100173 - Thai language selection broken.

Status: RESOLVED WORKSFORME

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

Description

[Prabhat Hegde]

For a description about this bug, please see:
http://bugzilla.mozilla.org/show_bug.cgi?id=100170

Have filed a separate bug for clarity.

Once again patch attached below :
Index: nsTextFrame.cpp
===================================================================
RCS file: /cvsroot/mozilla/layout/html/base/src/nsTextFrame.cpp,v
retrieving revision 1.318
diff -u -r1.318 nsTextFrame.cpp
--- nsTextFrame.cpp 2001/08/21 01:47:08 1.318
+++ nsTextFrame.cpp 2001/09/17 21:39:43
@@ -83,6 +83,10 @@
 //ahmed end
 #endif // IBMBIDI

+#ifdef SUNCTL
+#include "nsILE.h"
+#endif
+
 #ifndef PR_ABS
 #define PR_ABS(x) ((x) < 0 ? -(x) : (x))
 #endif
@@ -2239,6 +2243,37 @@
   while (sdptr){
     sdptr->mStart = ip[sdptr->mStart] - mContentOffset;
     sdptr->mEnd = ip[sdptr->mEnd]  - mContentOffset;
+
+#ifdef SUNCTL
+    static NS_DEFINE_CID(kLECID, NS_ULE_CID);
+
+    nsCOMPtr<nsILE> mCtlObj;
+    mCtlObj = do_CreateInstance(kLECID, &rv);
+    if (NS_FAILED(rv)) {
+      NS_WARNING("Cell based Cursor Movement Will Not Be Supported\n");
+      mCtlObj = nsnull;
+    }
+    else {
+      PRInt32 mStart, mEnd;
+
+      if (sdptr->mEnd < textLength) {
+    mCtlObj->GetRangeOfCluster(text, PRInt32(textLength), sdptr->mEnd,
+                   &mStart, &mEnd);
+    if (sdptr->mStart > sdptr->mEnd) /* Left Edge */
+      sdptr->mEnd = mStart;
+    else
+      sdptr->mEnd = mEnd;
+      }
+
+      /* Always start selection from a Right Edge */
+      if (sdptr->mStart > 0) {
+    mCtlObj->GetRangeOfCluster(text, PRInt32(textLength), sdptr->mStart
,
+                   &mStart, &mEnd);
+    sdptr->mStart = mEnd;
+      }
+    }
+#endif
+
 #ifdef IBMBIDI // Simon - display substrings RTL in RTL frame on non-Bidi platf
orm
     if ((level & 1) && !(isRightToLeftOnBidiPlatform)) {
       PRInt32 swap  = sdptr->mStart;
@@ -3894,12 +3929,31 @@
 #endif
         aPos->mContentOffset = 0;
     PRInt32 i;

Please review and comment.

prabhat.

Comment 1

[Roy Yokoyama]

bstell/shanjian: can one of you review the patch?

Comment 2

[Andreas Becker]

Switching QA contact to katakai for now, please re-assign further as appropriate.

Comment 3

[Roy Yokoyama]

assiging to bstell

Comment 4

[kill this account]

--> ftang

Comment 5

[Frank Tang]

smontagu- please take a look at this one.

Comment 6

[Frank Tang]

nsILE is not ready for all platform yet. move to m98

Comment 7

[Prabhat Hegde]

Thanks to Roland Mainz this feature has been putback into trunk (Ref: Bugzilla 
#84380)

Comment 8

[Simon Montagu :smontagu]

Marking fixed since the patch has been checked in. Who can verify that the
current Thai selection behaviour is correct?

Comment 9

[Masaki Katakai]

On my Solaris build, it seems working fine for me. But
I'll assign prabhat to QA person for verification.
Prabhat, can you try on your build?

Comment 10

[Prabhat Hegde]

Verified on Solaris with --enable-ctl flag, Checked that selection is by
"cluster". Arthit could you independently verify this as thai user and
also Mark bug as VERIFIED?

Comment 11

[Arthit Suriyawongkul]

i can't verify this prabhat,
since it still has a bug.

please see details in bug 167983.
[CTL-Thai] text with Thai/non-Thai combination, highlighting is not match with
selection

Comment 12

[Arthit Suriyawongkul]

can i reopen this?

Comment 13

[Arthit Suriyawongkul]

also this guaranteed the cell-based selection with keyboard only
(<shift>+arrow keys)

but NOT covers the case of selection by dragging mouse.

i've tested by dragging mouse to select Thai text,
still able to select only part of cell.

i will reopen this bug.

Comment 14

[Samphan Raruenrom]

Attachment: A text input field with 5 grapheme clusters

You can test the selection problem with this minimal HTML test case. See also
bug 283416 - Windows : Selection must be done by grapheme cluster boundaries

Comment 15

[Keng Susumpow]

Tested on Mac OS X 10.5.2 with Minefield nightly trunk build 2008030111 seems this bug has been fixed.

Comment 16

[Isriya Paireepairit]

I reproduce the step posted in bug 283416, can confirm Patipat comment #15 that this bug has been fixed.

Nightly build 2008030204 Mac OS X 10.5.2

Comment 17

[Arthit Suriyawongkul]

no problem with Firefox 3 beta 3 on Linux (Ubuntu 8.04 Alpha 4)

Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9b3) Gecko/2008021416 Firefox/3.0b3

Comment 18

[Keng Susumpow]

Works fine on Windows XP.

Tested with Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9b4pre) Gecko/2008030305 Minefield/3.0b4pre

Comment 19

[Arthit Suriyawongkul]

seems to already works fine on Windows XP, Mac OS X, and Linux.

we should be able to close this,
but as this bug is originally filed as bug on Solaris, should wait response from Solaris side.

Comment 20

[Arthit Suriyawongkul]

BugAThon Bangkok.

all confirmed this works ok. close.