How is the Ketcher file structured?

At a high level, a Ketcher file is structured to represent an object with a required property "root", which has a property "nodes" to hold an array of graphical objects. The hight level structure of a Ketcher file looks like this:

{
    "root": {
        "nodes": [
            <object>,
            <object>,
            ...,
            <object>
        ]
    }
}

Each graphical object has a required property "type" to specify the nature of the object. Here are some examples:

{    
    "type": "molecule",
    (A molecule structure)
}
{    
    "type": "text",
    (A text label)
}
{    
    "type": "arrow",
    (A line ended with an arrow )
}
{    
    "type": "plus",
    (A plus sign)
}

If the object is a "molecule", it must to be defined as a referenced-object. and move it the top level. The reason for the exception is probably to make the "nodes" structure short and easier to read. So the Ketcher file structure for a molecule with a text label looks like:

{
    "root": {
        "nodes": [
            {    
                "$ref": "mol<seq>"
            },
            {    
                "type": "text",
                ...
            }
        ]
    },
    "mol<seq>": {  
        "type": "molecule",
        ...
    }
}

Note that the name of the referenced-molecule must use the pattern of "mol<seq>", where <seq> is a sequential number.

⇒ Export Ketcher File from Editor

⇐ What Is Ketcher File Format

⇑ Ketcher File Format for Chemical Structures

⇑⇑ Ketcher - Chemical Structure Editor