[jsr-314-open] Metadata complete jar files
by Andy Schwartz
Gang -
Section 11.5.1 of the spec defines the following annotation scanning
behavior:
> Requirements for scanning of classes for annotations
> * If the <faces-config> element in the WEB-INF/faces-config.xml file
> contains metadata-complete attribute whose value is “true”, the
> implementation must not perform annotation scanning on any classes
> except for those classes provided by the implementation itself.
> Otherwise, continue as follows.
> * If the runtime discovers a conflict between an entry in the
> Application Configuration Resources and an annotation, the entry in
> the Application Configuration Resources takes precedence.
> * All classes in WEB-INF/classes must be scanned.
> * For every jar in the application's WEB-INF/lib directory, if the jar
> contains a “META-INF/faces-config.xml” file or a file that matches the
> regular expression “.*\.faces-config.xml” (even an empty one), all
> classes in that jar must be scanned.
Since application developers have the ability to disable annotation
scanning at a global level, this means that any component set that wants
to support this mode would need to provide a metadata complete
faces-config.xml file. I don't think this is a hardship for most
component vendors, since presumably component vendors are going to want
to provide design-time metadata (eg. JSR-276 metadata), which, for the
moment, requires a faces-config.xml file anyway.
A question that came up here is whether we can tweak section 11.5.1 to
accommodate metadata complete jar files. That is, can we specify that
any jar that contains a faces-config.xml with a metadata-complete="true"
attribute would not be scanned? This would allow component vendors to
indicate that their jar files are metadata complete, and thus avoid the
cost of annotation scanning for the contents of the jar.
Note that while the annotation scan is typically a one time hit (during
application startup), design-time tools may end up starting/stopping JSF
repeatedly during the development process. Thus, avoiding unnecessary
scanning should provide for a more efficient development experience.
Any thoughts on whether we could/should make this change? Does anyone
know of reasons why we avoided specifying this originally?
Also - if we agree to make this change, is this small enough that we
could get this into the the next MR?
Andy
13 years, 1 month
[jsr-314-open-mirror] [jsr-314-open] Ajax inside a DataTable
by Cagatay Civici
Hi,
I've faced with an issue in our app I'd like to share when trying to update the datatable itself from a command element located inside a column. Case is to select a row, execute logic and update the datatable. Basic code:
<h:dataTable id="cars" var="car" value="#{tableBean.carsSmall}">
<h:column>
<f:facet name="header">
Model
</f:facet>
<h:outputText value="#{car.model}" />
</h:column>
<h:column>
<f:facet name="header">
Action
</f:facet>
<h:commandButton value="Some Action" actionListener="#{tableBean.handleRowAction(car)}">
<f:ajax render="cars" />
</h:commandButton>
</h:column>
</h:dataTable>
As datatable has a rowIndex >= 0 during rendering of commandButton f:ajax defines the component id to render as cars:{rowIndex} where I should expect "cars" only. This is due to UIData.getClientId implementation as UIData
adds rowIndex for unique row ids. This causes an issue with a nested f:ajax case.
Regards,
Cagatay Civici
13 years, 10 months
[jsr-314-open] ADMIN: Final list of issues for JSF 2.0 and actions for each
by Ed Burns
This is the final list of issues for JSF 2.0. No other issues will be
considered for 2.0.
I do not think content of any of these issues, nor their quantity,
warrant another Proposed Final Draft, but I will of course produce an
Editor's Draft for the public on jsr-314-open to review.
I will send another email with the dates for the remainder of JSR-314.
Any further discussion on any topic sent to this list will be rolled
into the next release of the spec.
SECTION: List of EG Members from whom I have obtained explicit buy-in
I have obtained explicit buy in for this list in its present state from
the following EG members.
Pete Muir Status: DONE, but need final official answer from JBoss. Promised by 1800 EDT 20090415.
Andy Schwartz 617 794 7974 Status: DONE Spoke with him 1319 20090415
Kito Mann 917 848 3359 Status: DONE Spoke with him 1600 20090415
Martin Marinschek +43 699 1805 3906 Status: DONE Spoke with him at 1400 EDT 20090415
Ken Paulsen x42083 Status DONE Spoke with him 1403 20090415
Alexandr Smirnov DONE IRC chat with him 16:00 20090415
SECTION: Issue sums
Number of issues remaining: 30
I've broken down the issues into two groups, bigger and smaller. I
assert that we do not need to do a PFD2 because none of the issues are
so big as to warrant such an action.
Bigger issues:
Number of bigger issues: 16
deferred: 5
will fix: 10
pushed to EG member requesting the issue: 1
Smaller issues:
Number of smaller issues: 14
deferred: 1
will fix: 13
SECTION: Bigger issues
PM> ISSUE: 1) Fully stateless views. This is a performance optimization
PM> which is ideal for the use case of pages which are output-only (no
PM> form). This is moderately complex to get right, and a blocker in our
PM> opinion. This also has the support of Apache.
We see two aspects to consider.
a. Marking specific views in an app as stateless. Putting a
transient="true" attribute on <f:view> doesn't work here because the
template page with <f:view> can be shared across pages.
There are several options we could try but none of them are feasible at
this point in time.
b. If the Form Renderer never calls StateManager.writeState(), and the
ViewHandler is aware of this fact, then you have a stateless view.
Therefore, statelessness is an implementation detail. Adam Winer
provided this idea, which we implement in Mojarra.
ACTION: Defer to a later release.
EB> Pete agrees to defer, but Gavin does not.
EB> Andy agrees to defer.
EB> Ken agrees to defer.
EB> Martin agrees to defer.
EB> Kito agrees to defer
EB> Alexandr, agrees to defer, but is unhappy about it.
PM> ISSUE: 3) Better handling of whitespace in facelets. Currently
PM> whitespace in view sources can get "eaten" and you have to resort to
PM> tricks like to add space. I would put this as a critical and
PM> the (spec) complexity as low (simply switch it on/off).
ACTION: Defer to a later release.
EB> Pete says, "70 - yes, defering that is fine"
EB> Andy is ok for deferring this
EB> Ken is ok to defer.
EB> Martin agrees to defer. Not a spec issue.
EB> Kito agrees to defer.
EB> Alexandr, agrees to defer
PM> ISSUE: 2) First class support for page actions. This allows a more
PM> action orientated approach (Struts style) approach to page
PM> authoring, where some arbirtrary method can be run when the page is
PM> accessed. Use cases include security (redirect to another page if
PM> the user isn't logged in for example). JSF2 currently includes
PM> "second class" support - it's possible to do this stuff, but not in
PM> simple fashion. As the design for this has already been done, and
PM> reviewed/approved by the EG, the complexity is low. I would put this
PM> as a blocker. It also has the support of Oracle.
Pete, when we had the discussion regarding View Parameters, we decided
to defer Page Actions to a later release. We will not reverse that
decision.
ACTION: Defer to a later release.
EB> Pete says "I am happy for this to slip, given you can do a
EB> workaround with events. l know Dan is unhappy with the current
EB> state. On this one, Gavin was wavering, and would probably be happy
EB> to slip it".
EB> Andy can accept deferral for this
EB> Ken says, "Events should be able to accommodate this use case until
EB> we have a chance to improve the user-experience in a later release.
EB> I think we should defer this.
EB> Martin says ok to defer. But Ed and Martin agree that we should
EB> have a {Pre,Post}BuildViewEvent that is published appropriately.
EB> Kito agrees to defer.
EB> Alexandr, agrees to defer
PM> ISSUE: 5) make behaviour of new components more consistent -
PM> currently f:validateBean (JSR-303 support) / f:ajax use different
PM> strategies to mark areas of the areas of the page to enable. Align
PM> these. This is a blocker (once it is written in stone we can't go
PM> back and fix it) and complexity low (the EG has agreed on the
PM> correct design, so the language needs tweaking in the spec and
PM> javadoc). This has the support of Oracle.
ACTION: If the EG agrees that the default validator setup currently hard
coded into UIInput.encodeEnd() can be moved into
Application.createComponent(), this can be done.
EB> Pete agrees with the plan on this.
EB> Andy agrees with the plan on this.
EB> Ken agrees with the plan on this.
EB> Martin agrees with the plan on this.
EB> Kito agrees with the plan on this.
EB> Alexandr agrees with the plan, but thinks the tag handler is a
EB> better place to imbue the component with the default validator.
PM> ISSUE: 7) Allow multiple error messages for a component to be
PM> displayed - especially useful for JSR-303 support. This is
PM> complexity low, and priority minor (easy for an addon to fix).
ACTION: Pete, please submit a diff patch modifying the following files:
standard-html-renderkit-base.xml jsf-api/doc/standard-html-renderkit-base.xml
message-message-props.xml jsf-api/doc/message-message-props.xml
to specify the change, and the necessary implementation changes for this
feature.
For to the patch to be accepted at this VERY late date it must meet the
following requirements. Because of the late timing of this feature
request we will be unable to accept the feature if the patch doesn't
meet all of the following requirements.
* All existing automated tests still run successfully after applying the patch.
* The spec wording in the patch requires no grammatical or formatting
changes.
* The patch is sent to the EG list no later than 23:59 EDT Wednesday
20090415.
RL> ISSUE: 2. View ID Derivation Fix
RL> (java.net Issue 1002)
RL> Status: Ed had a fix - push back from Ken Paulsen on the fix;
ACTION: Because this breaks the V3 Admin GUI, it must be fixed for 2.0
=> edburns fix by Thursday 20090316 17:00 EDT.
RL> ISSUE: 3. Client Behavior in Composite Components
RL> (patch from Exadel)
RL> Status: Patch 1 checked in; Patch 2 checked in;
RL> Javadocs / prose updates need to be made.
ACTION: Fix checked in to revision 6952, will review spec implications
RL> ISSUE: 4. Bean Validation Bug Fixes
RL> (java.net issue 1058)
RL> Status: Checked in; Needs discussion; Javadocs, prose updates
ACTION: Will review docs, RELEASE_PENDING. Make sure spec for
CompositeComponent Attributes ELResolver mentions this, 5.6.2.2
JD> ISSUE: Impl 1071: UIData.visitTree()
ACTION: Will Fix. fully specify
JD> ISSUE: Impl 1072: Facelet Meta* Javadocs not cleanroom implementable
ACTION: Will fix.
EG> ISSUE: JavaScript disabled support [Was: Outcome of JSFDays
EG> discussions]
ACTION: This can easily be doable in a later release. Considering the
late date of this request we need to defer it.
EB> Pete is ok with deferring this.
EB> Andy is ok with deferring this.
EB> Ken is ok to defer.
EB> Martin is ok with defer
EB> Kito agrees to defer.
EB> Alexandr agrees to defer.
EG> ISSUE: minor feature enhancement for Facelets/VDL
ACTION: Defer to a later release
EB> Pete's ok with deferring this.
EB> Andy's ok with deferring this.
EB> Ken is ok to defer.
EB> Martin ok to defer.
EB> Kito ok to defer.
EB> Alexandr is ok to defer.
EG> ISSUE: required attributes in composite components
ACTION: Defer to a later release
EB> Pete's ok with deferring this.
EB> Andy's ok with deferring this.
EB> Ken is ok to defer this.
EB> Martin is ok to defer this.
EB> Kito ok to defer this.
EB> Alexandr, agrees to defer
EG> ISSUE: remove target attr from h:outputStylesheet (was:
EG> h:outputStylesheetdocs need to be updated)
ACTION: Will fix this
EG> ISSUE: #{compositeComponent.attrs....} verbosity
ACTION: Change "compositeComponent" to be "cc". Implementations may
provide a context-param that allows renaming the implicit object name to
an arbitrary value for the case where an app happens to have a
managed-bean named cc.
EB> Pete says he won't object to this solution.
EB> Andy is ok with this action
EB> Ken says, "I like 'cc' better to solve the verbosity issue. I
EB> recommend deferring on the renaming via context-param for now -- but
EB> would not object to it making the release if others feel strongly
EB> that it needs to be done now."
EB> Martin is ok with this solution.
EB> Kito is ok with this solution.
EB> Alexandr is ok to defer.
EG> ISSUE: composite insert children
ACTION: 2 days. Take the following steps
EB> Pete is ok with this action.
EB> Andy is ok with this action.
Ken says,
KP> I am very nervous about the reparenting (insert*) commands. If 2
KP> composite:insertFacet/Children tags are used in the same composite
KP> component, it will likely fail. Is that acceptable?
EB> We will specify that if multiple <composite:insertFacet> elements
EB> exist with the same name in the <composite:implementation> section,
EB> the "last one wins".
EB> We will specify that if multiple <composite:insertChildren> elements
EB> exist in the <composite:implementation> section the facelet layer
EB> must throw a FaceletException.
KP> If the component inserted has code which does: getParent(), they'll
KP> be exposed to the internal workings of a composite component (which
KP> they shouldn't know exists).
Yes, you are correct, but I feel this is sufficiently corner case to
ignore.
KP> I would prefer render* in most cases. In cases where this isn't
KP> possible (implementing a facet inside a composite component),
KP> perhaps we can do a proxy component which can be added to the inner
KP> facet and delegate rendering to the page-facet? I think there will
KP> be problems with this implementation. If I am wrong, then I do not
KP> object to this going in -- otherwise, I think it should be deferred.
EB> Martin is ok with this action.
EB> Alexandr is ok with this action.
1. rename composite:insertFacet to composite:renderFacet
2. create composite:insertFacet that reparents the named facet from the
top level component to be a facet child of the parent tag in the
composite:implementation section in which the composite:insertFacet
resides.
3. delete composite:renderUsingPageChildren
4. create composite:insertChildren that reparents the children from the
top level component to the parent of tag in the composite:implementation
section in which the composite:insertChildren resides.
5. For composite:insert{Facet,Children} support nested composite
components.
=> edburns take this to Bill and see what he thinks, if we need to do a
PFD2.
SECTION: Smaller issues
PM> ISSUE: 6) Better support for timezones. Currently only one of two
PM> timezones can be used - default or system. The user may well be in
PM> some other timezone entirely, so make the timezone used
PM> configurable. This is a critical issue, and the complexity is medium
PM> (but we have a full proposal for this waiting in the wings that
PM> didn't make it due to time constraints)
If this was a truly critical issue, it would have been brought up a long
time ago, not after the PFD has been published. Furthermore, this is a
platform level thing. I direct you to your JSR-316 EG rep for this feature.
ACTION: Defer to later release.
EB> Pete is ok with deferring this.
EB> Andy is ok with deferring this.
EB> Ken says defer.
EB> Martin ok to defer.
EB> Kito ok to defer.
EB> Alexandr is ok to defer
RL> ISSUE: 2. 11.4.7 - Absolute ordering. I think if a document name
RL> is explicitly referenced but not present, an exception should be
RL> thrown.
ACTION: We'll log a message in this case.
EB> Pete's ok with this action.
EB> Andy's ok with this action.
EB> Ken's ok with this action.
EB> Martin is ok with this action.
EB> Kito is ok with this action
EB> Alexandr states, "if related document was not present, it is
critical error, therefore Exception should be thrown." Given the
prevailing opinion, we'll stick with logging.
RL> ISSUE: 3. FacesRenderKit annotation hasn't existed from some time
RL> now.
ACTION: Remove this from the spec.
RL> ISSUE: 4. Javadocs for EditableValueHolder.addValidator() and
RL> ValueHolder.setConverter() mention ResourceDependency annotation
RL> processing. This documentation should be moved to
RL> Application.createConverter() and Application.createValidator().
ACTION: Will fix this in the spec.
RL> ISSUE: 5. 3.1.11 mentions that the AttributesMap, if the current
RL> component is composite, must eval any ValueExpressions stored in the
RL> Map. This is incorrect. This logic is handled by the
RL> CompositeComponentAttributesELResolver
ACTION: Will fix in the spec.
RL> ISSUE: 6. Preface references META-INF/managed-beans.xml. Support
RL> for this file is no longer required.
ACTION: Will fix.
RL> ISSUE: 7. 5.6.2.2 Doesn't mention any support for the special
RL> parent keyword in #{compositeComponent} expressions.
ACTION: Will fix.
RL> ISSUE: 8. 6.1.1.15 lists Flash as a property of FacesContext. It's
RL> now a property of ExternalContext.
ACTION: Will fix.
RL> ISSUE: 9. 7.1.1 and 7.1.2 Typo in assertion marker
RL> 'defualtActionListener' and 'defualtRenderKit' There may be others.
ACTION: Will fix
RL> ISSUE: 10. 7.4.2 will need to be discussed due to the
RL> implementation of issue 1066.
ACTION: Will fix
RL> ISSUE: 11. 7.4.3 should include a navigation case example with
RL> redirect parameters.
ACTION: Will fix
RL> ISSUE: 7.5.1 and others. This may already be done, but
RL> PageDeclarationLanguage should be ViewDeclarationLanguage
ACTION: Will fix. Make sure retargetAttachedObjects and retargetMethodExpressions are on VDL, not ViewHandler.
RL> ISSUE: 1. Spec ResponseWriter methods: startCDATA endCDATA (java.net
RL> issue 1055) Status: impl done; method javadocs / prose need to be
RL> done;
ACTION: Will fix
EG> ISSUE: View and custom scoped eager managed beans?
ACTION: Will Fix. Add eager for session scoped managed beans
RL> Update 3rd bullet in 11.5.1 so that pattern for custom faces config file
names is something that ends with .faces-config.xml. This brings it in
alignment with 11.4.
--
13 years, 11 months
Re: [jsr-314-open-mirror] [jsr-314-open] [2.0RevA-C071] none scoped beansand@PostConstruct
by Simon Lessard
Re: [jsr-314-open] [2.0RevA-C071] none scoped beansand@PostConstruct
Hi all,
I thought a bit more about this, and while I still think @PostConstruct
should be called for none scoped bean, I think we should document
@PreDestroy better as we won't have many way to call the annotated method
for none scoped bean. Of course, the managed bean engine could be made very
intelligent and know in what context the none scoped bean is getting used.
If injected into another bean, it could add the none scoped bean in an
hidden map in the same scope to detect destruction, but if the none scoped
bean is used directly, which is rare but possible, then I cannot think of
any way to enforce @PreDestroy.
Regards,
~ Simon
13 years, 11 months
Re: [jsr-314-open-mirror] [jsr-314-open] f:selectItems value="#{myMap}" underspecified
by Ed Burns
>>>>> On Tue, 25 May 2010 14:59:18 +0200, Martin Marinschek <mmarinschek(a)apache.org> said:
MM> Hi Leo,
>>> I never understood why the labels should be the key here - it should
>>> really be the other way round, and the value should be the key. I
>>> thought that was sorted for 1.2 - it is not?
>>>
>>>
>> No, this was not solved because the documentation of f:selectItems does
>> not have the properties mentioned (var, itemValue, itemLabel....). In fact
>> it was f:selectItem the component that has them in jsf 1.2
MM> Well, so this is really two issues - first, it is not documented,
MM> second, I still think if we build the items from a map, it should go
MM> the other way round.
Doch, f:selectItems sure does have the documentation.
https://javaserverfaces.dev.java.net/nonav/docs/2.0/pdldocs/facelets/f/se...
MM> Can you open issues?
He did open
<https://javaserverfaces-spec-public.dev.java.net/issues/show_bug.cgi?id=772>,
to which I just added this comment:
Leonardo, we can either specify that the "var" attribute is
unsupported for f:selectItems where value is a Map, or we can push
this to 2.1. I am constrained to these choice because to make
"f:selectItems var points to map" work needs a behavior change.
By default I've moved it to 2.1. Please comment if you want the "not
supported" option and have it in 2.0 Rev a.
Ed
--
| edburns(a)oracle.com |
| homepage: | http://ridingthecrest.com/
13 years, 11 months
[jsr-314-open-mirror] [jsr-314-open] Fwd: [Fwd: Re: Fwd: PostAddToViewEvent publishing conditions]
by Martin Marinschek
My comments on Andy's mail, now officially to the list.
=========================================
Hi Andy,
ok, I will answer here, and will also try to forward when I see your
mail on the list ;)
> Actually meant to say that I believe that DecorateHandler got this one
> right. (Calling pushClient() is the way to go.)
ok
> Unless someone understands why the implementation is inconsistent across
> these two cases, I think that we should align the implementations. Note
> that given the bug that Max logged (against Mojarra):
>
> https://javaserverfaces.dev.java.net/issues/show_bug.cgi?id=1684
>
> I believe that CompositionHandler has got this one right (ie. both
> DecorateHandler and CompositionHandler should be calling pushClient()).
ok
> If the template itself happens to depend on yet another template, I view
> that as an implementation detail of the template. So, if we've got this
> situation:
>
> - view.xhtml, which includes:
> - outerTemplate.xhtml, which includes:
> - innerTemplate.xhtml
>
> In my ideal world (which, at the moment, does not match reality), view.xhtml
> should only be aware of the slots that are exposed by outerTemplate.xhtml.
> That is, view.xhtml should not depend on the fact that outerTemplate.xhtml
> happens to include innerTemplate.xhtml, which may have its own slots.
>
> The reality is that when nesting templates, the outermost template exposes
> *all* of the slots defined by any nested templates (no matter how many
> levels deep) to the consumer. This seems like a violation of encapsulation
> to me. We are exposing implementation details up through the template
> hierarchy that, I think, should not necessarily be exposed.
>
> Of course, I understand that it may be useful to be able to pass content
> from view.xhtml all the way through to a slot defined by
> innerTemplate.xhtml. If outerTemplate.xhtml wants to allow certain content
> to be passed through from view.xhtml to innerTemplate.xthml, that's fine.
> However, this should be done by introducing a ui:insert inside of
> outerTemplate.xhtml - ie. by explicitly exposing a new slot that consumers
> of outerTemplate.xhtml can rely on. That way, if the author of
> outerTemplate.xhtml reconsiders the implementation and decides not to
> delegate to innerTemplate.xhtml after all, users of outerTemplate.xhtml are
> not hosed.
I totally understand your notion - and I agree. However, this is not
how Facelets works now. For me, this is a little bit like a method
call where suddenly inner method calls inherit the parameters from the
outside. FWIW, also the facelets parameters for composite components
(old style) were inherited this way, at least for a long time (I am
not sure if this has been fixed at some point of time). I find this
ugly.
> Oh well, this looks like another cases where I am many years too late. That
> seems to be my fate these days.
>
> Even if we cannot change this behavior for ui:insert now, I do think it is
> important that we avoid this behavior for composite components. Since we
> are tinkering with the composite:inertFacet/composite:insertChildren
> implementations now to be more Facelets/TemplateClient-like, let's just be
> careful to avoid introducing the Facelets nested insert behavior into
> composite components.
agreed
>>> Right, so I see that InsertHandler is implemented this way, but... why?
>>> Seems silly to have ui:insert serve two entirely different purposes -
>>> ie.
>>> to both identify an insertion point as well as to define content for
>>> included templates. Why should ui:insert act both as a ui:insert *and*
>>> as
>>> a ui:define? If a template author wants to pass content through from an
>>> outer page to an inner/nested template, seems like the right way to
>>> accomplish this would be to wrap the ui:insert inside of a ui:define.
>>>
>>
>> Well, if you define a template, you use ui:insert to mark the spots
>> where content can be - and then you include default content right
>> away. Is it this which makes ui:insert a thing of both worlds?
>>
>
>
> Let's take a step back and look at the role of the TemplateClient contract.
> TemplateClient allows consumers of templates (eg. ui:composition) to
> provide access to content that should be inserted into the template - ie.
> content defined by ui:define, inserted into ui:insert. So, for example,
> both CompositionHandler and DecorateHandler are TemplateClients.
>
> ui:insert is clearly a user of the TemplateClient interface. InsertHandler
> (via a call to FaceletContext.includeDefinition()) pulls in content provided
> by outer TemplateClients (eg. an ancestor ui:composition).
>
> This is all well and good.
>
> Where things get strange is that InsertHandler also implements
> TemplateClient. This means that, in addition to being a consumer of
> TemplateClient content, ui:insert also exposes its own content for
> consumption by other consumers of TemplateClients.
>
> So, if you've got:
>
> <ui:insert name="foo">
> body content
> </ui:insert>
>
> The "body content" serves two roles:
>
> 1. It provides default content in the event that no matching ui:define for
> "foo" can be found.
> 2. It is exposed as content for consumption while performing the definition
> inclusion (during the call to FaceletContext.includeDefinition()).
>
> Case #1 is expected/documented.
that is the case I meant - wasn´t sure if the existing implementation
tried to reach there.
> Case #2 is where ui:insert seems to acting as a ui:define. This is
> non-obvious, and, as far as I can tell, not documented.
brrr, ugly. Has to go.
best regards,
Martin
--
http://www.irian.at
Your JSF powerhouse -
JSF Consulting, Development and
Courses in English and German
Professional Support for Apache MyFaces
--
http://www.irian.at
Your JSF powerhouse -
JSF Consulting, Development and
Courses in English and German
Professional Support for Apache MyFaces
13 years, 11 months
[jsr-314-open-mirror] [jsr-314-open] Fix UIData state saving model (issue 153)
by Leonardo Uribe
In short, this topic is an attempt to add full state to components inside
UIData. I'll do a brief resume, so people can understand this one easily.
UIData uses the same component instances to render multiple rows. Suppose
this example:
<h:dataTable id="cities" var="city" value="#{country.cities}">
<h:column>
<h:outputText value="#{city}" />
</h:column>
</h:dataTable>
In the component tree it is created this hierarchy:
HtmlDatatable
UIColumn
HtmlOutputText
If we have 10 cities, the same component is used over and over to render all
10 cities. The reason to do that in this way is keep state as small as
possible.
Now let's suppose something like this:
<h:dataTable id="cities" var="city" value="#{country.cities}">
<h:column>
<h:inputText value="#{city}" />
</h:column>
</h:dataTable>
It was changed the output component for an input one. If this table is in a
form and the values are submitted, the same component is used to apply
request values, validate and apply them to the model (update process). To
make this possible, UIData class has some code to preserve this values
between phases (using EditableValueHolder interface), so when each phase is
processed, all rows are traversed and you get the expected behavior.
Now suppose something more complex: We have a code that use
invokeOnComponent to change the style of my inputText. In theory, only one
row should change. But in fact, all rows are rendered with the same color.
Why? because we use the same component to render all rows, and we don't
preserve the children component state between rows.
There is a lot of issues, questions, and side effects related to this issue,
but just to put some of them here:
TOMAHAWK-1062 InputTextArea doesn't work properly inside facet DetailStamp
TOMAHAWK-96 Data table Scroller not working the dataTable which was actually
contained in other DataTable
https://javaserverfaces-spec-public.dev.java.net/issues/show_bug.cgi?id=1...
with UIData state saving
Also, it is well know that one reason why people uses c:forEach in facelets,
is because this one create "full" components per each row. It is very easy
to find articles on internet.
Now, with jsf 2.0 we have partial state saving, so we have a chance to fix
this one once and for all. I tried fix this one per months (maybe years!),
but talking with Martin Marinschek on JSFDays, some ideas came out and
finally it was found a possibility to fix this one using the existing api
and with little changes on the spec.
The proposal is this:
1. Do not call UIComponent.markInitialState() on
ComponentTagHandlerDelegate, as ComponentHandler javadoc says, instead call
it after PostAddToViewEvent are published on vdl.buildView(). We need to
call it from root to nodes, so the parent component is marked first. I know
the place where this call comes is from trinidad tag handler, but this call
needs to be fixed in a more predictable way.
2. Use an attribute on facesContext to identify when the VDL is marking the
initial state (in myfaces there is already an attribute called
"org.apache.myfaces.MARK_INITIAL_STATE"). This is necessary to indicate
UIData instances that it is time to save the full state of all component
children,
3. Allow UIData to hold a map where the key are client ids and the value are
the deltas of all components per row. This map should be saved and restored.
4. Change UIData.setRowIndex() to restore and save the component state.
I'll attach a patch on myfaces issue tracker with the algorithm proposed
(because it is based on myfaces codebase). It was tested and it works. But
note it is necessary to fix the javadoc for UIData.markInitialState(),
ComponentHandler and maybe vdl.buildView(), so the intention is propose this
change for jsf 2.0 rev A. Note this works only with PSS enabled because
without it we don't have a place to notify UIData instances that it is
necessary to get the full state. Also, note this patch preserve backward
compatibility, because the old way to store/save is applied after the full
state is restored.
Really, I have the strong temptation to apply some similar code on myfaces
UIRepeat component (because this class is private), but I prefer first ask
to EG to know what you guys think about it.
Suggestions are welcome,
regards,
Leonardo Uribe
13 years, 11 months
Re: [jsr-314-open-mirror] [jsr-314-open] Fwd: PostAddToViewEvent publishing conditions
by Andy Schwartz
Gang -
Max did some investigation into the problems that Leonardo found when
attempting to use the TemplateClient mechanism to implement composite
component children/facet insertion. It looks like there are some more
fundamental problems here. Max captured his thoughts in the following
Mojarra issue:
https://javaserverfaces.dev.java.net/issues/show_bug.cgi?id=1684
Please review - interesting stuff.
Note - we still think that using a TemplateClient-like approach should
be possible for composite component insertion, so hopefully the
implementations can continue to pursue this option.
Andy
> ---------- Forwarded message ----------
> From: Max Starets <max.starets(a)oracle.com>
> Date: Fri, May 21, 2010 at 3:58 PM
> Subject: Re: [jsr-314-open] PostAddToViewEvent publishing conditions
> To: jsr-314-open(a)jcp.org
> Cc: dev(a)myfaces.apache.org
>
>
> Hey Leonardo,
>
> You are right - the same issue reproduces with two nested
> ui:composition templates.
> I logged the following Mojarra issue for the problem:
> https://javaserverfaces.dev.java.net/issues/show_bug.cgi?id=1684
>
> I do not think we have to abandon the idea of using TemplateClient
> mechanism for for the composite component implementation
> because of this issue. You are already using a separate TemplateClient
> stack for composites in your patch, so we do not
> need to wait for the fix before implementing the correct behavior for
> composites. All we need is a stack with only one TemplateClient
> being considered 'current'. includeDefinition() should be checking the
> current TemplateClient only, and we should pop the current
> TemplateClient before calling TemplateClient.apply(), then restore
> (push) it when the apply() is done.
>
> Max
>
>
> Leonardo Uribe wrote:
>
>> Hi
>>
>> I tried the proposal of use some variation of TemplateClient API (I attach it as reference on MYFACES-2638-6.patch ). The solution works for simple cases, but it does not when nested composite components are used.
>>
>> Look this example (I removed the non relevant code):
>>
>> testCompositeInsertChildren.xhtml (the page when it is used)
>>
>> <testComposite:compositeInsertChildren>
>> <h:outputText value="GAMMA " />
>> </testComposite:compositeInsertChildren>
>>
>> testComposite:compositeInsertChildren
>>
>> <composite:implementation>
>> <testComposite:compositeInsertChildrenInner>
>> <h:outputText value="BETA " />
>> <composite:insertChildren />
>> </testComposite:compositeInsertChildrenInner>
>> </composite:implementation>
>>
>> testComposite:compositeInsertChildrenInner
>>
>> <composite:implementation>
>> <h:outputText value="ALFA " />
>> <composite:insertChildren/>
>> <h:outputText value="OMEGA " />
>> </composite:implementation>
>>
>> The example should render this:
>>
>> ALFA BETA GAMMA OMEGA
>>
>> But it is rendered this:
>>
>> ALFA GAMMA OMEGA
>>
>> The current algorithm do it correctly because the tree is completely built before relocate, so the listener relocates BETA too. Facelets executed the inner FaceletHandler instance first and then the outer ones. On composite components, it executes first the composite component facelet, then the FaceletHandler children.
>>
>> Suggestions are welcome.
>>
>> regards,
>>
>> Leonardo Uribe
>>
13 years, 11 months
Re: [jsr-314-open-mirror] [jsr-314-open] f:selectItems value="#{myMap}" underspecified
by Martin Marinschek
Hi Leonardo,
> Map - The keys of this object (once converted to Strings) are assumed to be
> labels, and the values of this object (once converted to Strings) are
> assumed to
> be values, of SelectItem instances that will be constructed dynamically and
> added to the set of available options for the parent component, in the order
> provided by an iterator over the keys.
I never understood why the labels should be the key here - it should
really be the other way round, and the value should be the key. I
thought that was sorted for 1.2 - it is not?
> This behavior comes from jsf 1.1, but in jsf 2.0 it was added some new
> attributes (from f:selectItems tlddoc):
>
> Version 2 of the specification introduces several new attributes, described
> below. These are: var, itemValue, itemLabel, itemDescription, itemDisabled,
> and
> itemLabelEscaped.
>
> Now, what happen if some user do something like this:
>
> <f:selectItems value="#{myMap}" var="item"
> itemLabel="#{item.value.someLabelProperty}" itemValue="#{item.value}">
>
> It just does not work, because there is no clear definition about what
> should
> f:selectItems do in this case.
>
> The proposal for solve this one is if var property is set and value
> implements
> Map interface, use the entry object as var, so the user can choose between
> the
> key and some attribute on the item value.
I totally agree - +1 from me.
best regards,
Martin
13 years, 11 months
