summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* Sean is passing Class::ISA maintenance to SteffenSteffen Mueller2009-09-221-2/+2
|
* Deprecate Class::ISA from coreSteffen Mueller2009-09-212-2/+4
|
* Remove Class::ISA use from autouse testsSteffen Mueller2009-09-212-3/+14
| | | | Instead, include an ultra-simple module for testing in t/lib.
* Remove use of Class::ISA from the debuggerSteffen Mueller2009-09-211-24/+15
| | | | | Instead, use mro::get_linear_isa which accomplishes the same feat. This allows us to remove another module from core.
* Upgrade to Attribute::Handlers 0.87 (which is just a core sync) -- for realSteffen Mueller2009-09-212-4/+25
|
* Revert "Upgrade to Attribute::Handlers 0.87 (which is just a core sync)"Steffen Mueller2009-09-214-625/+4
| | | | | This reverts commit 248f30289f80d02b83f3d80d7ff575cb3e758ced because I accidentally broke everything.
* Compare core to Attribute::Handlers 0.87 from CPANSteffen Mueller2009-09-211-1/+1
|
* Upgrade to Attribute::Handlers 0.87 (which is just a core sync)Steffen Mueller2009-09-214-4/+625
|
* Upgrade to AutoLoader 5.70 for CPAN synchronization (no non-trivial changes)Steffen Mueller2009-09-211-1/+1
|
* Compare AutoLoader in core to AutoLoader 5.70 from CPANSteffen Mueller2009-09-211-1/+1
|
* Compare PathTools to CPAN's 3.30_01Steffen Mueller2009-09-211-3/+2
|
* Fix my bash-ism in 9988b539d419b6c4 - test takes a single =, not ==.Nicholas Clark2009-09-211-3/+3
| | | | | Annoyingly bash just accepts it, and dash generates a diagnostic to stderr but carries on, hence why I didn't spot the problem initially.
* typo in usage of deprecateRobin Barker2009-09-212-2/+2
|
* Change my email addressRafael Garcia-Suarez2009-09-212-1/+2
|
* Move the Makefile targets _test{,_prep,_tty,_notty} into a script runtests.Nicholas Clark2009-09-203-46/+96
| | | | | This avoids a recursive call to $(MAKE) in the top level directory, which feels over-complicated when all the targets invoked have no dependencies.
* Adopt Makefile macros MINIPERL_EXE and MINIPERL from VMS, to reduce copy&paste.Nicholas Clark2009-09-201-51/+56
| | | | | '$(LDLIBPTH) $(RUN) ./miniperl$(EXE_EXT) -Ilib' becomes '$(MINIPERL)' 'miniperl$(EXE_EXT)' becomes '$(MINIPERL_EXE)'
* The pod convertors chdir post BEGIN-time, so don't need an absolute path for ↵Nicholas Clark2009-09-201-1/+1
| | | | | | | | lib. In fact, I don't see why they need to chdir at all. Or still need most of the self-extraction logic, now that we have MakeMaker. (Yes, the code is *that* old.)
* On Unix, use the -C option to mktables rather than an explicit chdir.Nicholas Clark2009-09-201-1/+1
| | | | This makes generating the -I options for the command line clearer.
* dropped a test by accident the last go, so ressurect the pat_re_eval.t ↵Yves Orton2009-09-196-202/+414
| | | | anyway, and resort and update the MANIFEST
* split t/re/pat.t into new piecesYves Orton2009-09-1911-22131/+243
|
* copy pat.t into five new filesYves Orton2009-09-195-0/+21955
| | | | | | next commit i will mutually gut them all, and this step will allow git to trivially identify that all of these files were copies of pat.t originally, and thus the blame log should show the full history.
* lib/unicore/mktables needs to get File::Spec from ext/Cwd/libNicholas Clark2009-09-193-3/+3
| | | | | | | | | It was only parallel make order roulette that it was working on my machine. VMS doesn't have this problem, as it's invoking mktables from the top level, using its -C option to change directory, and VMS's $(MINIPERL) has an -I option for [.ext.Cwd.lib] Maybe all platforms should swap to using the -C approach.
* win32/config_h.PL indirectly needs File::Spec, before XS extensions are built.Nicholas Clark2009-09-192-4/+4
|
* Reuinte File::Spec with the rest of the PathTools distribution in ext.Nicholas Clark2009-09-1921-38/+29
| | | | | | | | | | It can't really be renamed from ext/Cwd to ext/PathTools, because Configure and Makefile.SH need to know the name of the shared object produced, and they infer this from the name of the directory. This concludes the migration of modules from lib to ext. Exporter and version remain in lib, but I don't think that there is any benefit in trying to move either of them.
* 82dee7de92a0bd35 failed to add ext/lib/Makefile.PL. Oops.Nicholas Clark2009-09-181-0/+49
|
* Move lib from lib to ext.Nicholas Clark2009-09-189-29/+11
| | | | | Nothing up to and including minitest uses lib, and not needing many special case rules to extract from lib/lib_pm.PL simplifies all the Makefiles.
* Avoid re-creating a temporary file with constant contents for use.tNicholas Clark2009-09-183-6/+5
|
* Avoid using lib.pm in miniperl's tests.Nicholas Clark2009-09-184-36/+50
|
* Split XSLoader out from ext/DynaLoader into its own directory.Nicholas Clark2009-09-186-11/+79
|
* Make make_ext.pl set PERL_CORE=1 in the environment.Nicholas Clark2009-09-181-0/+1
| | | | | This simplifies the check in Makefile.PL, and means that "I'm building in the core" is passed down to build scripts invoked from the Makefile.
* Build DynaLoader in ext on VMS.Nicholas Clark2009-09-171-35/+4
|
* PERL_SRC already has brackets on VMS.Craig A. Berry2009-09-171-1/+1
| | | | Follow-up to 28161c97e726a0cbc7d3e30460aa0bf491d83b97.
* In S_mro_get_linear_isa_c3() replace sv_inc() with code for the cases we need.Nicholas Clark2009-09-171-3/+9
|
* Now no need to run AutoSplit on lib on Win32, with DynaLoader builing in extNicholas Clark2009-09-174-35/+2
| | | | | This reverts the win32 part of commit 9139c7231c70a9388d718abc41e91edaf43422ec, reinstating the win32 part of commit 109e4020ef40828991be28fb05d9f269b4d92530.
* Convert win32 to build DynaLoader with EU::MM in ext/DynaLoaderNicholas Clark2009-09-175-58/+28
| | | | | With help from Steve Hay. This converges the win32 build process with the ./Configure-based approach.
* Since gitignores are advisory, we don't need the specific unignore patternsRafael Garcia-Suarez2009-09-171-102/+1
| | | | | So we can ignore all Makefile.PL and all *.c files, since the ones we want are already added.
* Revert "much simpler .gitignore for ext/"Rafael Garcia-Suarez2009-09-171-8/+108
| | | | | | | The negative patterns don't appear to un-ignore correctly what they should (at least with git version 1.6.0.4) This reverts commit 594c64d48c897be7522145d9f634599aae38b149.
* Rearrange ext/Memoize to the same layout as the CPAN distribution.Nicholas Clark2009-09-1610-12/+8
|
* Remove the now-redundant MAP for Math::BigInt.Nicholas Clark2009-09-163-6/+1
| | | | | Undo a local rename of a test that didn't have an 8.3 name, but is still unique when truncated to 8.3
* Rearrange ext/File-Temp to the same layout as the CPAN distribution.Nicholas Clark2009-09-163-4/+1
|
* /Compress/ modules are at version 2.021. Remove vestigal MAPs and comments.Nicholas Clark2009-09-161-12/+4
|
* Rearrange ext/IO-Zlib to the same layout as the CPAN distribution.Nicholas Clark2009-09-163-5/+1
|
* Linker options file not created before nonxs extension building.Craig A. Berry2009-09-161-1/+1
|
* Upgrade to IO::Zlib 1.10Rafael Garcia-Suarez2009-09-162-3/+3
|
* Start to make DynaLoader's Makefile.PL platform agnostic.Nicholas Clark2009-09-161-11/+35
|
* Revert 109e4020 and 2da760ce for now. Win32 needs to AutoSplit DynaLoader.pmNicholas Clark2009-09-166-3/+65
| | | | | | | | | | Khan! Win32 is special. It builds DynaLoader by hand, copies in DynaLoader.pm (and XSLoader.pm) to lib, and *then* runs, and relies on, the general AutoSplit to split them. It would be better if everything standardised on building DynaLoader via its Makefile.PL and MakeMaker. This reverts commit 2da760ceef1d2d90b3141bdeb239c7d7580be05a. This reverts commit 109e4020ef40828991be28fb05d9f269b4d92530.
* Localize $@ and $! before loading a file in SWASHNEWRafael Garcia-Suarez2009-09-161-0/+2
| | | | | | This fixes a bug where a spurious error was reported from utf8_heavy. This been found by Salvador Ortiz Garcia who suggested to localize $@; I merely added $!.
* 109e4020 failed to delete the clone of splittree.pl in the NetWare directory.Nicholas Clark2009-09-151-24/+0
| | | | | | (I wonder how much of the rest of the NetWare directory is a copy-and-edit of the win32 directory. I wonder how far NetWare is off working. I wonder if anyone still uses NetWare.)
* No need to run AutoSplit on lib on Win32 or NetWare either - see ebca06e9.Nicholas Clark2009-09-155-41/+3
|
* Ensure that the perl build doesn't read any installed sitecustomize script.Nicholas Clark2009-09-155-5/+5
| | | | | | | But not the tests yet. We need to avoid running a sitecustomize script as it may have side effects that cause the build to fail, or worse, complete but with buggy side effects.
View Lorry Controller Status