44600 2010-08-25 04:50:16 -0700 REGRESSION: Splitting an empty string sometimes returns an empty array 2010-08-25 20:10:14 -0700 1 1 1 Unclassified WebKit JavaScriptCore 528+ (Nightly build) Mac (Intel) OS X 10.6 RESOLVED FIXED GoogleBug, InRadar, Regression P1 Normal --- 1 john msaboff ap barraclough commit-queue ggaren msaboff oliver oldest_to_newest 269155 0 john 2010-08-25 04:50:16 -0700 In the console: 'a'.split(/\s+/) ["a"] ''.split(/\s+/) [""] ''.split(/\s+/) [] Notice the last two are identical, but give different answers. Other browsers and older webkit give the [""]. This seems to break some pages of Google apps for your domain admin. 269302 1 ap 2010-08-25 09:54:17 -0700 This is a regression from Safari 5.0.1. 269421 2 ggaren 2010-08-25 11:57:39 -0700 <rdar://problem/8354205> 269449 3 msaboff 2010-08-25 12:55:30 -0700 This bug is due to the single entry regexp cache. That regexp cache has been removed in https://bugs.webkit.org/attachment.cgi?id=65336 as part of resolving https://bugs.webkit.org/show_bug.cgi?id=44555. A test for this corner case will be added. 269527 4 65478 msaboff 2010-08-25 14:58:35 -0700 Created attachment 65478 Patch with new regression test for this issue. The underlying problem was fixed in bug https://bugs.webkit.org/show_bug.cgi?id=44555. 269693 5 65478 commit-queue 2010-08-25 20:10:08 -0700 Comment on attachment 65478 Patch with new regression test for this issue. Clearing flags on attachment: 65478 Committed r66075: <http://trac.webkit.org/changeset/66075> 269694 6 commit-queue 2010-08-25 20:10:14 -0700 All reviewed patches have been landed. Closing bug. 65478 2010-08-25 14:58:35 -0700 2010-08-25 20:10:08 -0700 Patch with new regression test for this issue. 44600.patch text/plain 2557 msaboff SW5kZXg6IExheW91dFRlc3RzL0NoYW5nZUxvZwo9PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09Ci0tLSBMYXlvdXRUZXN0cy9D aGFuZ2VMb2cJKHJldmlzaW9uIDY2MDQ2KQorKysgTGF5b3V0VGVzdHMvQ2hhbmdlTG9nCSh3b3Jr aW5nIGNvcHkpCkBAIC0xLDMgKzEsMTQgQEAKKzIwMTAtMDgtMjUgIE1pY2hhZWwgU2Fib2ZmICA8 bXNhYm9mZkBhcHBsZS5jb20+CisKKyAgICAgICAgUmV2aWV3ZWQgYnkgTk9CT0RZIChPT1BTISku CisKKyAgICAgICAgQWRkZWQgbmV3IHRlc3QgdG8gY2hlY2sgZm9yIHJlZ3Jlc3Npb24gYXMgcmVw b3J0ZWQgaW4KKyAgICAgICAgaHR0cHM6Ly9idWdzLndlYmtpdC5vcmcvc2hvd19idWcuY2dpP2lk PTQ0NjAwLgorCisgICAgICAgICogZmFzdC9qcy9zY3JpcHQtdGVzdHMvc3RyaW5nLXNwbGl0LWRv dWJsZS1lbXB0eS5qczogQWRkZWQuCisgICAgICAgICogZmFzdC9qcy9zdHJpbmctc3BsaXQtZG91 YmxlLWVtcHR5LWV4cGVjdGVkLnR4dDogQWRkZWQuCisgICAgICAgICogZmFzdC9qcy9zdHJpbmct c3BsaXQtZG91YmxlLWVtcHR5Lmh0bWw6IEFkZGVkLgorCiAyMDEwLTA4LTI0ICBSeW9zdWtlIE5p d2EgIDxybml3YUB3ZWJraXQub3JnPgogCiAgICAgICAgIFJldmlld2VkIGJ5IFRvbnkgQ2hhbmcu CkluZGV4OiBMYXlvdXRUZXN0cy9mYXN0L2pzL3N0cmluZy1zcGxpdC1kb3VibGUtZW1wdHktZXhw ZWN0ZWQudHh0Cj09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT0KLS0tIExheW91dFRlc3RzL2Zhc3QvanMvc3RyaW5nLXNwbGl0 LWRvdWJsZS1lbXB0eS1leHBlY3RlZC50eHQJKHJldmlzaW9uIDApCisrKyBMYXlvdXRUZXN0cy9m YXN0L2pzL3N0cmluZy1zcGxpdC1kb3VibGUtZW1wdHktZXhwZWN0ZWQudHh0CShyZXZpc2lvbiAw KQpAQCAtMCwwICsxLDExIEBACitUaGlzIHRlc3QgY2hlY2tzIGZvciBhIHJlZ3Jlc3Npb24gYWdh aW5zdCBzcGxpdCgpIGZ1bmN0aW9uIHJldHVybnMgd3JvbmcgYW5zd2VyIGZvciBzZWNvbmQgZW1w dHkgc3BsaXQuCisKK09uIHN1Y2Nlc3MsIHlvdSB3aWxsIHNlZSBhIHNlcmllcyBvZiAiUEFTUyIg bWVzc2FnZXMsIGZvbGxvd2VkIGJ5ICJURVNUIENPTVBMRVRFIi4KKworCitQQVNTICIiLnNwbGl0 KC9zKy8pIGlzIFsiIl0KK1BBU1MgIiIuc3BsaXQoL3MrLykgaXMgWyIiXQorUEFTUyBzdWNjZXNz ZnVsbHlQYXJzZWQgaXMgdHJ1ZQorCitURVNUIENPTVBMRVRFCisKSW5kZXg6IExheW91dFRlc3Rz L2Zhc3QvanMvc3RyaW5nLXNwbGl0LWRvdWJsZS1lbXB0eS5odG1sCj09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT0KLS0tIExh eW91dFRlc3RzL2Zhc3QvanMvc3RyaW5nLXNwbGl0LWRvdWJsZS1lbXB0eS5odG1sCShyZXZpc2lv biAwKQorKysgTGF5b3V0VGVzdHMvZmFzdC9qcy9zdHJpbmctc3BsaXQtZG91YmxlLWVtcHR5Lmh0 bWwJKHJldmlzaW9uIDApCkBAIC0wLDAgKzEsMTMgQEAKKzwhRE9DVFlQRSBIVE1MIFBVQkxJQyAi LS8vSUVURi8vRFREIEhUTUwvL0VOIj4KKzxodG1sPgorPGhlYWQ+Cis8bGluayByZWw9InN0eWxl c2hlZXQiIGhyZWY9InJlc291cmNlcy9qcy10ZXN0LXN0eWxlLmNzcyI+Cis8c2NyaXB0IHNyYz0i cmVzb3VyY2VzL2pzLXRlc3QtcHJlLmpzIj48L3NjcmlwdD4KKzwvaGVhZD4KKzxib2R5PgorPHAg aWQ9ImRlc2NyaXB0aW9uIj48L3A+Cis8ZGl2IGlkPSJjb25zb2xlIj48L2Rpdj4KKzxzY3JpcHQg c3JjPSJzY3JpcHQtdGVzdHMvc3RyaW5nLXNwbGl0LWRvdWJsZS1lbXB0eS5qcyI+PC9zY3JpcHQ+ Cis8c2NyaXB0IHNyYz0icmVzb3VyY2VzL2pzLXRlc3QtcG9zdC5qcyI+PC9zY3JpcHQ+Cis8L2Jv ZHk+Cis8L2h0bWw+CkluZGV4OiBMYXlvdXRUZXN0cy9mYXN0L2pzL3NjcmlwdC10ZXN0cy9zdHJp bmctc3BsaXQtZG91YmxlLWVtcHR5LmpzCj09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT0KLS0tIExheW91dFRlc3RzL2Zhc3Qv anMvc2NyaXB0LXRlc3RzL3N0cmluZy1zcGxpdC1kb3VibGUtZW1wdHkuanMJKHJldmlzaW9uIDAp CisrKyBMYXlvdXRUZXN0cy9mYXN0L2pzL3NjcmlwdC10ZXN0cy9zdHJpbmctc3BsaXQtZG91Ymxl LWVtcHR5LmpzCShyZXZpc2lvbiAwKQpAQCAtMCwwICsxLDggQEAKK2Rlc2NyaXB0aW9uKAorJ1Ro aXMgdGVzdCBjaGVja3MgZm9yIGEgcmVncmVzc2lvbiBhZ2FpbnN0IDxhIGhyZWY9Imh0dHBzOi8v YnVncy53ZWJraXQub3JnL3Nob3dfYnVnLmNnaT9pZD00NDYwMCI+c3BsaXQoKSBmdW5jdGlvbiBy ZXR1cm5zIHdyb25nIGFuc3dlciBmb3Igc2Vjb25kIGVtcHR5IHNwbGl0PC9hPi4nCispOworCitz aG91bGRCZSgnIiIuc3BsaXQoL3MrLyknLCAnWyIiXScpOworc2hvdWxkQmUoJyIiLnNwbGl0KC9z Ky8pJywgJ1siIl0nKTsKKwordmFyIHN1Y2Nlc3NmdWxseVBhcnNlZCA9IHRydWU7Cg==