summaryrefslogtreecommitdiff
path: root/pyuno
AgeCommit message (Collapse)AuthorFilesLines
2012-09-20warning C4702: unreachable codeStephan Bergmann1-3/+0
(cherry picked from commit 618376bcb55906913bc76365d7db9b71eb422daa) Conflicts: pyuno/source/module/pyuno_runtime.cxx Change-Id: Ie2cc83dde7e0522e0bf669a1dd3d7a8ca680f7e2
2012-06-22fdo#50682: Setting DYLD_LIBRARY_PATH in python script appears unnecessaryStephan Bergmann1-4/+5
...and apparently has negative consequences (system CFNetwork framework picking up LO libsqlite3.dylib instead of system one, as DYLD_LIBRARY_PATH overrides recorded installnames). Contrary to the old comment ("so that 'import pyuno' finds libpyuno.so"), what setting LD_LIBRARY_PATH on Linux is still necessary for is so that python.bin (a stripped version of the python executable from external python module) finds libpython2.6.so, as it lacks an RPATH. ('import pyuno' finds pyuno.so apparently on PYTHONPATH, anyway, and pyuno.so in turn dlopen's libpython.{dylib,so} with full path.) (This might make dc82cf021f76ea83ff7a4bcb2d7525f2e111f0cc "Make PyUNO work --with-macox-version-min-required=10.6" irrelevant.) Change-Id: I1c3a6c61d4cc976d85956e587497a13a77689128 (cherry picked from commit a164d246f2ed941fbdef19ec5b47dac1bab25509)
2012-06-04nuke unused definesTakeshi Abe2-3/+0
Change-Id: I7924da4bff73ca8f82c76eb8cb211401530acdeb
2012-06-01targeted string re-workNorbert Thiebaud1-2/+2
Change-Id: I44518bb1b464bb0b458c0b5f1c249ef160ea3b26
2012-06-01targeted string re-workNorbert Thiebaud1-37/+30
Change-Id: Ibe5c739c3066f3e706a44b64d4ad775abde9662c
2012-05-29nuke unused typedefTakeshi Abe1-1/+0
Change-Id: I12a55e08a83b88221d992790c97c830761755582
2012-05-23removed unused temporary rtl::OUStringTakeshi Abe1-1/+0
Change-Id: I9d118a5fbe2306465d8bcdb9ddd0596588137e5e
2012-04-29make gbuild the default assumption of build.plBjoern Michaelsen1-0/+0
this removes dmake completely out of the build for migrated modules build.pl now assumes modules to be gbuild, unless there is a prj/dmake file Change-Id: I674a036b182ee13c5ec093e83cb3d38133112d3b
2012-04-24Reduce chatterTor Lillqvist1-1/+3
2012-04-06Replaced equalsAsciiL(RTL_CONSTASCII_STRINGPARAM(...)) with == operatorSzabolcs Dezsi1-5/+5
Pattern used: find . -name "*.cxx" -exec sed -i 's/\( *\)\(else if\|if\) *( *\([^!()|&]*\)\.equalsAsciiL( *RTL_CONSTASCII_STRINGPARAM *( *\([^)]*\) ) *) *)$/\1\2 ( \3 == \4 )/' \{\} \;
2012-03-21chmod -xTor Lillqvist1-0/+0
2012-03-16fdo#46926: PyUNO_cmp: return acquired referenceDavid Bolen1-4/+15
2012-03-09fdo#46926: fix the fix for Python 3Michael Stahl1-1/+1
Thanks to Maxime de Roucy for the hint that the "cmpfunc" type doesn't exist.
2012-03-07fdo#46926: fix UNO struct comparison in Python 2David Bolen1-3/+10
This requires setting a rich comparison flag in Python 2, while Python 3 uses rich comparison by default. (regression from a09ce46818fd4d5e08b3af9a478501cd8ef5b4fe)
2012-03-07fdo#46859: adapt string type checks to work with both Python 2 and 3David Bolen1-4/+10
(regression from a09ce46818fd4d5e08b3af9a478501cd8ef5b4fe)
2012-02-23Adapted GCThread to safer-to-use salhelper::ThreadStephan Bergmann3-26/+27
2012-02-18Fix typos in commentsElton Chung1-1/+1
2012-02-18pyuno: decreaseRefCount: valgrind warning:Michael Stahl1-1/+4
Apparently the thread spawned in decreaseRefCount runs and deletes itself before the "m_hThread != 0" from osl::Thread::create is executed; try a lame workaround for that. ==1877== Invalid read of size 8 ==1877== at 0x2A70E546: osl::Thread::create() (in pyuno/unxlngx6/lib/libpyuno.so) ==1877== by 0x2A70E351: pyuno::decreaseRefCount(_is*, _object*) (in pyuno/unxlngx6/lib/libpyuno.so) ==1877== by 0x2A7092B4: pyuno::Adapter::~Adapter() (in pyuno/unxlngx6/lib/libpyuno.so) ==1877== by 0x2A709393: pyuno::Adapter::~Adapter() (in pyuno/unxlngx6/lib/libpyuno.so) ==1877== by 0x5EF2F64: cppu::OWeakObject::release() (weak.cxx:213) ==1877== by 0x2A70DE69: ==1877== Address 0x1ee30818 is 8 bytes inside a block of size 32 free'd ==1877== at 0x4A0662E: free (vg_replace_malloc.c:366) ==1877== by 0x4C44B62: rtl_freeMemory_SYSTEM(void*) (alloc_global.cxx:285) ==1877== by 0x4C44DC7: rtl_freeMemory (alloc_global.cxx:355) ==1877== by 0x2A70E41E: osl::Thread::operator delete(void*) (in pyuno/unxlngx6/lib/libpyuno.so) ==1877== by 0x2A70E6EF: pyuno::GCThread::~GCThread() (in pyuno/unxlngx6/lib/libpyuno.so) ==1877== by 0x2A70E303: pyuno::GCThread::onTerminated() (in pyuno/unxlngx6/lib/libpyuno.so) ==1877== by 0x2A70E671: threadFunc (in pyuno/unxlngx6/lib/libpyuno.so) ==1877== by 0x4C2E242: osl_thread_start_Impl (thread.c:292) ==1877== by 0x3C26607D8F: start_thread (pthread_create.c:309) ==1877== by 0x3C262EF48C: clone (clone.S:115)
2012-02-16Work around macro redefinition warningStephan Bergmann1-2/+4
2012-02-16Missing returnStephan Bergmann1-1/+2
2012-02-16Adapt pyuno.so wrapper to Python 3 supportStephan Bergmann1-21/+55
2012-02-10Removed unused codeAlexander Bergmann4-134/+0
2012-02-06Reintroduce pyuno.so wrapper around libpyuno.soStephan Bergmann1-15/+28
...it was cleaned away by a09ce46818fd4d5e08b3af9a478501cd8ef5b4fe "Port PyUno to support Python 3" but is still needed to make sure libpyuno.so is loaded RTLD_GLOBAL (Python apparently loads its modules RTLD_LOCAL). At least with pre 4.5 GCC this can cause problems with C++ exception handling, see the mail thread starting at <http://lists.freedesktop.org/archives/libreoffice/2012-February/025166.html> "LibO 3.5RC2: terminate called after throwing an instance of 'com::sun::star::registry::InvalidRegistryException'" for details.
2012-02-06Added READMEs for modules related to URE, with content from the wikiJosh Heidenreich1-0/+1
2012-01-26Code cleanup: ( () ) replaced by (())Alexander Bergmann2-2/+2
2012-01-21Removed some unused parameters; added SAL_UNUSED_PARAMETER.Stephan Bergmann5-21/+31
SAL_UNUSED_PARAMETER (expanding to __attribute__ ((unused)) for GCC) is used to annotate legitimately unused parameters, so that static analysis tools can tell legitimately unused parameters from truly unnecessary ones. To that end, some patches for external modules are also added, that are only applied when compiling with GCC and add necessary __attribute__ ((unused)) in headers.
2012-01-02Fix for fdo43460 Part XXVII getLength() to isEmpty()Olivier Hallot3-5/+5
Please find attached a partial fix for Easy Hack FDO43460 Part XXVII Modules padmin, pyuno, rdbmaker, regexp, registry, rsc, sal
2011-12-21catch exception by constant referenceTakeshi Abe8-51/+51
2011-12-12fix loading of python-uno module by package.module nameLionel Elie Mamane1-5/+6
2011-12-12more debug spew in non-debug productCaolán McNamara1-1/+1
2011-12-12Let's not have debug spew all the timeCaolán McNamara1-1/+1
2011-12-05Improve pyuno debugging in various ways - add stack traces etc.Michael Meeks3-5/+11
2011-12-05wizards: completely work python wizard packaging and registrationMichael Meeks1-3/+19
Also improve exception handling and error printing in pythonloader Allow registration of explicit .py components - the only method that works
2011-11-17Do build cpputools also for non-DESKTOP OSes, needed for bridgetestTor Lillqvist1-1/+1
2011-11-17Enforce a "lib" prefix for UNO components for AndroidTor Lillqvist1-1/+1
This commit for the old build system. (Don't bother for components not relevant for Android.) The Android package installer (as invoked through "adb install", from "ant debug install") silently ignores native libraries in app packages (.apk files) whose names don't start with "lib" and end with ".so". The package builder (as invoked through "ant debug") in the SDK gladly includes also thusly named native libraries in the .apk, though. Yay for consistency.
2011-11-10New LO_{LIB,JAVA}_DIR make special inbuild component handling superfluous.Stephan Bergmann1-1/+1
2011-10-24Undo basis/brand split: removed RPATH BRAND (same as OOO now).Stephan Bergmann1-1/+1
2011-10-24Undo basis/brand split: Move SDK and ure-link, remove OOO_BASE_DIR and ↵Stephan Bergmann2-11/+3
BaseInstallation.
2011-10-20Undo basis/brand split: moved Python from basis to brand.Stephan Bergmann4-50/+38
The Windows-only code in pyuno/zipcore/python.cxx may still need fixing.
2011-09-27Removed uses of rtl::O[U]String[Buffer]::operator sal_{char|Unicode} const *().Stephan Bergmann3-12/+16
2011-09-16Revert "Trying to chop out the uwinapi library"Fridrich Štrba1-0/+1
This reverts commit 2dea0dab4fafda3c10a5bd03ad15ed39a4658b51.
2011-09-16Trying to chop out the uwinapi libraryFridrich Štrba1-1/+0
2011-08-22Fix dbgutil build of pyunoTor Lillqvist1-6/+0
Whether things actually work when pyuno is built against the debugging runtime, but the internal Python (to the best of my knowledge) still against the normal one, I don't know. But anyway, this makes the build succeed here.
2011-08-21There was still overload ambiguousJulien Nabet1-2/+2
2011-08-21OUStringBuffer.append call: add explicit radix to disambiguate overloadLionel Elie Mamane1-1/+1
2011-08-21use boost::unordered_map instead of STL one so as not to require C++0xLionel Elie Mamane1-3/+3
2011-08-21pyuno: close opened parenthesis in raised exception messageLionel Elie Mamane1-0/+1
2011-08-21pyuno: some commentsLionel Elie Mamane1-1/+4
2011-08-21Janitorial: remove unnecessary const_castsLionel Elie Mamane7-50/+50
The python C API has consts at these places
2011-08-21Janitorial: remove unused variableLionel Elie Mamane1-1/+0