Charles Lowell
|
6e505d3e9e
|
decouple the interceptors from the context.
|
2010-08-30 14:36:21 -05:00 |
|
Charles Lowell
|
6a09f4b060
|
move constructors and embedded constructors to the portal.
|
2010-08-30 14:36:21 -05:00 |
|
Charles Lowell
|
e5bdf2b410
|
make embedded constructors per context. one more failing spec!!!
|
2010-08-30 14:36:21 -05:00 |
|
Charles Lowell
|
e4d96e24ed
|
fix segfaulting testcases
|
2010-08-30 14:36:21 -05:00 |
|
Charles Lowell
|
948de0a5f3
|
pass the dontintercept block down when delegating to the access implementation.
|
2010-08-30 14:36:21 -05:00 |
|
Charles Lowell
|
1be713d53d
|
refactor value converter to be per-context.
|
2010-08-30 14:36:21 -05:00 |
|
Charles Lowell
|
c22bbd9177
|
break out access into instance based scheme
|
2010-08-30 14:36:21 -05:00 |
|
Charles Lowell
|
f1d24232c2
|
coalesce ruby access into one interface.
|
2010-08-30 14:36:21 -05:00 |
|
Charles Lowell
|
3050e77046
|
named property query.
|
2010-08-30 14:36:21 -05:00 |
|
Charles Lowell
|
6f235622f2
|
bug in named property enumerator
|
2010-08-30 14:36:20 -05:00 |
|
Charles Lowell
|
3324dfb9eb
|
syntax fix for 1.8.6 which doesn't allow for passing blocks to blocks.
|
2010-08-06 11:17:17 -05:00 |
|
Charles Lowell
|
801a4ea7be
|
separate v8 ruby access from ruby access in preparation for making it pluggable.
|
2010-08-06 10:23:13 -05:00 |
|
Charles Lowell
|
1b5ea69f67
|
allow default ruby access property interceptors to specify which properties they do *not* intercept.
|
2010-08-05 11:37:49 -05:00 |
|
Charles Lowell
|
c1909ad39c
|
add rubysend() as replacement for rubycall where appropriate.
|
2010-08-02 17:03:14 -05:00 |
|
Charles Lowell
|
bcb7da3965
|
protect indexed access to ruby objects to prevent segfaults in the event that they raise an exception.
|
2010-08-02 16:58:21 -05:00 |
|
Charles Lowell
|
f56a58ab78
|
don't enumerate ruby property accessors (foo=) as javascript properties.
|
2010-08-02 16:50:43 -05:00 |
|
Charles Lowell
|
87b1982464
|
protect calls to dynamic accessors and interceptors so that if they throw ruby exceptions, it raises them in javascript instead of causing segfaults.
|
2010-08-02 16:44:55 -05:00 |
|
Charles Lowell
|
cd512d3135
|
fix breakage when V8 module is not included in the top level scope.
|
2010-07-08 10:10:19 -05:00 |
|
Charles Lowell
|
25f1d91b3f
|
add dynamic property access via the [] and []= methods.
|
2010-06-29 08:22:41 +03:00 |
|
Charles Lowell
|
4e83c7b889
|
convert method symbols to strings for 1.9 compatibility.
|
2010-06-29 07:47:08 +03:00 |
|
Charles Lowell
|
626b36cf5c
|
support for indexed property access
|
2010-06-29 07:38:23 +03:00 |
|
Daniël van de Burgt
|
8191e68c2e
|
added support for inherited properties
|
2010-06-28 18:02:54 +02:00 |
|
Charles Lowell
|
765fffb5c0
|
begin work on indexed property access
|
2010-06-28 16:36:41 +03:00 |
|
Charles Lowell
|
cfc12665b8
|
always return the value when we intercept on a setter. Has no effect, but it does get rid of an annoying warning.
|
2010-06-18 14:59:54 +03:00 |
|
Charles Lowell
|
222053b901
|
ability to embed classes as constructors directly into the context.
|
2010-06-09 00:59:22 +03:00 |
|
Charles Lowell
|
ca4eff3658
|
separate Constructors from normal object access
|
2010-06-09 00:32:31 +03:00 |
|
Charles Lowell
|
68e8fc1454
|
fixes for 1.8.6: tap.rb implementation didn't actually return self. Symbol#to_proc not defined until 1.8.7
|
2010-05-28 12:58:21 +03:00 |
|
Charles Lowell
|
ab33a88411
|
wrap calling from v8 -> rb so that we get uniform stack treatment
|
2010-05-27 18:30:19 +03:00 |
|
Charles Lowell
|
ae9c88bad4
|
keep ruby objects alive as long as V8 has a reference to them, even if ruby does not currently have a reference. rethrow ruby exceptions as javascript exceptions.
|
2010-05-26 19:54:56 +03:00 |
|
Charles Lowell
|
025fedcc23
|
move ruby property access to its own module
|
2010-05-22 12:12:42 +03:00 |
|