Below are two commits that were necessary to make (Linux) "make check" work again post <http://cgit.freedesktop.org/libreoffice/core/commit/?id=efb23f29983f87104a684e7fab00b84fc59d131d> "Integrate branch of IAccessible2." In both cases, the problematic changes made by that commit looked rather unmotivated to me, and I do not know whether it would actually have been better to change the code under test or the test code. So, insight welcome.

Stephan

On 11/27/2013 09:15 AM, Stephan Bergmann wrote:
commit 1533129759c0017413efe2213bfe503c26dc16f0
Author: Stephan Bergmann <sberg...@redhat.com>
Date:   Wed Nov 27 09:12:38 2013 +0100

     Revert sense in VCLXAccessibleBox::getAccessibleActionCount again

     ...this had been switched from "m_bIsDropDownBox ? 1 : 0" to 
"m_bIsDropDownBox ?
     0 : 1" in efb23f29983f87104a684e7fab00b84fc59d131d "Integrate branch of
     IAccessible2" for no apparent reason and broke JunitTest_testtools_unoapi.

     Change-Id: I276d2f452648387a5fa693349cf868787b2d48e1

diff --git a/accessibility/source/standard/vclxaccessiblebox.cxx 
b/accessibility/source/standard/vclxaccessiblebox.cxx
index 7aa7836..764c0ca 100644
--- a/accessibility/source/standard/vclxaccessiblebox.cxx
+++ b/accessibility/source/standard/vclxaccessiblebox.cxx
@@ -409,7 +409,7 @@ sal_Int32 SAL_CALL 
VCLXAccessibleBox::getAccessibleActionCount (void)

      // There is one action for drop down boxes (toggle popup) and none for
      // the other boxes.
-    return m_bIsDropDownBox ? 0 : 1;
+    return m_bIsDropDownBox ? 1 : 0;
  }

  sal_Bool SAL_CALL VCLXAccessibleBox::doAccessibleAction (sal_Int32 nIndex)

On 11/27/2013 11:33 AM, Stephan Bergmann wrote:
commit 4f94f16ba15218e5e7a9eb4d72ddb4cb62884dbb
Author: Stephan Bergmann <sberg...@redhat.com>
Date:   Wed Nov 27 11:28:57 2013 +0100

     IAccessible2: revert some bounds-checking changes again

     ...that were introduced with c23ab5eaed57cc3fb5860e26c591c73c5b22687b 
"Integrate
     branch of IAccessible2" but broke JunitTest_sc_unoapi, which expects calls 
to
     getCharacterAttributes and getIndexAtPoint to actually fail for 
out-of-bounds
     arguments.  The above commit does not make it obvious why those changes 
were
     actually made.

     Change-Id: I3492c3e6a49d68885df1206e9368f4483c0dc4f9

diff --git a/editeng/source/accessibility/AccessibleEditableTextPara.cxx 
b/editeng/source/accessibility/AccessibleEditableTextPara.cxx
index e9941df..614294f 100644
--- a/editeng/source/accessibility/AccessibleEditableTextPara.cxx
+++ b/editeng/source/accessibility/AccessibleEditableTextPara.cxx
@@ -1381,11 +1381,7 @@ namespace accessibility
          EBulletInfo aBulletInfo = rCacheTF.GetBulletInfo( static_cast< 
sal_uInt16 >(GetParagraphIndex()) );
          if (aBulletInfo.bVisible)
              nIndex += aBulletInfo.aText.getLength();
-        if (nIndex != 0 && nIndex >= getCharacterCount())
-            nIndex = getCharacterCount()-1;
-        //
-        if (nIndex != 0)
-            CheckIndex(nIndex); // may throw IndexOutOfBoundsException
+        CheckIndex(nIndex); // may throw IndexOutOfBoundsException

          bool bSupplementalMode = false;
          uno::Sequence< OUString > aPropertyNames = rRequestedAttributes;
@@ -1534,9 +1530,6 @@ namespace accessibility

          SolarMutexGuard aGuard;

-        if ((rPoint.X <= 0) && (rPoint.Y <= 0))
-            return 0;
-
          sal_Int32 nPara;
          sal_uInt16 nIndex;

_______________________________________________
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Reply via email to