summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* Renamed functions to reduce confusion and added to header file. tip_257_implementation_branchdkf2006-10-312-4/+6
|
* Added more API as requested by Arnulf Wiedermann <arnulf@wiedermann-pri.de> dkf2006-10-311-1/+40
| | | | Also added more comments.
* Rewritten to not pass function pointers through ClientData directly; somedkf2006-10-231-22/+53
| | | | compilers/architectures do not like it at all.
* Use internal TclListObjGetElements macro where appropriate.dkf2006-10-232-4/+4
|
* Add some hacks for Itcl supportdkf2006-10-225-78/+159
|
* Fixed the remaining TODO, made the call chains more efficient, reworked the ↵dkf2006-10-225-378/+567
| | | | introspectors to be ensembles.
* Fix problems identified by AKu, and finish moving [oo::copy] to where it belongsdkf2006-10-216-1373/+1385
|
* Clarify; metadata applies special interpretation to NULL so don't call ↵tip_257_implementation_branch_patchpoint_20061020T1300dkf2006-10-191-3/+3
| | | | clientData
* Add a metadata mechanism (C API only; scripts can use variables) that allowsdkf2006-10-193-6/+318
| | | | | code to attach arbitrary non-NULL pointers to objects and classes. Allows extensions (like Itcl) to store data in objects/classes without major core changes.
* Strip bits that I'm never planning to enable.dkf2006-10-191-5/+2
|
* Move the copy-object operation to a place where it fits better and expose itdkf2006-10-153-175/+307
| | | | as a part of the public API.
* fix line endingsdgp2006-10-115-5928/+5928
|
* tclOOCall.c build targetsdgp2006-10-111-1/+5
|
* Makefile updates for new filedkf2006-10-084-7/+10
|
* Following experiments, now have (possibly imperfect implementation of)dkf2006-10-081-1232/+1246
| | | | | class mixins. Factor out the call chain management code to make that easier to work on. Add more comments to changes in non-OO core.
* Following experiments, now have (possibly imperfect implementation of)dkf2006-10-086-4649/+4976
| | | | | class mixins. Factor out the call chain management code to make that easier to work on. Add more comments to changes in non-OO core.
* Many improvements to introspection; all features should now be introspectable.dkf2006-10-044-233/+279
|
* Add some memory testsdkf2006-10-041-1/+37
|
* Tidy up long linesdkf2006-10-041-3390/+3391
|
* Fixed jenglish's nasty filter recursion case. Filters can't be recursively ↵dkf2006-10-033-10/+110
| | | | | | | | | entered *unless* they're done so through progressing to an inner non-filter method through [next]. This gives the real intent of "no recursive entry" without the inconvenient blocking of it entirely. The net effect is that filter authors still have to watch themselves, but not in a surprising way. :-)
* fix line endingsdgp2006-10-031-309/+309
|
* Following discussion on c.l.t, remove oo::definer and oo::struct completely.dkf2006-10-022-524/+44
|
* Remove oo::definer and oo::struct from the global state; they're not ↵dkf2006-10-022-353/+335
| | | | important enough.
* fix line endingsdgp2006-10-021-315/+315
|
* The C API is now just about theredkf2006-10-016-217/+514
| | | | | | Done the cross-consistency check with the TIP Uncovered more bugs Main remaining problem: inconsistency in class filter chain processing
* Reorganize to make traces neaterdkf2006-09-306-589/+781
| | | | | Bug fix in object->unknown when handling zero-method objects More tests, more cleanup, more comments (I love it!)
* Tidying up, and starting to build a C API ready for exposure.dkf2006-09-293-55/+209
|
* Convert confusing 'cPtr' to clearer 'clsPtr'dkf2006-09-284-3788/+3792
|
* Stop people from creating objects with empty names; it doesn't work.dkf2006-09-273-4455/+4465
|
* Regendkf2006-09-271-3/+3
|
* TclRenameCommand should take const argumentsdkf2006-09-272-7/+7
|
* Finished off [self] and fixed up the mess that was class mixins to no longer ↵dkf2006-09-251-4/+33
| | | | | | have odd lurking nasties.
* Remove parameters; I'm not going to implement them (for now)dkf2006-09-256-3312/+3517
| | | | Add class filters, as I understand how to do this now.
* Oops, wrong length in comparison produces random results in qsort() when one dkf2006-09-221-2/+2
| | | | value is a prefix of another.
* Filled out more [self] subcommandsdkf2006-09-221-7/+100
|
* Added many comments; code should be more readable now.dkf2006-09-201-2606/+3178
|
* fix line endingsdgp2006-09-201-288/+288
|
* More extensive (complete?) testing of the oo::struct classdkf2006-09-202-51/+285
| | | | | Working and tested 'variable' method of oo::object class Added FOREACH use
* Use the FOREACH macrodkf2006-09-201-27/+31
|
* Move FOREACH declaration to more useful placedkf2006-09-192-5/+10
|
* oo::struct now seems to be working, but more tests needed.dkf2006-09-182-60/+281
| | | | Also arranged for all (public) ::oo commands to be exported.
* Remove a useless epoch fielddkf2006-09-172-13/+125
| | | | Add (most) of the 'struct' class definition; still needs tests
* Add some dependency info for conveniencedkf2006-09-171-1/+5
|
* Some tidying, and fix remaining known call context generation and trace ↵dkf2006-09-024-97/+308
| | | | generation bugs.
* More functionality, more tests, more bugs uncovered...dkf2006-09-014-20/+114
|
* fix line endingsdgp2006-09-011-819/+819
|
* Improvements in error trace handlingdkf2006-09-014-923/+953
|
* More tests (all passing), and got rid of the knownBug from all tests.dkf2006-09-015-301/+433
|
* More introspection and starting to assemble the [oo::definer] metaclassdkf2006-08-313-2824/+2957
|
* Fix bloopersdkf2006-08-311-5/+3
|
generated by cgit v0.12 at 2025-08-14 18:52:07 (GMT)