59924 - Win IME for Japanese: the candidate window is incorrectly positioned covering the current line

Status: VERIFIED FIXED

(Core :: DOM: Editor, defect, P3)

Description

[Katsuhiko Momoi]

** Observed with 11/8/2000 Win32 MN build **
** This a left over problem from Bug

This bug has been filed to correct a remaining problem in
the positioning of Windows Japanese IME candidate window
relative to the current line of editing.

This problem does not occur if the IME candidate window is
placed below the current line.

On a composer window, if you are editing a line which is
placed at least 2/3 down from the 1st line, then the candidate
window is placed above the current line of editing, this is when
it covers the current line. This seems to be a left-over problem
from Bug 53072.
CC'ing a bunch of people in that bug.

shanjian, is this your bug?

Comment 1

[Katsuhiko Momoi]

Attachment: An image file illustrating the problem.

Comment 2

[Shanjian Li]

The problem is different from bug 53072, and the same problem exist in 4.x.
The cursor position we find is correct, but it seems that we have to adjust
this position when pass it to win IME. Currently we just add cursor height. 
It seems we have to check current screen position before we do this or not. 
I assumed it is windows' responsibility. Need to investigate more.

Comment 3

[Teruko Kobayashi]

Changed QA contact to [e-mail].

Comment 4

[Shanjian Li]

Add kato san to cc list.

Kato-san, Can you help?

Comment 5

[Teruko Kobayashi]

Changed QA contact to [e-mail].

Comment 6

[Shanjian Li]

Attachment: proposed fix

Comment 7

[Roy Yokoyama]

I like it shanjian. Instead of specifying the position of the upper left 
corner of candidate window (by useing CFS_CANDIDATEPOS), we now use
CFS_EXCLUDE for not to cover the composition window area.

One question: any reason why 
candForm.rcArea.right = candForm.rcArea.left = candForm.ptCurrentPos.x?

Please answer above question and /r=yokoyama on the patch.

Comment 8

[Shanjian Li]

The candidate form windows appears either up or down relate to cursor, so there is really
no reason to caculate the width of cursor.

Comment 9

[Shanjian Li]

erik, this is another one need your sr=.

Comment 10

[Erik van der Poel]

sr=erik

Comment 11

[Shanjian Li]

set milestone

Comment 12

[Shanjian Li]

checked in  the fix.

Comment 13

[sujay]

Kat/Shanjian, please verify this bug and mark VERIFIED-FIXED..thanks!

Comment 14

[Katsuhiko Momoi]

ylong should be able to verify this bug.

Comment 15

[sujay]

ylong, please verify....thanks!

Comment 16

[Yuying Long]

Fixed verified on 03-22 Mtrunk build.  Sorry for late but since the fix was 
checked in yesterday, I couldn't verify it on any build before today :)