Questions regarding Wicket and HTML Ids

[Marek Šabo <ms...@buk.cvut.cz>]

Hi All,

I would like to ask about two separate pieces of code in the Wicket 
codebase working with HTML Ids:

1. org.apache.wicket.markup.repeater.AbstractRepeater
     This class uses regex SAFE_CHILD_ID_PATTERN and produces warning in 
case the repeater IDs are not just numerical. I found a post [1] where 
Igor gave an explanation eons ago. I tried to locate code that would 
rely on that fact but didn't find anything. Is it possible that the 
check and warning produced are obsolete nowadays? If so, I guess it 
could be removed (I'd happily provide pull request). Otherwise, would 
someone point me to the code that relies on this?

2. org.apache.wicket.DefaultMarkupIdGenerator
     This class, after generating an Id, goes on to "escape some 
noncompliant characters" (namely "_.- ") with underscores. I'm not sure 
what compliance it's referring to. I checked the HTML 4 spec [2] and 
HTML 5 specs [3]. All mentioned characters except whitespace seem to be 
legit in that context. Could someone shed some light on this -- e.g. 
what purpose does it exactly serve?


Many thanks in advance.


Best regards,

Marek


[1] 
http://apache-wicket.1842946.n4.nabble.com/Is-SAFE-CHILD-ID-PATTERN-in-AbstractRepeater-correct-or-does-it-break-CSS-spec-td1869301.html#a1869308
[2] http://www.w3.org/TR/html401/types.html#type-name
[3] http://www.w3.org/TR/html5/dom.html#the-id-attribute

Re: Questions regarding Wicket and HTML Ids

[Martin Grigorov <mg...@apache.org>]

Hi,

On Fri, Jan 23, 2015 at 11:56 PM, Marek Šabo <ms...@buk.cvut.cz> wrote:

> Hi All,
>
> I would like to ask about two separate pieces of code in the Wicket
> codebase working with HTML Ids:
>
> 1. org.apache.wicket.markup.repeater.AbstractRepeater
>     This class uses regex SAFE_CHILD_ID_PATTERN and produces warning in
> case the repeater IDs are not just numerical. I found a post [1] where Igor
> gave an explanation eons ago. I tried to locate code that would rely on
> that fact but didn't find anything. Is it possible that the check and
> warning produced are obsolete nowadays? If so, I guess it could be removed
> (I'd happily provide pull request). Otherwise, would someone point me to
> the code that relies on this?
>

This code is a "best effort" to make sure that the application code doesn't
use the same id for each iteration.
I agree that it is not as smart as one would wish.
Probably a better way would be to use a Set<String> and fail when a
duplicate id is detected during the rendering of the repeater.
A patch or Pull Request is welcome!


>
> 2. org.apache.wicket.DefaultMarkupIdGenerator
>     This class, after generating an Id, goes on to "escape some
> noncompliant characters" (namely "_.- ") with underscores. I'm not sure
> what compliance it's referring to. I checked the HTML 4 spec [2] and HTML 5
> specs [3]. All mentioned characters except whitespace seem to be legit in
> that context. Could someone shed some light on this -- e.g. what purpose
> does it exactly serve?
>

Well, it would be a perfect world if browsers were following the standards
but there are always exceptions. The escaping has been added because of a
bad behavior (read "wrong implementation").
Does the escaping break your application anyhow ?
The good thing is that now it is possible to provide custom
IMarkupIdGenerator. This interface has been added for better Portlet
support. AFAIR you are also Portlet user.


>
>
> Many thanks in advance.
>
>
> Best regards,
>
> Marek
>
>
> [1] http://apache-wicket.1842946.n4.nabble.com/Is-SAFE-CHILD-
> ID-PATTERN-in-AbstractRepeater-correct-or-does-it-break-CSS-spec-
> td1869301.html#a1869308
> [2] http://www.w3.org/TR/html401/types.html#type-name
> [3] http://www.w3.org/TR/html5/dom.html#the-id-attribute
>