301998 - Inner repeats should initialize to startindex on update

Status: RESOLVED FIXED

(Core Graveyard :: XForms, defect)

Description

[Allan Beaufour]

If setindex, insert, or delete modify the index of a repeat parent the inner
repeats should per the errata re-initialize to their startindex attribute (and
not 1 as it is now).

Comment 1

[Allan Beaufour]

Attachment: Patch

Created a getStartingIndex() on nsXFormsRepeatElement, used to set the index to
"startindex" or 1 if it's not present.

Comment 2

[Allan Beaufour]

btw, attachment 197407 [details] [diff] [review] depends on bug 302918

Comment 3

[Allan Beaufour]

Attachment: Testcase

Comment 4

[aaronr]

Comment on attachment 197407 [details] [diff] [review]
Patch


>--- xforms/nsXFormsRepeatElement.cpp	2005-09-26 14:46:07.000000000 +0200
>+++ xforms.repeatfix/nsXFormsRepeatElement.cpp	2005-09-26 14:51:43.000000000 +0200

> NS_IMETHODIMP
>+nsXFormsRepeatElement::GetStartingIndex(PRUint32 *aRes)
>+{
>+  nsresult rv = GetIntAttr(NS_LITERAL_STRING("startindex"),
>+                           (PRInt32*) aRes,
>+                           nsISchemaBuiltinType::BUILTIN_TYPE_POSITIVEINTEGER);
>+  if (NS_FAILED(rv)) {
>+    *aRes = 1;
>+  }
>+  if (*aRes < 1) {
>+    *aRes = 1;
>+  } else if (*aRes > mMaxIndex) {
>+    *aRes = mMaxIndex;
>+  }
>+  return NS_OK;
>+}
>+
>+

You should probably have a NS_ENSURE_ARG(aRes) at the beginning of this
function.  Might want to add one to ::GetIsParent while you are at it.


With that, r=me

Comment 5

[Allan Beaufour]

Checked in to trunk with smaug's comments

Comment 6

[Allan Beaufour]

(In reply to comment #5)
> Checked in to trunk with smaug's comments

Aaron's comments from comment 4, that is...

Comment 7

[aaronr]

checked into branch 20051004