Help with operatorList parsing

[Bboyredstar]

Hello!

I am trying to implement OCG data parsing. I have several tasks to do.

1. To highlight all OCG data when we hover over one of the elements (outline the elements).
2. Make it so that OCG elements can be selected by clicking. For example the user clicks on a wall and we remember it location to use it later on another canvas.

Pdf example:
Andheri_Residence_Electrical_Drawing.pdf

I have encountered a problem. By default we can't get OCG data by any inbuilt method, so I have to parse operatorList myself. I parse the content into groups, and after that I want to get data about polygons only, for example, by filtering the operators.

Something like:

operatorList.fnArray.forEach((fn, index) => {
       const args = operatorList.argsArray[index];

       if (fn === pdfjs.OPS.beginMarkedContentProps) {
         const property = args[0];
         if (property === 'OC') {
           ocgStack.push(args?.[1]?.id);
         }
       }

       if (fn === pdfjs.OPS.endMarkedContent) {
         if (ocgStack.length > 0) {
           const ocg = ocgStack.pop();
           if (ocgStack.length === 0) {
             const group = optionalContentConfig.getGroup(ocg);

             // console.log(group);
             const name = group?.name;

             if (!pagesContent[pageNum - 1]?.[name]) {
               pagesContent[pageNum - 1][name] = [];
             }

             pagesContent[pageNum - 1][name].push([...ocgContent]);
             ocgContent = [];
           }
         }
       }

       if (ocgStack.length > 0 && [91, 20].includes(fn) && args) {
         ocgContent.push({ fn, args });
       }

I have a few questions:

Is there some better way to not store so much data that we got after filtering operatorList?

The second problem is related to dimension conversion. I don't know how I can correctly convert the polygon's dimensions relative to its display on the canvas.

That is, let's say the dimensions of my canvas:

{
    "viewBox": [
        0,
        0,
        1191,
        842
    ],
    "scale": 1,
    "rotation": 0,
    "offsetX": 0,
    "offsetY": 0,
    "transform": [
        1,
        0,
        0,
        -1,
        0,
        842
    ],
    "width": 1191,
    "height": 842
}

And the polygon data is:

fn = 91
args = [
    [
        13,
        14,
        14,
        14,
        14
    ],
    [
        0,
        2526,
        39685,
        2526,
        39685,
        25537,
        0,
        25537,
        0,
        2526
    ],
    [
        0,
        2526,
        39685,
        25537
    ]
]

Obviously, its coordinates are outside the view box.

So I need to convert the data in points, to pixels. I tried dividing the coordinates by 72. But the scale turns out to be very small relative to the real size of the canvas.

I want to draw the corresponding elements on top of the linked elements on the pdf. I want them to be interactive.

Here is an example:
https://github.com/user-attachments/assets/e2ceb640-7cb1-427a-802d-c5ea732baf8d

As you can see now my data from the list operator does not match their coordinates on the pdf (I adjusted the coordinates a bit manually).

Is there any way to convert the data and how to do it?

Thanks!