Commit Graph

  • 1556d076bd got token insertion working parrt 2011-10-02 16:15:02 -0800
  • c33b06fdde bug fix; no lexer if no lex rules parrt 2011-10-02 15:21:32 -0800
  • 2a34d00d09 tweak parrt 2011-10-02 13:18:10 -0800
  • 05360be562 more cleanup parrt 2011-10-02 13:17:06 -0800
  • 02d424a92b tweak parrt 2011-10-02 12:58:53 -0800
  • 49571e386f err clean up parrt 2011-10-02 12:51:33 -0800
  • 298a15e017 Parse error systemTweaks, removing the debug API since we won't use it parrt 2011-10-02 12:01:35 -0800
  • 9894221249 rename test. tweak errs parrt 2011-10-01 10:31:36 -0800
  • 6b4e9905fb rename member, tweak output. new chk for inf loop from error sync works parrt 2011-09-30 20:40:14 -0800
  • 60c9fe76c0 almost got new error mech in parrt 2011-09-30 20:22:09 -0800
  • 9e69640bc1 added more unit tests parrt 2011-09-27 17:09:17 -0800
  • 82360bfaa8 clean up parrt 2011-09-27 16:23:15 -0800
  • c535d45f66 tweak parrt 2011-09-25 15:11:05 -0800
  • fd3f482375 fixed regression in atn interp parrt 2011-09-25 15:10:21 -0800
  • cd2c4e622d fix bug in code gen for * loops parrt 2011-09-25 14:43:37 -0800
  • e3d6db0478 more unit tests parrt 2011-09-25 14:10:30 -0800
  • f02db87c02 got nullable kids tree parsing working. got $rule.text in tree parsing working parrt 2011-09-24 17:12:22 -0800
  • e8a2a738cf rename TreeNodeStream parrt 2011-09-24 09:33:04 -0800
  • 93581cfa98 got v3 tests working in v4 parrt 2011-09-24 09:14:22 -0800
  • 149033fb2e add more tests from v3 parrt 2011-09-24 08:41:15 -0800
  • 9c5f636bea labels work with tree parsers now. parrt 2011-09-22 17:32:25 -0800
  • b63b271815 got basic tree parser working. parrt 2011-09-22 17:16:22 -0800
  • 0c1716d062 reorg to get tree parser gen'd parrt 2011-09-20 16:30:20 -0800
  • a126e2043a add file parrt 2011-09-20 16:22:25 -0800
  • 6c45753bea got basic tree parser ^(A B) stuff in. parrt 2011-09-18 16:49:52 -0800
  • a8195e5d7c playing parrt 2011-09-11 16:20:42 -0800
  • a1c682434d removes the individual stacks to track rule invocations. $a::x now looks up the context stack for rule index RULE_a. they're not done very often and so it's better to say all of the overhead pushing and popping the contexts. parrt 2011-09-11 16:12:01 -0800
  • 0e608ab0a4 clean up unit tests parrt 2011-09-11 15:16:53 -0800
  • 64ccba4168 rebuilt the code generator so that I generate a separate class for each labeled alternative and modify the listener so that it's always enter and exit rule; the type of the argument that distinguishes between them. I now create the overall rule context and then copy it into the alternative context if they have a label. parrt 2011-09-11 15:09:55 -0800
  • bf19465437 Got a rule labels added to the tree and also altered code generation so that it pays attention to the labels. fairly major surgery but it's a nice refactoring. parrt 2011-09-09 17:08:13 -0800
  • f13fd1973e mv ruleIndex, add alt parrt 2011-09-09 13:00:58 -0800
  • 62937ca639 rename visit->walk parrt 2011-09-08 16:52:30 -0800
  • 2d234ee28c Got the visitor thing cleaned up enough for a small example parrt 2011-09-08 13:45:05 -0800
  • da7e7c8813 clean up tree output, rename _ctx to context parrt 2011-09-08 12:12:51 -0800
  • b671cb1fba add treeTextProvider and various "visual" properties to TreeViewer Udo Borkowski 2011-09-08 10:05:23 -0800
  • 0e7911f99c add parrt 2011-09-07 20:47:39 -0800
  • c16782c848 tweak parrt 2011-09-07 17:13:32 -0800
  • 297087770e scope listener name in output parrt 2011-09-07 17:01:59 -0800
  • c79848cca0 gen listener file; reorg file generation parrt 2011-09-07 16:51:10 -0800
  • 4f2beaf69d rename discover/finish parrt 2011-09-07 15:04:35 -0800
  • a440d4e988 add TreeViewer Udo Borkowski 2011-09-06 08:45:28 -0800
  • 1e595a486a add parrt 2011-09-06 06:31:44 -0800
  • 730794128e parse tree visiting; now, we ALWAYS generate a return struct definition: rule_ctx. parrt 2011-09-05 17:08:26 -0800
  • d3dc566a69 tweak parrt 2011-09-05 13:11:33 -0800
  • 99d5d64efb update ant build parrt 2011-09-05 13:05:58 -0800
  • afb767b99c should not be in depot parrt 2011-09-05 12:54:27 -0800
  • 0f74909c8c mv ruleIndex parrt 2011-09-05 12:49:54 -0800
  • f93bc7aa7a add org.abego.treelayout Udo Borkowski 2011-09-05 10:50:57 -0800
  • 528d6df1e3 added list of rule consts; parser ctx tracks rule index parrt 2011-09-05 08:28:24 -0800
  • 01020ee1bb more cleanup parrt 2011-09-04 15:55:45 -0800
  • 5c35165873 update the comment parrt 2011-09-04 13:49:40 -0800
  • 32a375bbf7 I split the tree interface into multiple simpler ones and removed the methods that can modify a node. Either you should use the adapter or, you know the concrete type and then you can call those methods directly not through the interface. The whole point of the interface is to be flexible but then for consistency users must use the adapters. I'm also starting to pull helper methods into a Trees "module". I also renamed CommonTree to be CommonAST. I also renamed the adapters appropriately. parrt 2011-09-04 12:02:07 -0800
  • 783f1bc38f Split token into read-only and writable portions. added release method. moved the methods around that get input streams and token streams between interfaces. simplify the character stream interface by moving the gas line and get characterization methods to the lexer. parrt 2011-09-03 17:36:14 -0800
  • 5c6f35abef did some reorganization parrt 2011-09-03 11:52:23 -0800
  • c1aa0fcd10 New parse tree mechanism parrt 2011-09-03 09:18:26 -0800
  • 104f39acde working on parse trees parrt 2011-09-02 17:15:36 -0800
  • 7aad57fe49 add parrt 2011-09-02 16:18:25 -0800
  • 62549b65bf add dbg classes. add mark/rewind parrt 2011-09-02 16:17:37 -0800
  • 085dd05bf1 got parser nongreedy loops going parrt 2011-08-12 19:51:12 -0800
  • f44c49a8b8 got lexer nongreedy loops going parrt 2011-08-07 11:19:34 -0800
  • a9782118ac add parrt 2011-08-04 08:56:08 -0800
  • d926ec9661 WOOT! all left-recursive tests pass parrt 2011-07-31 18:14:47 -0800
  • ec5d74c83e removed all traces of FORCED_ACTION parrt 2011-07-31 12:36:07 -0800
  • a7f2bc31ac ok, got new simple semantics in for sempreds; no forced actions and preds always eval unless ctx dependent, in which case they need to bein original rule. parrt 2011-07-31 11:58:59 -0800
  • 328444244f backing out changes; tried 2 stacks; too complicated parrt 2011-07-31 10:38:48 -0800
  • d2651a50a6 backing out changes; tried 2 stacks; too complicated parrt 2011-07-31 10:38:23 -0800
  • c713b3dd2d rm print parrt 2011-07-30 17:22:56 -0800
  • e32bae3276 rebuilt ATN structure for ()* and ()+; two decisions each now unless single alt. parrt 2011-07-30 17:20:57 -0800
  • cf54eea3b3 set type parrt 2011-07-30 15:25:14 -0800
  • 64e12dbfa5 add method parrt 2011-07-30 11:49:49 -0800
  • 2ebedba421 debugging parrt 2011-07-29 15:45:31 -0800
  • 0168e0837a finished getting unit tests for action/pred exec during ATN simulation parrt 2011-07-29 14:49:16 -0800
  • b90b073450 getting more unit tests in parrt 2011-07-29 14:05:19 -0800
  • 4b5cd3a9ba We're forced to exec args, even if dependent on _localctx. If no actual context to use, create dummy context to use for arg eval only. parrt 2011-07-29 11:45:27 -0800
  • dcfb804ed3 added method to get context for non-arg rules parrt 2011-07-28 19:36:56 -0800
  • cce6c70b2e fixed context in sempred. don't exec forced actions either if in outer context and dependent. parrt 2011-07-28 15:15:41 -0800
  • 3d15daccad got predicate issue resolved. executes preds if not dependent on context and no dipping into outer context parrt 2011-07-28 12:43:28 -0800
  • 0cb91817d9 not tested but put locals into attr lookup for rule and code gen chunks + java.stg parrt 2011-07-28 10:29:12 -0800
  • 10d7a79324 reorg; cleanup. added ast.g ptr setting to nodes created during left-recursion elim. parrt 2011-07-27 17:03:54 -0800
  • e084c12483 rm gated sem pred and other weird variants parrt 2011-07-27 16:50:13 -0800
  • 9328c164af set type parrt 2011-07-27 16:45:02 -0800
  • 2db09cdeb9 rm scopes for rules/grammar, added locals [int i, int j] for rules. i do not generate anything for the LOCALS tree at moment parrt 2011-07-27 16:43:16 -0800
  • 2136287966 tweak parrt 2011-07-27 15:55:30 -0800
  • 078e2ab47c java.g working again. parrt 2011-07-26 10:56:24 -0800
  • fd3f8b17c9 ~sets weren't working; got closer to java.g working again. parrt 2011-07-25 18:09:00 -0800
  • c6fa2fce6e decided to clean up a bit, backtracking from updating current to have inner/outer context. will try with full ctx again. parrt 2011-07-25 16:49:33 -0800
  • 76ca9ba0d5 a bit of clean up parrt 2011-07-24 16:24:18 -0800
  • a29feac1bf a bit of clean up parrt 2011-07-24 16:18:21 -0800
  • 613c802c8c rm dead code parrt 2011-07-24 15:26:52 -0800
  • b4b02bb813 cleaned up atn construction / set detection / unit tests. still not done full context in parser simulation parrt 2011-07-24 15:25:17 -0800
  • 1fcc12a7ad changes to ATN simulator just before I try adding context back in. parrt 2011-07-24 13:20:49 -0800
  • 66da1f723a got unit tests working again for interp; got args funcs in output parrt 2011-07-24 11:48:43 -0800
  • d51784a250 got lexer actions working again parrt 2011-07-23 12:58:09 -0800
  • ba9f5d54e7 tweak parrt 2011-07-22 18:18:43 -0800
  • 3b52eb4e6a snapshot parrt 2011-07-22 17:19:26 -0800
  • 40f9e867cd 23 import grammar unit tests pass parrt 2011-07-20 18:55:23 -0800
  • 36f3590f6c got rule props in for current rule; fixed yet another freakin bug in implicit lexer extraction. parrt 2011-07-19 19:18:16 -0800
  • 6dab71e160 symbol issue test working again parrt 2011-07-19 16:44:35 -0800
  • 66933e36a2 got token aliases checks done parrt 2011-07-19 11:54:37 -0800
  • c6904bb21b imports get merged into root grammar now. parrt 2011-07-18 13:31:24 -0800