Fix incorrect comment about XMLHttpRequest::handleDidCancel()

Source/core/xml/XMLHttpRequest.h

Index: Source/core/xml/XMLHttpRequest.h
diff --git a/Source/core/xml/XMLHttpRequest.h b/Source/core/xml/XMLHttpRequest.h
index 5b6abc437918101b27e896c7360ac00c805ce47c..ce6455905aecc1d1b8a294f06bbd6dab7d972a08 100644
--- a/Source/core/xml/XMLHttpRequest.h
+++ b/Source/core/xml/XMLHttpRequest.h
@@ -211,7 +211,14 @@ private:
     void handleDidFailGeneric();
     // Handles didFail() call not caused by cancellation or timeout.
     void handleNetworkError();
-    // Handles didFail() call triggered by m_loader->cancel().
+    // Handles didFail() call for cancellation.

[sof, 2014/08/12 12:24:00]

"Handles didFail() call for error cancellations" ?

[sof, 2014/08/12 12:29:51]

Hmm, scratch that - i think it is fine to leave out that detail and keep it as
it is.

[tyoshino (SeeGerritForStatus), 2014/08/12 12:34:46]

not
errors of/meaning cancellation
or
cancellation type error?

[sof, 2014/08/12 12:42:56]

The former covers it more closely, I think, but how cancellation is signalled to
didFail() isn't that important to know here.

Fine as you had it; sorry about the noise.

[tyoshino (SeeGerritForStatus), 2014/08/12 13:48:36]

np. thanks for review. Using "cancellations".

+    //
+    // This is not for the abort() method of the XHR. abort() calls

[sof, 2014/08/12 12:24:00]

Why mention the implementation mechanics of abort() here; isn't that covered by
the comment next to the m_error check in didFail() already?

[tyoshino (SeeGerritForStatus), 2014/08/12 12:34:46]

Oh, I see. Removed!

+    // internalAbort() which sets m_error which prevents handleDidCancel() from
+    // being called. This method handles cancellation coming from other
+    // components. For example, the ResourceLoader handling the load notifies
+    // m_loader of cancellation when the frame containing the XHR navigates
+    // away.
     void handleDidCancel();
     // Handles didFail() call for timeout.
     void handleDidTimeout();