- Sep 12, 2011
-
-
Mark Miller authored
-
Mark Miller authored
etc, because the substitution logic in test262.py operates on the source after prepending framework.js. The plan is to replace sputnikLib.js with framework.js. Also, the substitution logic in test262.py should probably be fixed to apply the substitution before prepending.
-
Mark Miller authored
testRunner, test262.py, that can run all the converted tests.
-
Mark Miller authored
assertTrue became assertTruthy. Canonical form of doc-comment attribute value source no longer have a leading colon and trailing semicolon. Any existing such are stripped during conversion.
-
- Sep 11, 2011
-
-
Mark Miller authored
preparation for deriving a test262 command-line runner.
-
- Sep 09, 2011
-
-
Mark Miller authored
style to append runTestCase(funcName); rather than assertTrue(funcName.call(this)), giving the testing framework more control.
-
- Sep 08, 2011
-
-
Mark Miller authored
-
Mark Miller authored
-
Mark Miller authored
morning. The rewrite of "function testcase()..." now puts the assertTrue at the bottom. Preconditions, names, and ids are removed. And the path in the file is ignored, and is instead set accurately according to where the file is found.
-
- Sep 07, 2011
-
-
Mark Miller authored
-
Mark Miller authored
-
Mark Miller authored
Much rougher: A first attempt at building the testcases for the website based on the new canonical test262 style testcase.
-
Mark Miller authored
-
Mark Miller authored
-
Mark Miller authored
-
- Jul 18, 2011
-
-
David Fugate authored
-
David Fugate authored
-
David Fugate authored
-
David Fugate authored
-
David Fugate authored
-
- Jul 08, 2011
-
-
David Fugate authored
15.4.4.15-3-14, 15.4.4.15-3-29, and 15.4.4.15-3-8 were disabled with no reason or bug given. These actually look OK, and pass on all browsers...
-
- Jul 07, 2011
-
-
David Fugate authored
-
David Fugate authored
if (!strict_mode) { throw new SyntaxError('unspecified case'); } This doesn't work for a few reasons: 1. 'strict_mode' is undefined in the test case exeuction context. That is, the test framework does not force the test case into strict mode 2. The test case code looks like: "use strict"; if (!strict_mode) ... ... Note that 'strict_mode' is not set before being accessed Really the metadata for these test cases should have a "@strict" property added to instruct the test framework *not* to run the test if an implementation doesn't support strict mode.
-
David Fugate authored
-
David Fugate authored
-
David Fugate authored
-
David Fugate authored
-
- Jul 06, 2011
-
-
David Fugate authored
Disabled 9 invalid tests.
-
- Jul 05, 2011
-
-
David Fugate authored
-
David Fugate authored
-
- Jul 01, 2011
-
-
David Fugate authored
-
https://bugs.ecmascript.org/show_bug.cgi?id=87David Fugate authored
on the live website to be sure.
-
- Jun 30, 2011
-
-
David Fugate authored
-
David Fugate authored
host objects (which make them invalid for test262). Using this, I found one benign 'alert' usage (it was never even called) and removed it.
-
David Fugate authored
-
- Jun 29, 2011
-
-
David Fugate authored
-
David Fugate authored
-
David Fugate authored
New Sputnik 'negative' metadata that included exception type was causing problems for SputnikConverter. Took the easy fix and split the metadata into two: 'negative' and 'errortype'.
-