[
https://issues.apache.org/jira/browse/WICKET-2130?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12677005#action_12677005
]
Martijn Dashorst commented on WICKET-2130:
------------------------------------------
Patch below fixes the unit test, but causes multiple detaches for normal pages.
I'd suggest setting the ATTACHED flag in the Page constructor. Typically a page
is constructed in an Attached state, but might not be rendered until a next
request (and only then pass through onAttach()).
Index: src/main/java/org/apache/wicket/Session.java
===================================================================
--- src/main/java/org/apache/wicket/Session.java (revision 746413)
+++ src/main/java/org/apache/wicket/Session.java (working copy)
@@ -57,11 +57,10 @@
* holds the Page. So {...@link Component#getSession()} traverses the
component hierarchy to the root
* Page and then calls {...@link Page#getSession()}.
*
- * <li><b>Access via Thread Local </b>- In the odd case where neither a
RequestCycle nor a
- * Component is available, the currently active Session for the calling thread
can be retrieved by
- * calling the static method Session.get(). This last form should only be used
if the first two
- * forms cannot be used since thread local access can involve a potentially
more expensive hash map
- * lookup.
+ * <li><b>Access via Thread Local </b>- In the odd case where neither a
RequestCycle nor a Component
+ * is available, the currently active Session for the calling thread can be
retrieved by calling the
+ * static method Session.get(). This last form should only be used if the
first two forms cannot be
+ * used since thread local access can involve a potentially more expensive
hash map lookup.
*
* <li><b>Locale </b>- A session has a Locale property to support
localization. The Locale for a
* session can be set by calling {...@link Session#setLocale(Locale)}. The
Locale for a Session
@@ -103,9 +102,9 @@
* <li><b>Removal </b>- Pages can be removed from the Session forcibly by
calling remove(Page) or
* removeAll(), although such an action should rarely be necessary.
*
- * <li><b>Flash Messages</b>- Flash messages are messages that are stored in
session and are
- * removed after they are displayed to the user. Session acts as a store for
these messages because
- * they can last across requests.
+ * <li><b>Flash Messages</b>- Flash messages are messages that are stored in
session and are removed
+ * after they are displayed to the user. Session acts as a store for these
messages because they can
+ * last across requests.
*
* @author Jonathan Locke
* @author Eelco Hillenius
@@ -388,8 +387,8 @@
* <p>
* By calling this method, the session will be bound (made
not-temporary) if it was not bound
* yet. It is useful for cases where you want to be absolutely sure
this session object will be
- * available in next requests. If the session was already bound
({...@link ISessionStore#lookup(Request) returns a session}),
- * this call will be a noop.
+ * available in next requests. If the session was already bound (
+ * {...@link ISessionStore#lookup(Request) returns a session}), this
call will be a noop.
* </p>
*/
public final void bind()
@@ -1367,6 +1366,9 @@
for (int i = 0; i < touchedPages.size(); i++)
{
Page page = (Page)touchedPages.get(i);
+
+ // ensure pages are detached when shared
through the pagemaps
+ page.detach();
page.getPageMap().put(page);
dirty = true;
}
> Pages stored in Session.touchedPages aren't detached when part of ModalWindow
> -----------------------------------------------------------------------------
>
> Key: WICKET-2130
> URL: https://issues.apache.org/jira/browse/WICKET-2130
> Project: Wicket
> Issue Type: Bug
> Components: wicket
> Affects Versions: 1.3.5, 1.4-RC2
> Reporter: Martijn Dashorst
> Fix For: 1.3.6, 1.4-RC3
>
> Attachments: wicket-2130.tgz
>
>
> Creating a ModalWindow with a Page causes the newly constructed page not to
> be detached.
> The page is stored in Session.touchedPages, but that list isn't properly
> processed at the end of the request for detaching.
> I'll try to create a testcase for this.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
