How can i understand the algorithm wu line?

Question

I have Wu's line algorithm but I don't understand all the lines of code very well.

public static void drawLine(Graphics g, double x0, double y0, double x1, double y1, int value) {
     
        boolean steep = abs(y1 - y0) > abs(x1 - x0);
        if (steep)
            drawLine(g, y0, x0, y1, x1, value);
     
        if (x0 > x1)
            drawLine(g, x1, y1, x0, y0, value);
     
        double dx = x1 - x0;
        double dy = y1 - y0;
        double gradient = dy / dx;
     
        // handle first endpoint
        double xend = round(x0);
        double yend = y0 + gradient * (xend - x0);
        double xgap = rfpart(x0 + 0.5);
        double xpxl1 = xend; // this will be used in the main loop
        double ypxl1 = ipart(yend);
     
        if (steep) {
            plot(g, ypxl1, xpxl1, rfpart(yend) * xgap);
            plot(g, ypxl1 + 1, xpxl1, fpart(yend) * xgap);
        } else {
            plot(g, xpxl1, ypxl1, rfpart(yend) * xgap);
            plot(g, xpxl1, ypxl1 + 1, fpart(yend) * xgap);
        }
     
        // first y-intersection for the main loop
        double intery = yend + gradient;
     
        // handle second endpoint
        xend = round(x1);
        yend = y1 + gradient * (xend - x1);
        xgap = fpart(x1 + 0.5);
        double xpxl2 = xend; // this will be used in the main loop
        double ypxl2 = ipart(yend);
     
        if (steep) {
            plot(g, ypxl2, xpxl2, rfpart(yend) * xgap);
            plot(g, ypxl2 + 1, xpxl2, fpart(yend) * xgap);
        } else {
            plot(g, xpxl2, ypxl2, rfpart(yend) * xgap);
            plot(g, xpxl2, ypxl2 + 1, fpart(yend) * xgap);
        }
     
        // main loop 
        for (double x = xpxl1 + 1; x <= xpxl2 - 1; x++) {
            if (steep) {
                plot(g, ipart(intery), x, rfpart(intery));
                plot(g, ipart(intery) + 1, x, fpart(intery));
            } else {
                plot(g, x, ipart(intery), rfpart(intery));
                plot(g, x, ipart(intery) + 1, fpart(intery));
            }
            intery = intery + gradient;
        }
    }

I know this information:

The intensity of the pixel above the line must be proportional to (1-a), and the intensity of the pixel below the line must be proportional to (1-b). I.E. the closer the line is to a pixel, the brighter it is.

Why are the following steps necessary:

• handle the first endpoint?
• handle the second endpoint?
• main loop?