diff options
| author |   Jean Chalard <jchalard@google.com> | 2013-12-27 15:44:29 +0900 |
|---|---|---|
| committer | Jean Chalard <jchalard@google.com> | 2014-01-22 14:41:45 +0900 |
| commit | 2b38b5e8e60900e45a9d5d591ba6f7f7788e535c (patch) | |
| tree | b0ae779e89fcdb8b7494423ae4c90203f89f231d /java/src/com/android/inputmethod/research | |
| parent | 591580ee8a084d06343b95e409931d05c7269e3c (diff) | |
| download | latinime-2b38b5e8e60900e45a9d5d591ba6f7f7788e535c.tar.gz latinime-2b38b5e8e60900e45a9d5d591ba6f7f7788e535c.tar.xz latinime-2b38b5e8e60900e45a9d5d591ba6f7f7788e535c.zip | |
[IL62] Pull up X,Y processing, step 1
This is fine because getKey{X,Y} is idempotent for any
non-keyboard coordinate value.
However this makes a net change : the x,y values passed to
LatinImeLoggerUtils.onNonSeparator are now different.
The point is however that they used to be wrong. The logged
values used not to account for the keyboard padding that
is present on tablets, and in the simulating tools we don't
know about that padding so we couldn't use the coordinates.
The catch here is that other calls like LoggerUtils.onSeparator
should follow suit, but this is too big a change to do it at once.
Follow-up changes will fix them too.
Bug: 8636060
Change-Id: If4b3d3cb1ed4b44c35f23e66aba3b5797236bba7
Diffstat (limited to 'java/src/com/android/inputmethod/research')
0 files changed, 0 insertions, 0 deletions