975276 - [Keyboard]A cursor is not displayed on the head of a string by setComposition.

Status: RESOLVED FIXED

(Firefox OS Graveyard :: Gaia::Keyboard, defect)

Description

[Takuya Nakayama]

Attachment: Cursor_1_FxOS(bug).mp4

[steps to reproduce]
Set "0" to "cursor" of setComposition(DOMString text, optional long cursor), InputContext
  InputContext.setComposition(text, 0);
 * Refer to video.

[actual results]
Seemingly a cursor is displayed on the end of a string. It is not displayed on the head of it. 

[expected results]
A cursor is displayed on the head of a string.
Therefore it suggests an opportunity to a user for inputting a character from the head of the string.

Comment 1

[Takuya Nakayama]

Attachment: Cursor_2_Android(expected).mp4

Cursor_1_FxOS(bug).mp4
When "0" is set to "cursor" of "setComposition", a cursor moves to the end of a string.

Cursor_2_Android(expected).mp4
Expected result: When Android IME sets cursor location "0", a cursor locates the head of a string.  

Both of above cases, you can input a letter from the head of a string. 
However a blinking cursor location of "Cursor_1_FxOS(bug).mp4" is different from the expected result "Cursor_2_Android(expected).mp4".

Comment 3

[Masayuki Nakano [:masayuki] (he/him)(JST, +0900)]

http://mxr.mozilla.org/mozilla-central/source/dom/inputmethod/MozKeyboard.js#577
> 577         cursor: cursor || text.length,

Looks like this a simple mistake. If cursor is 0, the text.length is used unexpectedly.

We should avoid a case of |cursor === 0|.

Comment 4

[Masayuki Nakano [:masayuki] (he/him)(JST, +0900)]

Yuan, could you take a look? See comment 3.

Comment 5

[Takuya Nakayama]

Thank you for your reply.

We have confirmed the expected result after fixing the place pointed out with local environment.
We are waiting for a fixed source file in the future.

Comment 6

[Yuan Xulei [:yxl] (Leave from May 18)]

(In reply to Masayuki Nakano (:masayuki) (Mozilla Japan) from comment #4)
> Yuan, could you take a look? See comment 3.

It is a mistake. Let me fix it.

Comment 7

[Yuan Xulei [:yxl] (Leave from May 18)]

Attachment: setComposition.patch

It is a mistake. When cursor is not passed, we use the text length as cursor value. When cursor is passed, even 0 value, we should use the passed value, but never the text length.

Comment 8

[Ryan VanderMeulen [:RyanVM]]

Please run this through Try (and post a link here) before requesting checkin. Thanks!

Comment 9

[Yuan Xulei [:yxl] (Leave from May 18)]

(In reply to Ryan VanderMeulen [:RyanVM UTC-4][PTO June 19-22] from comment #8)
> Please run this through Try (and post a link here) before requesting
> checkin. Thanks!

The try link:

https://tbpl.mozilla.org/?tree=Try&rev=b05f4985ea0c

Comment 10

[Yuan Xulei [:yxl] (Leave from May 18)]

Try is green. Ask for checkin.
Thanks ;)

Comment 11

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/integration/b2g-inbound/rev/543203ca7ebe

Comment 12

[Wes Kocher (:KWierso) (Not reading bugmail; email directly if needed)]

https://hg.mozilla.org/mozilla-central/rev/543203ca7ebe