37016 2010-04-02 00:27:27 -0700 WebHTMLView.firstRectForCharacterRange returns wrong y for line beginning of multiline <pre> 2010-04-22 10:50:41 -0700 1 1 1 Unclassified WebKit WebKit API 528+ (Nightly build) PC OS X 10.5 UNCONFIRMED P3 Normal --- 0 morrita webkit-unassigned ap oldest_to_newest 207767 0 52394 morrita 2010-04-02 00:27:27 -0700 Created attachment 52394 reproduce (for DRT) WebHTMLView.firstRectForCharacterRange returns NSRect with wrong y value. To reproduce, run attached file on DRT, that will results following text: You can see a pump "y:600" on y value at the beginning of the line. ---- Ignore me; this block is just a place holder to make the test result significant. ABC DEF abc def 2 i:0 x:16 y:364 x:24 y:364 x:32 y:364 x:8 y:349 x:16 y:349 x:24 y:349 x:32 y:349 i:1 x:16 y:321 x:24 y:321 x:32 y:321 x:0 y:600 x:16 y:306 x:24 y:306 x:32 y:306 ---- This problem is observable only from WebKit API, and not visible from DOM. The test case uses textInputController.firstRectForCharacterRange() to reveal it. The expectation for LayoutTest/editing/selection/move-left-right.html is also suffered by this problem, although its difference is minor issue for that test case. 215960 1 ap 2010-04-22 10:50:41 -0700 A problem observable via textInputController.firstRectForCharacterRange() would probably affect inline text input, too. 52394 2010-04-02 00:27:27 -0700 2010-04-02 00:27:27 -0700 reproduce (for DRT) first-rect-after-set-position.html text/html 1557 morrita PGh0bWw+CjxoZWFkPgogICAgPG1ldGEgaHR0cC1lcXVpdj0iQ29udGVudC1UeXBlIiBjb250ZW50 PSJ0ZXh0L2h0bWw7IGNoYXJzZXQ9dXRmLTgiPgogICAgPHNjcmlwdD4KICAgICAgICBpZiAod2lu ZG93LmxheW91dFRlc3RDb250cm9sbGVyKQogICAgICAgICAgICBsYXlvdXRUZXN0Q29udHJvbGxl ci5kdW1wQXNUZXh0KCk7CgogICAgICAgIGZ1bmN0aW9uIGxvZyhtZXNzYWdlKQogICAgICAgIHsK ICAgICAgICAgICAgZG9jdW1lbnQuZ2V0RWxlbWVudEJ5SWQoImNvbnNvbGUiKS5hcHBlbmRDaGls ZChkb2N1bWVudC5jcmVhdGVUZXh0Tm9kZShtZXNzYWdlICsgIlxuIikpOwogICAgICAgIH0KCiAg ICAgICAgZnVuY3Rpb24gY2FyZXRDb29yZGluYXRlcygpCiAgICAgICAgewogICAgICAgICAgICBp ZiAoIXdpbmRvdy50ZXh0SW5wdXRDb250cm9sbGVyKQogICAgICAgICAgICAgICAgcmV0dXJuIHsg eDogMCwgeSA6MCB9OwogICAgICAgICAgICB2YXIgY2FyZXRSZWN0ID0gdGV4dElucHV0Q29udHJv bGxlci5maXJzdFJlY3RGb3JDaGFyYWN0ZXJSYW5nZSh0ZXh0SW5wdXRDb250cm9sbGVyLnNlbGVj dGVkUmFuZ2UoKVswXSwgMCk7CiAgICAgICAgICAgIHJldHVybiB7IHg6IGNhcmV0UmVjdFswXSwg eTogY2FyZXRSZWN0WzFdIH07CiAgICAgICAgfQogICAgICAgCiAgICAgICAgb25sb2FkID0gZnVu Y3Rpb24oKQogICAgICAgIHsKICAgICAgICAgICAgdmFyIHRlc3RzID0gZG9jdW1lbnQuZ2V0RWxl bWVudHNCeUNsYXNzTmFtZSgidGVzdCIpOwogICAgICAgICAgICB2YXIgc2VsID0gZ2V0U2VsZWN0 aW9uKCk7CiAgICAgICAgICAgIGxvZyh0ZXN0cy5sZW5ndGgpOwogICAgICAgICAgICBmb3IgKHZh ciBpID0gMDsgaSA8IHRlc3RzLmxlbmd0aDsgKytpKSB7CiAgICAgICAgICAgICAgICB0ZXN0c1tp XS5mb2N1cygpOwogICAgICAgICAgICAgICAgbG9nKCJpOiIgKyBpKTsKICAgICAgICAgICAgICAg IHNlbC5zZXRQb3NpdGlvbih0ZXN0c1tpXS5maXJzdENoaWxkLCAwKTsKICAgICAgICAgICAgICAg IGZvciAodmFyIGo9MDsgaiA8IDc7IGorKykgewogICAgICAgICAgICAgICAgICAgc2VsLm1vZGlm eSgibW92ZSIsICJmb3J3YXJkIiwgImNoYXJhY3RlciIpOwogICAgICAgICAgICAgICAgICAgdmFy IHBvcyA9IGNhcmV0Q29vcmRpbmF0ZXMoKTsKICAgICAgICAgICAgICAgICAgIGxvZygieDoiICsg cG9zLnggKyAiIHk6IiArICBwb3MueSk7CiAgICAgICAgICAgICAgICB9CiAgICAgICAgICAgIH0K ICAgICAgICB9CiAgICA8L3NjcmlwdD4KPC9oZWFkPgo8Ym9keT4KCjxkaXYgc3R5bGU9ImhlaWdo dDogMjAwcHg7Ij5JZ25vcmUgbWU7IHRoaXMgYmxvY2sgaXMganVzdCBhIHBsYWNlIGhvbGRlciB0 byBtYWtlIHRoZSB0ZXN0IHJlc3VsdCBzaWduaWZpY2FudC48L2Rpdj4KCjxwcmUgY2xhc3M9InRl c3QiIGNvbnRlbnRFZGl0YWJsZT5BQkMKREVGPC9wcmU+CjxwcmUgY2xhc3M9InRlc3QiIGNvbnRl bnRFZGl0YWJsZT5hYmMKPCEtLSAtLT5kZWY8L2Rpdj4KCjxwcmUgaWQ9ImNvbnNvbGUiPjwvcHJl Pgo8L2JvZHk+CjwvaHRtbD4K