[jira] Updated: (WICKET-789) New ExtensibleChoiceAutocomplete

["Frank Bille Jensen (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/WICKET-789?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Frank Bille Jensen updated WICKET-789:
--------------------------------------

    Fix Version/s:     (was: 1.3.0-rc1)
                   1.3.0-rc2

> New ExtensibleChoiceAutocomplete
> --------------------------------
>
>                 Key: WICKET-789
>                 URL: https://issues.apache.org/jira/browse/WICKET-789
>             Project: Wicket
>          Issue Type: Improvement
>          Components: wicket-extensions
>         Environment: all platforms
>            Reporter: Francisco Diaz Trepat
>            Assignee: Janne Hietamäki
>            Priority: Minor
>             Fix For: 1.3.0-rc2
>
>         Attachments: AutocompleteTestCase.rar, AutocompleteTestCase.rar
>
>
> The idea is to have an autocomplete behavior that allows you to have a key and value if you need it. This is done by either including a hidden field or some other component as the key receiver when a choice gets selected, or it can be left out and works exactly like the one that is know available in the extensions package. 
> The thing started out as the current implementation used a fixed "node path" to get the elements that needed to render select etc. Now it uses an option/choice name. 
> To do this I had to re-write the JavaScript and All the abstract classes so I did a new package all together.
> And renamed all of the classes to have both of them without any conflicts. I also included a new collection class in the javascript to help out and I putted on a Wicket.Collection "package", and also named this autocomplete as ExtensibleChoiceAutocomplete. Extensible because you can put in it any html markup you feel like and it will work (choose arrows etc), Choice because it can use key and value like a DropdownChoice, and Autocomplete because it is an autocomplete.  
> All naming convention might be wrong and some of my code might not make the grade for you guys, so what I would like is to have someone review it and see if it is considered to be included somewhere. 
> Wicket is awesome, I get lots of help all the time, and just want to contribute.
> There are some algorithms in the current Autocomplete JavaScript that didn't worked on IE 7 and IE 6 because of a well known issue on IE that is that it does not refresh document DOM on the assignment of the innerHTML property of an element. But that is taken care of by using element.all collection that works for IE 7 and 6.
> Both wicket-extenssions and scriptaculos don't support just any html. They both support putting <ul><li> with inside them any html like images <img> and divs spans <p> etc. It also supports puting <div><div> or span div or span span etc on wicket if you change a little bit. 
> But that doesn't mean any html. For instance you cannot put a table with rows and colunms and then have the rows painted over key controls like up down left & right  arrows.
> In scriptaculous looks for ul and the wicket one lives that open with the renderer but it is fixed on the JavaScript by looking at the, menu.firstChild.childNodes , that means that inside the menu container you cannot put any html that whould argue that path that is.
> Menu Container
>   Your Choices Container
>     Your Choices
> Inside a table you have:
> Menu Container
>    Your Table
>       Your Table Body (this is put in the DOM whether you specify it or not. 
>          Your Rows 
> This is not accesible due to the firstChild.childNodes which in this case would translate to table.tableBody and not the tr that you maight need. Also the event handlers for key down key pressed etc are similar.
> Also my code is usually considered heavily commented so as I see wicket code in general is not, maybe the reviewer would have to erase some comments.  
> Feel Free is the key. I want to contribute, keep contributing in the future, and learn a proper way not to step on any one's feet.
> Best regards,
> f(t)

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.