458 Commits

Author SHA1 Message Date
Jeremy Ashkenas
60f80e2698 Fixing the double-implicit-call-into-implicit-object problem. 2010-09-15 22:29:03 -04:00
Jeremy Ashkenas
c782c2ec1c Fix for issue #685. Over-aggressize heredoc cleanup when there is no leading indentation. 2010-09-15 22:08:13 -04:00
Jeremy Ashkenas
a3c224e57a Fixing issue #676, chained accesses against the super() keyword. 2010-09-12 16:33:38 -04:00
Jeremy Ashkenas
38ce0cfd9a running cacheindexes on soaks that need them. 2010-09-12 16:18:05 -04:00
Jeremy Ashkenas
61a39e04fc Issue #680. @::prop versus this::prop, fixed lexing regex. 2010-09-12 15:48:31 -04:00
Jeremy Ashkenas
3a20d7dacb Partial fix for 653. 2010-09-12 11:08:05 -04:00
Timothy Jones
e5837b4ee9 Actually caching the splatted constructor this time. 2010-09-12 01:49:25 +12:00
Timothy Jones
d1f31c5143 Ensuring constructors invoked with splats behave correctly, along with caching. 2010-09-12 01:28:22 +12:00
Jeremy Ashkenas
4af41e9bfb Fixing issue #678 -- missing parentheses in a mixed operation(call(soak)) 2010-09-11 08:39:14 -04:00
Jeremy Ashkenas
d41a414b5c Complexifying the indebt test, one more level. 2010-09-08 22:48:28 -04:00
Jeremy Ashkenas
9bd3cca7c4 Introducing the notion of 'indebt' to mirror 'outdebt', but for suppressed indentation with trailing operators etc. Issue #639. 2010-09-08 22:46:13 -04:00
Jeremy Ashkenas
44618d5765 Adding a test for Issue #669 2010-09-08 20:25:17 -04:00
Jeremy Ashkenas
70cfd54ad4 Issue #670. 'THIS' tokens should trigger an implicit call. 2010-09-08 20:15:16 -04:00
Jeremy Ashkenas
eb9a524ea1 Adding unmatched 'then' to the list of things that closes a single-line implicit call early. Issue #611. 2010-08-30 22:04:13 -04:00
Jeremy Ashkenas
0caa731291 re-enabling the mis-dented call case. Issue #657 2010-08-30 20:59:16 -04:00
Jeremy Ashkenas
704fbf499b Merging in Tesco and rofrankel's soaks for function calls. A soaked check that fails will return undefined. 2010-08-28 09:00:04 -04:00
Richard Frankel
bcecbd051b added some tests for function soak 2010-08-25 18:54:42 -04:00
Jeremy Ashkenas
fa95f743f3 Fixing Issue #643. Be a little bit safer about declaring block variables as close to the block scope as possible. 2010-08-24 22:19:53 -04:00
Jeremy Ashkenas
bd3471b3d1 Issue #641. Rename __superClass__ to __super__ 2010-08-23 21:19:43 -04:00
Jeremy Ashkenas
117204a784 Revert "Issue #619. 'new' operator misbehavior."
This reverts commit e7834de929.
2010-08-21 19:16:02 -04:00
Jeremy Ashkenas
e7834de929 Issue #619. 'new' operator misbehavior. 2010-08-21 18:54:24 -04:00
Jeremy Ashkenas
d5d5de55ae Fixing the longstanding and important Issue #637. I'm glad this one got got. 2010-08-21 12:13:43 -04:00
Jeremy Ashkenas
13adc44867 Issue #633. 'not instanceof' 2010-08-21 08:56:25 -04:00
Jeremy Ashkenas
4ddd65a4c4 Issue #621. Added the ability to leave the start and end index off of arrays. list[5..] is now valid CoffeeScript, slicing to the end of the array. 2010-08-18 21:51:44 -04:00
Jeremy Ashkenas
24f1174b16 Adding a test for issue #631. 2010-08-18 20:42:23 -04:00
Jeremy Ashkenas
098caa9979 Revert "Fixes Issue #618. Close implicit arguments in the middle of param lists."
This reverts commit 474c372b17.
2010-08-18 20:39:55 -04:00
Jeremy Ashkenas
1b05cd81f0 Enabling bound functions as static members of classes... Issue #627 2010-08-17 21:07:36 -04:00
Jeremy Ashkenas
a8c6a641d7 stop supressing left-hand whitespace in newline'd strings. Make 'em more literal. 2010-08-16 22:02:37 -04:00
Jeremy Ashkenas
474c372b17 Fixes Issue #618. Close implicit arguments in the middle of param lists. 2010-08-16 21:19:34 -04:00
Jeremy Ashkenas
5b9b45814f Fixing Issue #622 with a light refactor to start using a 'tags' property on all nodes. 2010-08-16 20:18:42 -04:00
Jeremy Ashkenas
7a0d95c612 Issue #575. Allowing implicit objects to force implicit calls. 2010-08-14 19:56:00 -04:00
Jeremy Ashkenas
129e950c59 Fixing naked chained existential checks -- the associativity recently got reversed. 2010-08-14 16:33:20 -04:00
Jeremy Ashkenas
c90a75ebc5 Fixing first part of #614 -- improperly cached existential operator, when first operand is a function call. 2010-08-14 16:24:05 -04:00
Jeremy Ashkenas
c71f2794eb Fixing compound assignment as a right-hand sub expression of a larger operation. Invalid in JS, valid in Coffee. 2010-08-14 14:43:25 -04:00
Jeremy Ashkenas
ead9b1041c tweaking existential compilation. 2010-08-14 11:29:10 -04:00
Jeremy Ashkenas
1eebbfe2bc allowing more flexible whitespace-started regexes. You have to wrap 'em in parens or assign to 'em, but it's better than nothing. 2010-08-12 20:24:53 -04:00
Jeremy Ashkenas
fd6e9a1e66 Adding support for compound assignment to indented implicit objects. 2010-08-11 23:14:50 -04:00
Jeremy Ashkenas
d704afa0e9 Fixes Issue #574 ... number keys in implicit object literals. 2010-08-11 21:49:18 -04:00
Jeremy Ashkenas
2ec857ef8d Fixing implicit object literals with leading string keys. 2010-08-11 21:41:29 -04:00
Jeremy Ashkenas
ba02ebc3dc Fixes Issue #603 -- a longstanding precedence issue involving prefix vs postfix if and unless, involving using the entire single-line if statment as an expression. 2010-08-11 21:28:22 -04:00
Jeremy Ashkenas
398ec3be5a Issue #602. Adding JS's Bitwise Assignment operators. 2010-08-11 19:24:59 -04:00
Jeremy Ashkenas
642e7fada5 adding other implicit indentation/call/object test. 2010-08-11 00:27:01 -04:00
Jeremy Ashkenas
7142131f3d adding herecomment in class definition test. 2010-08-10 23:06:06 -04:00
Jeremy Ashkenas
cb52be6475 adding a top-level object literal test for Issue #542 2010-08-10 22:18:15 -04:00
Jeremy Ashkenas
18d6fd72de First step of the general rewriter refactor. Added a generic 'detectEnd' function which is expression-pair sensitive. Use it to reimplement closeOpenCallsAndIndexes 2010-08-08 17:37:28 -04:00
Jeremy Ashkenas
d286b33601 Fixing closures-generated-within-comprehensions to preserve block local scope for range comprehensions 2010-08-08 10:52:59 -04:00
Jeremy Ashkenas
b902377304 Fixing Issue #589. Compound assignment to an operation should have lower precedence. 2010-08-08 00:07:00 -04:00
Jeremy Ashkenas
1c903450c8 Merging in Stan's patch for a --require flag that allows you to hook into CoffeeScript compiler events for on 'compile', 'success', and 'failure' ... a better way to extend the language or do growlnotify stuff. Issue #590 2010-08-07 23:33:35 -04:00
Jeremy Ashkenas
0ada5a7734 Issue #587 ... Removing the Attempted Removal of Erroneous Double Parens. 2010-08-07 08:13:36 -04:00
Jeremy Ashkenas
9df3e6a538 first step towards requiring #{ ... } interpolation -- removing naked interps from the compiler. 2010-08-07 08:02:16 -04:00