brendan@mozilla.org
0b62c0a1e4
Patch from nanto@moon.email.ne.jp for misordered alternates in string-lexing regexp, and lack of IE /[/]/ compat in regexp-lexing regexp (433831, r=me, NPOTB).
2008-05-21 15:38:18 -07:00
mrbkap@gmail.com
3e0670e952
Make peek-on-same-line actually work correctly. bug 401246, r=brendan (NPOTB)
2007-12-18 17:15:41 -08:00
brendan@mozilla.org
c14ddb9adf
Fix post-increment/decrement automatic semicolon insertion bug (401466, r=mrbkap).
2007-11-04 14:21:19 -08:00
brendan@mozilla.org
db5d699473
Fix conditional expressions (401405, r=mrbkap).
2007-10-29 19:35:02 -07:00
brendan@mozilla.org
df567b52fb
Include $ and _ in identifier chars (still to do: Unicode alnums and escapes); based on patch from Frankie Robertson <frankierobertson5@googlemail.com> (399625, r=mrbkap).
2007-10-16 16:43:28 -07:00
brendan@mozilla.org
5f7baa7b6c
Fix from Frankie Robertson <frankierobertson5@googlemail.com> to Node toString (399659, r/a=me).
2007-10-16 15:56:16 -07:00
mrbkap@gmail.com
8fe65a69cb
Correctly parse regular expressions with the 'm' and 'y' flags. bug 378486, r=brendan
2007-05-18 18:57:30 -07:00
brendan@mozilla.org
eca0bc73e1
Fix blunder in reworking of how SpiderMonkey detects a label statement (reported by Steve Yegge).
2007-05-05 21:34:24 -07:00
mrbkap%gmail.com
f50ceb399b
Use __applyConstructor__ to allow for an arbitrary number of parameters to be passed to native constructors. bug 331429, r=brendan
2006-05-09 19:23:39 +00:00
mrbkap%gmail.com
a9e8f95b55
Use the right object constructors. bug 334024, r=brendan
2006-04-18 18:12:53 +00:00
brendan%mozilla.org
33af4e7a97
Add exec-time ecmaStrictMode flag and use it to allow 'for (i in null)' and the like.
2006-04-15 05:02:08 +00:00
mrbkap%gmail.com
ee5ae83517
Only avoid calling toObject when v is null. bug 333911, r=brendan
2006-04-14 18:14:43 +00:00
mrbkap%gmail.com
cde0d2439f
Don't throw when given |for (i in null)| to match SpiderMonkey and to match users' expectations. bug 333911, r=brendan
2006-04-13 22:49:56 +00:00
mrbkap%gmail.com
dc0e278951
Fixing some silliness from my last checkin, tokenstr takes a tokentype, not the token itself. r=brendan
2006-03-22 06:23:59 +00:00
mrbkap%gmail.com
1d4b9a7b0a
Handle assignment inside the then part of ternary expressions correctly according to the ECMAScript grammar. bug 330975, r=brendan
2006-03-22 03:00:03 +00:00
mrbkap%gmail.com
5c696de243
Don't forget to throw syntax errors for missing parentheses or brackets. bug 331127, r=brendan
2006-03-20 20:57:51 +00:00
mrbkap%gmail.com
f60971fc05
Don't pollute String.prototype. bug 330976, r=brendan
2006-03-20 20:15:15 +00:00
mrbkap%gmail.com
a92ebcfd67
Part of bug 300736: Reflect Narcissus' evaluate in the global object since we expose it natively in the shell now. r=brendan
2005-07-14 07:22:58 +00:00
mrbkap%gmail.com
6d682be208
bug 299898: Get Narcissus running the js tests. This adds a way for it to load scripts (using snarf) relative to the calling script (such as js.js). r=brendan a=shaver
2005-07-13 16:49:02 +00:00
mrbkap%gmail.com
f47551589c
bug 300534: Internal Narcissus functions show up in objects' prototypes. This fixes a failure on the js testsuite and also fixes Narcissus' failure on SpiderMonkey's bug 299641. r=brendan
2005-07-13 02:46:10 +00:00
brendan%mozilla.org
e4a82727bb
Add dummy or first params to constructors so they have the right arity (length property).
2005-07-07 21:47:59 +00:00
brendan%mozilla.org
71eab9c0ac
Keep Activation objects from being polluted by native Object.prototype props (using bug 299898, r=mrbkap).
2005-07-07 07:29:21 +00:00
mrbkap%gmail.com
b6459f7951
Narcissus can now run the Javascript tests, to some extent. Fix a couple of bugs found by the js tests (we were unable to parse the generated function resulting from new Function() and String.fromCharCode was not defined). evaluate() now returns its argument if it isn't a string (to catch things like evaluate(load(...))).
2005-07-07 00:53:44 +00:00
brendan%mozilla.org
67eeb2edef
Rename FunctionDeclaration to FunctionDefinition, since it handles declarations, statements, and expressions (mimicing ECMA).
2005-07-07 00:21:09 +00:00
brendan%mozilla.org
31cabefc0f
Fix expression statement (SEMICOLON) nodes to start and end at the same source index as their expression child.
2005-07-02 17:26:00 +00:00
mrbkap%gmail.com
423cdcda66
Provide snarf and load functions.
2005-07-02 04:55:29 +00:00
brendan%mozilla.org
c77919f477
Use non-greedy closure to match C-style comments.
2005-07-01 23:00:35 +00:00
brendan%mozilla.org
1056b77193
Never mind\!
2005-07-01 20:48:23 +00:00
brendan%mozilla.org
a48f461e71
Add function statements, an ECMA Ed. 3 extension supported by SpiderMonkey.
2005-06-30 18:16:19 +00:00
brendan%mozilla.org
0efbaf9c4b
One more followup fix: __defineProperty__ should probably be global.
2005-06-12 18:45:53 +00:00
brendan%mozilla.org
241e1ec5d6
Followup to last checkin to delete constructor from Activation.prototype.
2005-06-12 18:28:33 +00:00
brendan%mozilla.org
9c322522a7
Mirror fix for bug 290774 from SpiderMonkey to Narcissus.
2005-06-12 16:08:21 +00:00
brendan%mozilla.org
2158473744
Fix IDENTIFIER resolution to use in, not hasOwnProperty, since we are metacircular on __proto__.
2004-05-17 23:10:53 +00:00
brendan%mozilla.org
8dcdcb1f89
Parallel fix to SpiderMonkey's bug 238945, to automatically insert ; after do-while on any error token.
2004-03-30 03:25:17 +00:00
brendan%mozilla.org
533fdfabe2
Fiddle comment.
2004-02-25 22:34:53 +00:00
brendan%mozilla.org
b6b0241e94
Use label local in Statement; eliminate somewhat arbitrary empty lines.
2004-02-25 09:06:36 +00:00
brendan%mozilla.org
0bc43fbec1
Duh.
2004-02-24 21:23:45 +00:00
brendan%mozilla.org
df0ecadb4b
s/compile/parse/ -- compile is overgeneral, especially in light of what's next.
2004-02-24 03:04:40 +00:00
brendan%mozilla.org
c4044d863e
Simplify consts generation slightly.
2004-02-24 03:03:33 +00:00
brendan%mozilla.org
09acb4babf
Add ecmaStrictMode to CompilerContext.
2004-02-22 07:21:55 +00:00
brendan%mozilla.org
c0ad926a1b
Elaborate reference error message; fix iloop in __hasInstance__.
2004-02-20 07:35:28 +00:00
brendan%mozilla.org
962663d3d9
t.scanOperand should be true by default, eliminating scanForOperand bloat.
2004-02-17 08:43:48 +00:00
brendan%mozilla.org
5632fed18a
s/this/v/g in getValue.
2004-02-16 16:55:40 +00:00
brendan%mozilla.org
3ddaa18b05
- Add source and cursor properties to SyntaxError exceptions.
...
- Rework labeled statements to fix chicken-and-egg problem with BREAK to label
completion type targeting.
- Unify next-statement non-expression token cases and default: in Expression.
2004-02-16 02:38:19 +00:00
brendan%mozilla.org
42c4aff0a0
- Fix global.String to convert its argument to string type when called as
...
a function.
- Rework labeled statements to fix chicken-and-egg problem with BREAK to label
completion type targeting.
- Avoid defining Function.prototype.__call__, __construct__, and
__hasInstance__ once they're set, to avoid storing host-uninterpretable
target function objects in those properties.
- Fix evaluate to throw a host exception if returning to the host interpreter.
2004-02-16 02:15:27 +00:00
brendan%mozilla.org
02939f0679
Fully bootstrapped, but host still can't apply target functions.
2004-02-14 21:58:10 +00:00
brendan%mozilla.org
76d7770850
Fix Expression to work with Statement's automatic semicolon insertion logic.
2004-02-14 21:56:41 +00:00
brendan%mozilla.org
70d3411bc0
Fix Expression to stop on : only if hookLevel is not elevated.
2004-02-11 08:30:20 +00:00
brendan%mozilla.org
f3068c76bf
Checkpoint, still not quite metacircularly bootstrapped.
2004-02-11 07:57:19 +00:00
brendan%mozilla.org
bd08f327fb
- ECMA-prescribed {DontEnum} attributes for length on empty arg list, and on
...
<any-function>.prototype.constructor.
- Put rarely-set default properties in prototypes.
- Improve a couple of comments.
2004-02-03 23:58:03 +00:00
