1393088 - Port bug 1390209 to C-C [Remove unused nsIDOMHTML*Element interfaces]

Status: RESOLVED FIXED

(Thunderbird :: General, defect)

Description

[Richard Marti (:Paenglab)]

I get this error:

z:/Mozilla/comm-central/mailnews/compose/src/nsMsgSend.cpp(30): fatal error C1083: Cannot open include file: 'nsIDOMHTMLBodyElement.h': No such file or directory

Bug 1390209 removed this. backout of this bug makes TB buildable.

Comment 1

[Jorg K (CEST = GMT+2)]

Attachment: 1393088-nsIDOMHTMLBodyElement.patch

This compiles, but I'm not sure that it's right. Very few C++ examples of HTMLBodyElement in the tree. Sadly :bz is on holidays, I'll see how I get advice on this. Maybe Ehsan.

Comment 2

[:aceman]

Comment on attachment 8900359 [details] [diff] [review]
1393088-nsIDOMHTMLBodyElement.patch

Review of attachment 8900359 [details] [diff] [review]:
-----------------------------------------------------------------

::: mailnews/compose/src/nsMsgSend.cpp
@@ +1289,5 @@
>    // Now, we know the types of objects this node can be, so we will do
>    // our query interface here and see what we come up with
> +  nsCOMPtr<Element> bodyElement = (do_QueryInterface(node));
> +  RefPtr<mozilla::dom::HTMLBodyElement> body =
> +    static_cast<mozilla::dom::HTMLBodyElement*>(mozilla::dom::HTMLBodyElement::FromContentOrNull(bodyElement));

Why the cast? The RefPtr<mozilla::dom::HTMLImageElement> below doesn't have it.

Comment 3

[Jorg K (CEST = GMT+2)]

Without the cast I get:

 0:12.66 c:/mozilla-source/comm-central/mailnews/compose/src/nsMsgSend.cpp(1293): error C2440: 'initializing': cannot convert from 'nsGenericHTMLElement *' to 'RefPtr<mozilla::dom::HTMLBodyElement>'
 0:12.66 c:/mozilla-source/comm-central/mailnews/compose/src/nsMsgSend.cpp(1293): note: No constructor could take the source type, or constructor overload resolution was ambiguous

Comment 4

[Jorg K (CEST = GMT+2)]

I actually works, I sent and e-mail with background and it was downloaded and attached.

Comment 5

[Jorg K (CEST = GMT+2)]

Attachment: 1393088-nsIDOMHTMLBodyElement.patch (v2).

After advice from Olli Pettay (:smaug) on IRC I came up with this.

Comment 6

[Olli Pettay [:smaug][[e-mail]]]

Comment on attachment 8900385 [details] [diff] [review]
1393088-nsIDOMHTMLBodyElement.patch (v2).

>+  nsCOMPtr<nsINode> bodyNode = do_QueryInterface(node);
I wouldn't call this bodyNode, since it may not point to a body element. Perhaps just n or some such temporary sounding variable name?


>       // Now, we know the types of objects this node can be, so we will do
>       // our query interface here and see what we come up with
>-      nsCOMPtr<nsIDOMHTMLBodyElement>     body = (do_QueryInterface(domSaveArray[j].node));
>+      nsCOMPtr<nsINode> bodyNode = (do_QueryInterface(domSaveArray[j].node));
same here. (and I wonder why () around QI)

>     // Now, we know the types of objects this node can be, so we will do
>     // our query interface here and see what we come up with
>-    nsCOMPtr<nsIDOMHTMLBodyElement>     body = (do_QueryInterface(domSaveArray[i].node));
>+    nsCOMPtr<nsINode> bodyNode = (do_QueryInterface(domSaveArray[i].node));
and here

Comment 7

[Jorg K (CEST = GMT+2)]

Attachment: 1393088-nsIDOMHTMLBodyElement.patch (v2b).

Thank you, Olli!

Comment 8

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/comm-central/rev/1777ffb9363f
Port bug 1390209 to mailnews: Remove use of nsIDOMHTMLBodyElement. r=smaug

Comment 9

[Richard Marti (:Paenglab)]

Hmm, all red. And my local fails too.

Comment 10

[Jorg K (CEST = GMT+2)]

Hmm, compiled locally without having pulled M-C. The interface I used should have been in place already but I can see there are some string type issues. Just comment out the four lines and don't send anything. I'll fix it (now in a Council meeting).

Comment 11

[Jorg K (CEST = GMT+2)]

Attachment: 1393088-strings.patch

Comment 12

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/comm-central/rev/e7173819d591
Port bug 1390209 to mailnews, follow-up: Correct parameters of [Get|Set]Background(). rs=bustage-fix

Comment 13

[Boris Zbarsky [:bzbarsky]]

The string bits in these patches are very very wrong.  See upcoming comments in bug 1403516.

Comment 14

[Boris Zbarsky [:bzbarsky]]

Like "might well be a security bug" wrong.