summaryrefslogtreecommitdiff
AgeCommit message (Collapse)AuthorFilesLines
2014-12-15vcl: Don't keep a reference to the texture in the FBO objectprivate/moggi/opengl-4-4-build-testLouis-Francis Ratté-Boulianne2-9/+17
Change-Id: I240d2b44e77d28af3cd5952b6d666a1709c4c54a
2014-12-15vcl: Re-use a framebuffer of the same size when possibleLouis-Francis Ratté-Boulianne5-20/+93
Conflicts: include/vcl/opengl/OpenGLContext.hxx Change-Id: Id9c7932976ce9d9282776c20d93d9cca4d290056
2014-12-15vcl: Keep the same context for VirtualDevice to avoid FBO switchesLouis-Francis Ratté-Boulianne1-10/+9
Change-Id: I66496fae902db9df5b8301b00bb779f42adaa7a7
2014-12-15vcl: Reset context when the backend window is destroyedLouis-Francis Ratté-Boulianne4-62/+137
Conflicts: vcl/source/opengl/OpenGLContext.cxx Conflicts: vcl/source/opengl/OpenGLContext.cxx Change-Id: Ie2b93de8efe5ea56b0420adf23639c0153103385
2014-12-15vcl: prefix global variable g_vShareListMichael Stahl1-14/+14
Conflicts: vcl/source/opengl/OpenGLContext.cxx Change-Id: I13ecc76d8bd75ee65e626670995d7c93efc953ca
2014-12-15vcl: Limit Cairo surface size to the clipping region to improve performanceLouis-Francis Ratté-Boulianne8-2/+57
Conflicts: vcl/unx/generic/gdi/openglx11cairotextrender.cxx Change-Id: I469b34c9f1047a274550229391d3dfb578291df6
2014-12-15vcl: Reset data buffer after resizing bitmap on GPULouis-Francis Ratté-Boulianne1-1/+1
Change-Id: I516cdbc466f3d6427e36fea8c5cdbe718ce7d0ea
2014-12-15vcl: Use right size to draw native GTK window backgroundLouis-Francis Ratté-Boulianne1-4/+7
Change-Id: Id446d20599f072f657c6106d6c6457fce08830e6
2014-12-15vcl: Add BitmapInfoAccess for bitmaps' metadata accessLouis-Francis Ratté-Boulianne22-177/+259
Change-Id: Iec9f8c7d8f7cded0aef9e017373e44387dc0b05c
2014-12-15do not hardcode what specific "generic" bitmap scaling operations useLuboš Luňák4-15/+16
I.e. make it possible for the opengl backend to decide to use a different scaling algorithm. Change-Id: I36e76de071797129c7636b3048c93bc6e5d93aca
2014-12-15fdo#82614: sw: Revert "Simplify code path"Michael Stahl1-2/+14
This reverts commit 07c00671443f3f0a2e2e6c3a54c391f040e8fd2b. The bug is because the check of the GotoHeaderTxt() is inverted. Actually in the old code it is a bit easier to understand what is intended, even though it is more verbose, so just revert the commit. (cherry picked from commit 86910c87c193345e5e36104e21b8e26790d6846d)
2014-12-15Resolves: fdo#87281 Crash on importing diagram in .docx fileGary Houston1-1/+3
Change-Id: Ic2c7326ab83258e2e9dcbac88aa9f76f771a9b12 (cherry picked from commit de41a2038ac44f71a9adef97e6ccea83c99dd1c2)
2014-12-15vcl: Fix crash when requesting coordonnates for empty textureLouis-Francis Ratté-Boulianne1-0/+8
Change-Id: I424ac73f23bed5b3782c148d9242ca8fc29d200d
2014-12-15fdo#87292 use system dependent path instead of file URLAndras Timar1-1/+5
Change-Id: I3da4cd7af16a8ecfb19ce23e01117e735e9dd8d6 (cherry picked from commit efe3e06dd1b99215d82b6373aa7434a8f25a22f2)
2014-12-14fdo#86761 RTF import: positive border width and fLine=0 means no borderMiklos Vajna6-3/+58
Regression from 01a32b7d074511bed24044dc94e1159aea62722b (fdo#85179 RTF filter: import image border, 2014-10-23), there were a number of problems here: - CppunitTest_sw_htmlexport: revert back to the old behavior, where in case there is no border, we don't set the color of it. - The testcase of the above commit omitted fLine=1 shape property, which is present in the original bugdoc, and only with that should we put a border around the shape. - Let fLine=1 explicitly change the line style from NONE. - dmapper: if line style is NONE, then don't bother setting the border color and width. Change-Id: Iffee41066d42822b699c478821645b9742df3f58 (cherry picked from commit 4568d1d298bf4fc98dcd86384743a04587a2fe6f)
2014-12-13Updated coreAndras Timar1-0/+0
Project: translations fd601de98d9273747f3c40c186087bee0f77172a
2014-12-13fdo#84909 enable start from current slide and add its tango iconsYousuf Philips7-5/+926
Change-Id: I2af37e19f5f54557415b80d9aad2bba38d381a87 Reviewed-on: https://gerrit.libreoffice.org/13462 Reviewed-by: Yousuf Philips <philipz85@hotmail.com> Tested-by: Yousuf Philips <philipz85@hotmail.com> (cherry picked from commit 56cc4c4f4c6e1807ab47965f9e40357b5b6eaf9d) Reviewed-on: https://gerrit.libreoffice.org/13463
2014-12-12fdo#67930 work around fdo#68024Lionel Elie Mamane1-1/+2
don't emit string-value attribute Change-Id: I3668b88437451523f2b6bb3c2c82e1929f7ba3d8 Reviewed-on: https://gerrit.libreoffice.org/5371 Tested-by: Miklos Vajna <vmiklos@suse.cz> Reviewed-by: Miklos Vajna <vmiklos@suse.cz> (cherry picked from commit 76408b0f0d7e91cddb8b056069a2ee547109b156) (cherry picked from commit 1e10f6d229587ffc2a955c225b489645f8016e23)
2014-12-12fdo#87044: reportbuilder: Revert "fdo#67930 don't use variables ...Michael Stahl1-33/+24
... for formattedtext in header/footer" This reverts commit b0e3c6b8fa548defd5a3af2b8bf88b0b49589753. This reverts commit fc92c1abebcfe9b18649d35b76bf22e001e332da. The other fix from the libreoffice-4-1 branch seems to work better. Change-Id: Idc5d7543c11dff74977729dab5afd978197dab6a (cherry picked from commit d6ce95ae2288859fe74d601f1bdaf616ab1ee7f0)
2014-12-12reportbuilder: re-add numerous bits of dead code needed by bug fixMichael Stahl2-1/+87
Change-Id: Ia0a8b8ec3b4684ad6766467c8e3413f91dc8bf67 (cherry picked from commit 458f22c338f39ad777ad3f34b45a81b4eb8dfe8e)
2014-12-12reportdesign: avoid assertion on duplicate properties ...Michael Stahl1-1/+14
... from XMLAutoStylePoolProperties that happens when opening bugdoc from fdo#87044. Change-Id: I7002cf5176cd326f57b50fb6030ffb89160c237b (cherry picked from commit 541800fa17c6a4f9552c76e5290b5011addb1264)
2014-12-12rhbz#1116534 crash when pasting over a formulaDavid Tardon3-23/+52
When pasting CSV data into existing document, we cannot use the ScImportExport optimization for insertion, because the cells can already have content (a grouped formula in this case). So use appropriate ScDocument functions, even if that means the import is slower. This is a regression from commit 93959db4d8846cfdfb87ab647c4d457fb09bb869 "use DocumentImport for csv import, related fdo#69006". Change-Id: I92f70abca0542d796e3aa674b28a31053fff00d0 (cherry picked from commit bf97980c8d5d5fbc63a08b8834c10fe903152b75)
2014-12-12resolved fdo#86809 connect filter conditions correctlyEike Rathke1-3/+10
Maybe since 0960ec3e1b7b0d872d1f84d2d56f480a4df08b21 all conditions of a single filter stack were OR connected, regardless whether filter-and was given in the file. Change-Id: Ifb423ccc7deb42bb6bba0bfc810607633f2c342c (cherry picked from commit b500e60b32b5cf8fd61beb7476c2c29e65aee3e4)
2014-12-12fdo#37814: SHCreateItemFromParsingName doesn't like LO's file URLsStephan Bergmann1-13/+25
...but apparently wants pathnames instead, cf. the other use of it in WinSalInstance::AddToRecentDocumentList (vcl/win/source/app/salinst.cxx). Change-Id: Iaa64ff82915f7cbfee55b4389387b226f7adc919 (cherry picked from commit e123ffd66088783ce74a7b5f15e9d324b39ecf67)
2014-12-11fdo#86572: Write test for this.Kohei Yoshida2-0/+38
Change-Id: I00c3eaac473f0464e923e785cc50ad6769c78d41 (cherry picked from commit 170616e9f2d30c1302bbb5a7a4b588bc05cd5cc9)
2014-12-11fdo#86572: Increment row position *only when* the row ref is relative.Kohei Yoshida1-2/+5
This is just a simple oversight. Change-Id: I024777c46641cce3410a20f40c48f47ad6a5dade (cherry picked from commit 52781a55441cb270ffecc2522d1e069141a21607)
2014-12-11Add undo operation for deleting a bookmark + unit test, related fdo#51741Matúš Kukan6-22/+80
(cherry picked from commit 9338bea6e8dfab8d360fe8ab19dd5d75071bfc2a) Conflicts: sw/qa/extras/uiwriter/uiwriter.cxx sw/source/core/undo/unbkmk.cxx Change-Id: I79d8d3c30b6b0b2cc253963fdd50019aec033e12
2014-12-11fdo#51741 Mark document as modified on bookmark rename / deleteRenato Ferreira4-2/+17
Change-Id: I378037ea546d0f2dc4ab9e0b0e6f9bf65c8db5b4 Deletion: overrided DdeBookmark::DeregisterFromDoc in Bookmark Renaming: in MarkManager::renameMark Reviewed-on: https://gerrit.libreoffice.org/12436 Reviewed-by: Michael Stahl <mstahl@redhat.com> Tested-by: Michael Stahl <mstahl@redhat.com> (cherry picked from commit f0c4cd51cb160492d78c2796f71ecd88da25fa59)
2014-12-11Add message to exceptionMatúš Kukan1-2/+4
Change-Id: I0e7cf850b51343e4afc1ea0a0409ad2c4e596435 (cherry picked from commit 373daf55d98ba706d3dea131e755da3a00eb259f)
2014-12-11Use IDocumentMarkAccess::findMark hereMatúš Kukan1-10/+3
Presumably, the names should be unique, so it's the same. Change-Id: I74907b97bc25f9512d4f7514162360e8b21ad4a5 (cherry picked from commit 43a803ccf8d392561bd168dc3b64695c4dcf93ce)
2014-12-11Writer: Set document as modified and add undo when renaming bookmarksMatúš Kukan3-22/+87
Change-Id: Id0d4695b6a7c5eae0493e679345bd41d4687d6d7 (cherry picked from commit c7fa6687e544e7f9702da23aadbd31c1f2e0115e)
2014-12-11only checked for new VirtualDeviceCaolán McNamara1-2/+2
Change-Id: I64d7a3c6c0894e231cb0644922675dcce0db89fa (cherry picked from commit f5607f7a440cc54a0a48582ae05c817c8c306b71)
2014-12-11stray debugging codeCaolán McNamara1-1/+0
Change-Id: I953389e236739c01226365c33ab777fc3972b69d (cherry picked from commit 067b560335195b24eeedc4514956029ea975fbbf)
2014-12-11Resolves: fdo#87242 reuse vcl clip for cairo during animationsCaolán McNamara2-3/+25
Change-Id: I0a26d4c4092226732620c3852b0402ee45d4fa1d (cherry picked from commit 94d935eecbba0161de2616c2234b4a5d9d3cad88)
2014-12-11Related: fdo#87242 merge duplicate clip setup codeCaolán McNamara4-147/+112
favoring the vclcanvas one for the places where they diverge Change-Id: I18e3d4e7659ebd4cb90c86718c1b1035671b4be3 (cherry picked from commit f88b5ab8692ee7ecf58b570e703d0e7f10cc2f0d)
2014-12-11Related: fdo#87242 init VirtualDevice with size of surfaceCaolán McNamara6-7/+15
otherwise vcl's clipping doesn't work quite right when the render text with vcl apis fallback is used. Manually forced in my case, but it should happen in practice with vertical text, so if there is a bug about vertical text not appearing in slideshows then this is part of the fix for that. Windows and Mac remain unchanged as initialized with 1, 1. If the same problem affects those platforms then they'll need to be adjusted to remember their height/widths from the ctor and those values plugged in here instead Change-Id: I2f82f0db0cf446d7db21f0a7ee4f8c15c7ebdb42 (cherry picked from commit f95b0743da4239e047db8638c61f90f8bbe54306)
2014-12-11desktop: avoid loading to-be-recovered documents when recovery is disabledMiklos Vajna1-2/+4
Change-Id: I5c8a117df53e0c7efd1e82ec6e56080c8fd42345 Reviewed-on: https://gerrit.libreoffice.org/13438 Reviewed-by: Michael Stahl <mstahl@redhat.com> Tested-by: Michael Stahl <mstahl@redhat.com> (cherry picked from commit fbe9accf36ac4c52ef5bcb84d455b4d06c317668)
2014-12-11unit test for matrix comparison error propagation, fdo#87237Eike Rathke2-0/+24
Change-Id: Id02f169b111945673a16d0c852940c805aa89319 (cherry picked from commit 6c41a1272e43a3e3b785724991837b597dbf00ef)
2014-12-11resolved fdo#87237 propagate error values through matrix comparisonsEike Rathke2-45/+53
Apparently introduced with 8e8b43a03e77dd251876c1de0ac06eeeb09192cd the comparison results were stored as boolean values, effectively discarding any infinite double values and error values encoded as NaN values. Change-Id: I1fb6f46894a0bee02a37e28b7e6cc84f8c051f28 (cherry picked from commit 3c405ff82fcc9f8f044833420485c54658064636)
2014-12-11do not save UNO marks in undo historyLuboš Luňák1-4/+6
These are usually very short-lived marks that are used for various internal purposes or by UNO interfaces, so restoring them from undo history does not make any sense. It can in fact lead to a crash on undo if the mark has been already deleted (which is usually the case with UNO marks) but the restoring still refers to it (this is not a problem with other mark types since those should be deleted only when they are no longer in the document, in which case they are saved in undo history by content positions rather than mark references). Change-Id: Id707dc3d2b888b5c96a974d04a0ef17b22fc024c
2014-12-11fdo#87030: Make PDF signing using Windows API workTor Lillqvist2-12/+146
This is a squash of four commits that incrementally add Windows-specific code for PDF signing and make it work properly. 1) Remove noise SAL_WARNs that were not warnings, just informative messages SAL_DEBUG is what should be used for stuff like that (and not committed). 2) Started writing a Windows version of PDFWriterImpl::finalizeSignature(). The certificate we get from the selection dialog (in xmlsecurity) is from the Windows built-in store anyway, so it is pointless to try to use it with NSS. (See bug for longer discussion.) So far it was pretty straightforward. The WinCrypt API seems clean and easy to use. But for some reason the CryptSignHash() call fails with "Keyset does not exist" (NTE_BAD_KEYSET). What am I missing? Anyway, comitting this now as it does compile and doesn't make things worse. 3) Prevent PDF signing using Windows API from failing There was one details that I had missed in my initial coding: CryptAcquireContext() doesn't give you a HCRYPTPROV key container that would contain the private key of a public key certificate. For that you need to use CryptAcquireCertificatePrivateKey(). When the hash has been created using such a HCRYPTPROV, the CryptSignHash() call succeeds. The certificate in DER encoding that is passed in from the caller, obtained in the certificate chooser (in xmlsecurity), is possibly not good enough to be used for the other things. So look the same (?) certificate up in the user's key store instead. At least more properties are present in the certificate when looked up like that. Add more SAL_INFO logging, with cleartext dumping of certificate context property names and list of algorithms supported by the CSP. Unfortunately, even if all the WinCrypt API calls now succeed, the signatures we produce still are not good enough for Adobe Reader... A lot of information must be missing, they are quite short, just 256 bytes. 4) Generate a proper PKCS#7 signature The signature should be in DER-encoded PKCS#7 format and what CryptSignHash() produces is nothing like that. Luckily CryptSignMessage() is actually almost easier to use and is capable of doing what we need. This also means that we won't need any HCRYPTPROV or HCRYPTHASH after all so all the code related to that can be removed. CryptSignMessage() handles both calculating the hash and signing it. One less than ideal issue with CryptSignMessage() is that it needs all the data to be hashed and signed at the same time, so we need to keep both buffers around for signing. It also turns out that we don't need to look up the certificate anew from the user's certificate store after all. Now Adobe Reader doesn't complain any longer about the signature's format and contents. Change-Id: I4674805cca2fa9342860e635cb07cbdfddfdb08c Reviewed-on: https://gerrit.libreoffice.org/13434 Reviewed-by: Jan Holesovsky <kendy@collabora.com> Tested-by: Jan Holesovsky <kendy@collabora.com>
2014-12-11Remove resizable:False property from dialogs containing GtkEntry.Jan Holesovsky68-68/+1
As agreed with Adolfo (and others) at the Design hangout on 2014-12-10. Change-Id: Ibe0e8075566523e746866b9d2c15c348c345d2df
2014-12-10Deteremine the script type of source cell during paste replication.Kohei Yoshida12-24/+98
Doing this ahead of time eliminates the need to update script types of all replicated cells after the paste. This makes significant performance impact. Change-Id: Ic4f7d7b2fe152bd5640ddb1dae01cc2ed757657e (cherry picked from commit 808fd5fbd8868dfd95c8a38676815798fa2b79c4)
2014-12-11fdo#87203 make translatable AGGREGATE functionAndras Timar1-1/+1
Change-Id: Ic376a4e03cafbba65ec78723de188bd627c0ef4f (cherry picked from commit f6ef20b046c485c89fb25521dec1eb6642aac1ba)
2014-12-11fdo#86215 pool new ScPatternAttr if source and destination are different docsEike Rathke1-3/+5
Change-Id: Icdb68af83a0b208f9a427ee61ae18020c8835a41 (cherry picked from commit d9c15867e914d91a7e86c499ffb6ee62f5b443f6)
2014-12-11update creditsChristian Lohmaier1-1318/+1390
Change-Id: I0446149f3f2517f360a70fd73fe6f1713ef1d19e
2014-12-10fdo#75137: Can't Move Cursor Between Footnotes with Up/Down Arrow KeysTakeshi Abe1-1/+1
This fixes a regression introduced at 43941d9a5313fcd7fe39a61bd2eace64f7743486. Change-Id: If73883569049f0f3ae8a2d40f6704f74e63eecb2 Reviewed-on: https://gerrit.libreoffice.org/13401 Reviewed-by: Michael Stahl <mstahl@redhat.com> Tested-by: Michael Stahl <mstahl@redhat.com> (cherry picked from commit dad173d9588e6abc2a465198b7d2881d4629246a) Signed-off-by: Michael Stahl <mstahl@redhat.com>
2014-12-10fix packaging of OpenGL shaders, related fdo#83875Markus Mohrhard1-0/+7
Change-Id: I68f7d4e2d91d37013018fc3cc185c46eb81fcb7d
2014-12-10paste from system from the top left of selectionDavid Tardon1-1/+17
For example, 1. echo -e "5,6\n7,8" 2. select the lines and copy them to clipboard 3. select an area in Calc, starting from bottom right 4. paste 5. select comma as data separator in the CSV import dialog I think most people would expect the data to appear in the selected area, but currently it would be inserted starting from active cell (which is the bottom right cell of the selection). Change-Id: Ief0e05deda18928a832dca2a261f8493ac9a03a9 Reviewed-on: https://gerrit.libreoffice.org/13410 Reviewed-by: Eike Rathke <erack@redhat.com> Tested-by: Eike Rathke <erack@redhat.com> (cherry picked from commit 781639eb84fc09d0260092238c3f50216f15a3ff) Reviewed-on: https://gerrit.libreoffice.org/13421
2014-12-10Resolves: fdo#87141 Remove deletes the node, release doesn'tCaolán McNamara1-1/+1
regression from introduction of boost::ptr_vector originally one had to Remove and delete the return to delete it, while now the Remove automatically delete it. This bare Remove to just "release" it seems to have been overlooked Change-Id: I175f7c95862ce47402993e5e3b8feaa93af43e75 (cherry picked from commit efcbb12a27a380a442292f17c7f508555a66d4af)
generated by cgit v1.2.3 (git 2.39.1) at 2024-05-17 12:38:22 +0000