Lavoisier

Since renderers must be able to render any XML data, this choice would produce quite complex JSON outputs:

<ul>
<li style="color:red">First Item</li>
<li title="Some hover text." style="color:green">
Second Item
</li>
<li><span class="code-example-third">Third</span>
Item</li>
</ul>

{
    "ul":
    {
        "li":
        [
            [
                { "style" : "color:red" },
                "First Item" 
            ],
            [
                {
                    "title" : "Some hover text.",
                    "style" : "color:green" 
                },
                "Second Item" 
            ],
            [
                {
                    "span":
                    [
                        { "class" : "code-example-third" },
                        "Third" 
                    ]
                },
                " Item" 
            ]
        ]
    }
}

Can you please confirm that such a complexity does not make this rendering useless?

Modify the existing JSON renderer.

Add parameters:

• element-as-array=TRUE|false. If element-as-array, then sibling elements with same name are grouped into an array (<li>a</li><li>b</li> = "li":[]).

• text-as-pair=TRUE|false. If text-as-pair, then text nodes are converted to pairs "#text": "value".

• abort-on-conflict=true|FALSE. If abort-on-conflict, then conflicting names will raise an exception. Else conflicting names will be renamed to "name#1", "name#2"...

Text nodes are trimed, and empty text strings are ignored.

Namespaces will be ignored.