289857 - discarded characters cause that the selected text shift

Status: RESOLVED FIXED

(Core :: Layout: Text and Fonts, defect)

Description

[Hideo Saito]

Discarded characters (like soft hyphen (­)) cause that the selected text
shifts. I think that it is the cause that the mapping between the content and
the transformed buffer counts the discarded characters.

testcase: https://bugzilla.mozilla.org/attachment.cgi?id=39944

If 'ous' of the third table on testcase is selected and pasted and copied, other
texts will be copied.

Comment 1

[Hideo Saito]

Attachment: test patch

rbs, I propose changing the patch (attachment 179848 [details] [diff] [review]) of Bug 96423.
Please refer to the changes to nsTextFrame.cpp on attachment 177109 [details] [diff] [review] of Bug
166127. I think that this problem has same cause.

Comment 2

[Hideo Saito]

> I think that this problem has same cause.

Because a '\n' between CJ characters is discarded.

Comment 3

[rbs]

You don't have to introduce |mDiscarded|. |*aWasTransformed| should catch your
case because we should get (*aWordLenResult != *aContentLenResult). 

Try another patch without introducing |mDiscarded|.

Comment 4

[Hideo Saito]

Attachment: patch

To be sure, |mDiscarded| is unnecessary. I removed it.

Comment 5

[rbs]

You also don't need to move the #ifdef around, that will keep your patch small
and friendlier...

Comment 6

[rbs]

Oops... I just noted you indeed need to move it from the .cpp to the .h since
you use the IS_DISCARDED() macro elsewhere...

Comment 7

[rbs]

Comment on attachment 181229 [details] [diff] [review]
patch

I am having trouble applying your patch to my tree. Parts of the patch are
rejected. Care to update your tree to be in sync with the CVS trunk and
resubmit another patch.

Comment 8

[Hideo Saito]

Attachment: patch

I am sorry to trouble you. I posted this patch using source of 20-Apr-2005 on
latest trunk.

Comment 9

[rbs]

Comment on attachment 181382 [details] [diff] [review]
patch

r=rbs

Comment 10

[rbs]

Comment on attachment 181382 [details] [diff] [review]
patch

r+sr=rbs

Comment 11

[Hideo Saito]

Comment on attachment 181382 [details] [diff] [review]
patch

I think that this patch is low-risk although it influences the fix of Bug
96423, and it solves Bug 166127.

Comment 12

[chris hofmann]

Comment on attachment 181382 [details] [diff] [review]
patch

a=chofmann for 1.8a

Comment 13

[Hideo Saito]

rbs, this patch was approved for 1.8b2. Could you check in to the trunk?

Comment 14

[rbs]

Checked-in.

Comment 15

[Simon Montagu :smontagu]

*** Bug 245478 has been marked as a duplicate of this bug. ***

Comment 16

[Uri Bernstein]

*** Bug 241486 has been marked as a duplicate of this bug. ***