Mailing-List: contact dev-help@myfaces.apache.org; run by ezmlm
Precedence: bulk
Reply-To: "MyFaces Development" <dev@myfaces.apache.org>
Received-SPF: pass (nike.apache.org: domain of mkienenb@gmail.com designates
 209.85.212.175 as permitted sender)
MIME-Version: 1.0
In-Reply-To: 
 <CABujnVff=om-30OBy7L9aEG1jMLGUNZ76UpsrsirQyXtw9HAiw@mail.gmail.com>
References: 
 <CAM1yOjav7JcHvOL-JtY=eXN-jym2LdGYHMj_X_dyCVK1qpLJ1w@mail.gmail.com>
 <CAM1yOjbfE8+jHNQzxmbAE-rTE7Q_qp=CSSzDo04mW=0zakMe9Q@mail.gmail.com>
 <CABujnVff=om-30OBy7L9aEG1jMLGUNZ76UpsrsirQyXtw9HAiw@mail.gmail.com>
From: Mike Kienenberger <mkienenb@gmail.com>
Date: Mon, 25 Feb 2013 11:45:56 -0500
Message-ID: 
 <CAM1yOjabcrv=HgnUcLCuRzb+ca6cwXK9VcRwGNckKwq0qb2SFw@mail.gmail.com>
Subject: Re: duplicate h:body onload (and probably onunload) attributes
 rendered when isCommonPropertiesOptimizationEnabled is true
To: MyFaces Development <dev@myfaces.apache.org>
Content-Type: text/plain; charset=ISO-8859-1

Is trunk still 2.1?
I'm guessing I want to fix this on trunk, 2.0.x and 2.2.x


On Mon, Feb 25, 2013 at 11:39 AM, Leonardo Uribe <lu4242@gmail.com> wrote:
> Hi
>
> Yes, the solution seems correct.
>
> regards,
>
> Leonardo Uribe
>
> 2013/2/25 Mike Kienenberger <mkienenb@gmail.com>:
>> Ok.  Looking at the constants, the solution is probably to replace
>> HTML.BODY_ATTRIBUTES with BODY_ATTRIBUTES_WITHOUT_EVENTS.  I should
>> have looked at the other available constants first.
>>
>>
>> On Mon, Feb 25, 2013 at 11:30 AM, Mike Kienenberger <mkienenb@gmail.com> wrote:
>>> I've noticed this for some time in the MyFaces 2.1 code, but haven't
>>> taken the time to look into it until recently.
>>>
>>> HtmlBodyRendererBase renders onload and onunload, but then renders it
>>> a second time when isCommonPropertiesOptimizationEnabled() is true.
>>> This is not the same issue as MYFACES-2630 where HTML.ONUNLOAD_ATTR
>>> was misdefined as "onload".
>>>
>>> This appears to be because HTML.BODY_ATTRIBUTES contains onload and
>>> onunload and these are unconditionally rendered a second time.
>>>
>>> I'm not sure what the solution should be.   My guess is that we just
>>> need to stop re-rendering the onload onunload attributes in the
>>> optimized section.   I'm not sure if we can change the contents of
>>> BODY_ATTRIBUTES.   It seems odd that passthrough attributes contains
>>> more attributes than body attributes as replacing HTML.BODY_ATTRIBUTES
>>> with HTML.BODY_PASSTHROUGH_ATTRIBUTES_WITHOUT_EVENTS was my first
>>> though.
>>>
>>> HTML.BODY_ATTRIBUTES contains [onload, onunload, alink, vlink, link,
>>> text, background, bgcolor]
>>> HTML.BODY_PASSTHROUGH_ATTRIBUTES_WITHOUT_EVENTS contains [dir, lang,
>>> title, style, styleClass, alink, vlink, link, text, background,
>>> bgcolor]
>>>
>>>             HtmlRendererUtils.renderBehaviorizedAttribute(facesContext,
>>> writer, HTML.ONLOAD_ATTR, component,
>>>                     ClientBehaviorEvents.LOAD, behaviors, HTML.ONLOAD_ATTR);
>>>             HtmlRendererUtils.renderBehaviorizedAttribute(facesContext,
>>> writer, HTML.ONUNLOAD_ATTR, component,
>>>                     ClientBehaviorEvents.UNLOAD, behaviors, HTML.ONUNLOAD_ATTR);
>>>
>>>             if (isCommonPropertiesOptimizationEnabled(facesContext))
>>>             {
>>>                 HtmlRendererUtils.renderHTMLAttributes(writer, component,
>>>                         HTML.BODY_ATTRIBUTES);
>>>
>>> CommonPropertyUtils.renderCommonPassthroughPropertiesWithoutEvents(writer,
>>>
>>> CommonPropertyUtils.getCommonPropertiesMarked(component), component);
>>>             }
>>>             else
>>>             {
>>>                 HtmlRendererUtils.renderHTMLAttributes(writer, component,
>>>                         HTML.BODY_PASSTHROUGH_ATTRIBUTES_WITHOUT_EVENTS);
>>>             }