How to improve line breaks in Document Intelligence's OCR output?

Question

Hello everyone,

I'm currently working with an image of code that I need to convert into text using OCR. The issue is that the code uses line breaks in a way that, if modified, could lead to compile/runtime errors.

Here's an example of the OCR output I am getting:

Paragraph:
public static int foo(int bar) {
==========
Paragraph:
bar++; if (bar < 10) bar = foo(bar);
==========
Paragraph:
int i = 0; int j = 0; while (i > foo(j - bar)) { j++; bar += j;
==========
Paragraph:
3
==========
Paragraph:
return bar;
==========
Paragraph:
}
==========

However, the code should look like this:

public static int foo(int bar) {
    bar++;
    if (bar < 10)
        bar = foo(bar);

    int i = 0;
    int j = 0;
    while (i > foo(j - bar)) {
        j++;
        bar += j;
    }

    return bar;
}

Here's the actual image used for the ocr:

Is there any way to make the Document Intelligence OCR output line breaks better, ensuring the code is correctly formatted?

Best regards,
YK