreoffice.org/14410 Reviewed-by: Markus Mohrhard <markus.mohrhard@googlemail.com> Tested-by: Markus Mohrhard <markus.mohrhard@googlemail.com>
2015-02-10Avoid OpenCL compilation error in some casesTor Lillqvist
We do need to use GenSlidingWindowDeclRef(). We can't assume it is always a vector element that is taking part in the calculation. Change-Id: Ifcda5bdc4564d5d13755b10d9296802cee723182
2015-02-10Don't call clReleaseProgram() on the same program twiceTor Lillqvist
If the clBuildProgram() failed, the lastSecondProgram variable kept as its value the already released program handle, and later we called clReleaseProgram() on it again. This is certainly wrong, and caused a crash at least with my OpenCL implementation. As such I am not sure if that lastOneProgram, lastSecondProgram etc dance is sane... Change-Id: Ia30426fbba9118ad7e20f13ef35daa5f78a1f0a1
2015-02-10Add SAL_INFOs for tracing kernel and program life cycleTor Lillqvist
Change-Id: Icb28114d3939063dedaedbd0ce370210b3721fc5
2015-02-10fdo#39468 translated german commentsMichael Jaumann
Change-Id: I96ef1c6758eaffb88e2167acee7c9d810317ca44 Reviewed-on: https://gerrit.libreoffice.org/14384 Tested-by: Michael Stahl <mstahl@redhat.com> Reviewed-by: Michael Stahl <mstahl@redhat.com>
2015-02-10tdf#89281 fix performance regression of XLS importLászló Németh
The fix for Bug 76611 caused ~20% performance loss in XLS import. With optional cloning of ScTokenArray of the shared XLS formulas depending on the need of address wrapping, it is possible to gain back near the original performance. Note: The original patch for Bug 76611 has already got an unit test, too (see wrapped-refs.xls), and that unit test works with this improved patch, too. Change-Id: Ibfb59d1543ef9c4b8a075d5c4e37f77ab451aaa0
2015-02-10Avoid OpenCL compiler warning "OpenCL extension is now part of core"Tor Lillqvist
64-bit floating point became an optional part of OpenCL core in 1.2 and is no longer an extension. The pragma is not needed in 1.2 and later. It causes a warning that is indeed harmless, but still clutters the compiler diagnostic output that is displayed when there is some actual error in generated OpenCL code. Change-Id: Id8ea4b39f7e8e788d7683f8b389f53eadc57ea18
2015-02-09Handle zero and empty cells (which also means zero) in OpenCL for divisionTor Lillqvist
Not sure if it makes sense to keep having OpDiv a subclass of Reduction. There is no DIV() function that would take a range of cells, so it isn't really comparable to the other Reducion subclasses. But let's keep that as it is for now. We need to handle three cases specially in the OpenCL: Dividing by an empty cell which should produce an #DIV/0! error, dividing an empty cell by zero which also should produce #DIV/0!, and dividing an empty cell with anything else number which should produce 0. Change-Id: I86d86f652047d6f9e3c095c3ef135a8f5396b000
2015-02-08coverity#1242895 Untrusted loop boundCaolán McNamara
Change-Id: Ib54ed3f18c91142f1c217f618d88e36a646cf931
2015-02-08Avoid OpenCL compilation error when multiple RAND() calls in a formulaTor Lillqvist
Simply surround the Random123 code snippet with an ifdef guard. Change-Id: I370a3c37226d31bfbe703e5b7936b2180aee1784
2015-02-08Create proper error when dividing by zeroTor Lillqvist
Create a so-called "double error", i.e. a NaN with a error code payload. Change-Id: I6d538426c184b30d067f8ef6035b49f3a8986f12
2015-02-08Add a couple of SAL_INFOs to trace the vector refsTor Lillqvist
Change-Id: I6b71e320359d025bf8cf31637dabb1bc35d581fb
2015-02-08Add one more case for string argument handling when attempting to use OpenCLTor Lillqvist
If a function takes numeric arguments but there are only string arguments, but the settings say strings are to be treated as zero anyway, we can go the OpenCL path. Also add a few informative comments. Change-Id: I40b64cf040c986cfc2b26a2a778711f3cd397186
2015-02-08Handle string arguments better when attempting to use OpenCLTor Lillqvist
If a function/operator that takes numeric arguments but not string arguments is passed string arguments anyway, handle that with OpenCL only if the (document-specific!) setting for text-to-number conversion is ZERO (treat strings as zero, which is how our OpenCL code for such functions/operators works). Otherwise let the code fall back to the traditional interpreter. Change-Id: I351da75b225baf4b721cd1ab67ce5094dbc5e458
2015-02-08Add ScCalcConfig field to DynamicKernelArgumentTor Lillqvist
It will be needed to be able to adapt OpenCL code generation to the text-to-number conversion settings. To get the document-specific ScCalcConfig into the DynamicKernelArgument, we have to pass it along in lots of places to constructors of various DynamicKernelArgument subclasses. Originally it comes from the maCalcConfig field of FormulaGroupInterpreterOpenCL, inherited from FormulaGroupInterpreter. Change-Id: Iac6a83e8081ab973e8b7e8161de79e411d2ed2dd
2015-02-08Be prepared to handle error codes encoded in NaNs ("double error")Tor Lillqvist
Change-Id: I84576139d271d05977ff8bfa4de4820ef523d69e
2015-02-07databaroptions.ui: Reduce row spacing, right-align new Fill labelAdolfo Jayme Barrientos
See also tdf#87356 Change-Id: I4866afb179fa2425f4ba6be48dde33461b00c255 Reviewed-on: https://gerrit.libreoffice.org/14356 Reviewed-by: Katarina Behrens <bubli@bubli.org> Tested-by: Katarina Behrens <bubli@bubli.org>
2015-02-07Some compilers need explicit OUString() in return expressionTor Lillqvist
Change-Id: Ideb64a7cf6fd4ad563b0e4bfab5583af1ecefd54
2015-02-07Not all platforms have std::to_string(), urkTor Lillqvist
Change-Id: Ie930f2110bc1c10ac3ada9e266ca25d4ebdea5cd
2015-02-07ERROR is a macro on Windows, try ILLEGAL thenTor Lillqvist
Change-Id: Iae0c5308f77657c6a55bafd372ce7e4db5ef4aab
2015-02-07We (will) need ScCalcConfig also for formula groupsTor Lillqvist
The settings for interpretation of text (string) cells as numbers affect whether OpenCL can be used or not to get expected results, when operating on vectors that contain also text cells. Change-Id: I0a7fd657038846ba762aa6ee6a95c001aea8f124
2015-02-07Turn StringConversion into a C++11 scoped enumerationTor Lillqvist
Change-Id: I353a62bac6c8bf00b20c93d9778fc45ade5d502c
2015-02-07Output meStringConversion symbolicallyTor Lillqvist
Change-Id: Ibcd8cb12525fbce33fbfd208ee8e357c904ffd4f
2015-02-07Introduce SetResultError()Tor Lillqvist
Change-Id: I0645b7a58e9f3dfc6c431844805ab87add8745ad
2015-02-07loplugin:deletedspecialStephan Bergmann
Change-Id: Ieeca9fe957e7bc6a4cf9d7d6ac57f9ed150aab78
2015-02-06Resolves: tdf#83461 do not override MatColsRows if already setEike Rathke
ScMatrixFormulaCellToken::SetMatColsRows() via ScFormulaCell::SetMatColsRows() is used during document import and preselected cell area input of an array formula. Do not override existing values with subsequent result matrix dimensions. Change-Id: I9e844b5064ea276f3cbcb680eb1127c344328e00
2015-02-06shared_ptr<T>(new T(args)) -> make_shared<T>(args)Caolán McNamara
and boost:make_shared->std::make_shared Change-Id: Ic1e187c52c856a7b27817967b2caa8920f23a98d
2015-02-06Report errors from clRelease*() APIsTor Lillqvist
Change-Id: Id30afe5bf954e26515bf8cca6f1ee8bc018fb835
2015-02-06Unmap the host buffer only after done accessing itTor Lillqvist
Most likely 64c479e9da02f724e1870649c99fac92f5f27cd3 accidentally made the code unmap the host buffer before it is accessed, but the code continued to work by accident in many (most?) cases. Either because in the case of OpenCL devices that share memory with the CPU, the host buffer *is* the OpenCL buffer, so even if the host buffer is "unmapped", it still exists. In the case of GPU device with separate memory, using the host buffer after unmapping corresponds simply to a case of use after free of a heap-allocated buffer, which often happens to work. Found by code reading. Change-Id: I9e2b4574077a267938702c0f81c4b1cba9c9a183
generated by cgit (git 2.34.1) at 2025-02-01 19:58:31 +0000