221387 - XPathExpression.createExpression doesn't throw NAMESPACE_ERR

Status: RESOLVED FIXED

(Core :: XSLT, defect, P3)

Description

[Peter Van der Beken [:peterv]]

Making this happen will involve moving the voidable string class out of
XPConnect, since we need it to detect null namespaces from lookupNamespaceURI.

Comment 1

[Peter Van der Beken [:peterv]]

Attachment: Move voidable string from xpconnect to xpcom

Comment 2

[Peter Van der Beken [:peterv]]

Attachment: Use voidable string to detect unknown prefixes

Comment 3

[Axel Hecht]

Comment on attachment 132741 [details] [diff] [review]
Use voidable string to detect unknown prefixes

>Index: extensions/transformiix/source/xpath/nsXPathEvaluator.cpp
<...>
>-    *aResult = new nsXPathExpression(expression, mRecycler);
>+    *aResult = new nsXPathExpression(expression.forget(), mRecycler);
>     if (!*aResult) {
>-        delete expression;
>         return NS_ERROR_OUT_OF_MEMORY;
>     }
> 

This leaks the expression on out-of-mem. .forget() it on success, not before.
Or use a nsAutoPtr&.
I don't need a new patch for that, though, so r=[e-mail]

Comment 4

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

Comment on attachment 132740 [details] [diff] [review]
Move voidable string from xpconnect to xpcom

If this is going to be used in more places, I'd rather see the virtual
functions defined in the .cpp file rather than the .h file.  sr=dbaron with or
without that, though.

Comment 5

[Peter Van der Beken [:peterv]]

Attachment: Move voidable string from xpconnect to xpcom

Comment 6

[Peter Van der Beken [:peterv]]

Attachment: Use voidable string to detect unknown prefixes

Comment 7

[Boris Zbarsky [:bzbarsky]]

Comment on attachment 132919 [details] [diff] [review]
Move voidable string from xpconnect to xpcom

r=bzbarsky

Comment 8

[Boris Zbarsky [:bzbarsky]]

Comment on attachment 132920 [details] [diff] [review]
Use voidable string to detect unknown prefixes

sr=bzbarsky

Comment 9

[Peter Van der Beken [:peterv]]

Checked in.

Comment 10

[Johnny Stenback (:jst)]

It'd be really cool if our new widely usable nsVoidableString wasn't obsolete
starting from its birth, but oh well :-)

Comment 11

[Peter Van der Beken [:peterv]]

I didn't realize the stuff in string/obsolete was obsolete :-P.