How can I get the actual bounding borders of a SVG after rotation?

I need to get the position of the actual bounding borders of a SVG image after its parent being applied CSS rotation. The grey color in the image below indicates its parent element(In my case it is transparent - I added the grey color in the image below just for indicating the parent element's border)

getBoundingClientRect() or getBBox() seems to only show the bounding borders of the original rectangle - the red borders shown below. But what I want is the bounding borders without SVG's transparent background after rotation - like the green borders.

$(function() {

  let svgString = '<svg id="Layer_1" data-name="Layer 1" xmlns="http://www.w3.org/2000/svg" viewBox="0 0 425.58 635.88"><defs><style>.cls-1{fill:#000000;fill-rule:evenodd;}</style></defs><title>1</title><path id="path" class="cls-1" d="M918.94,831.44c-103,110.83-133.9,110.83-236.86,0-68.44-73.68-80.26-150.84-84.92-241.2-4.22-81.95-26.86-194.68,18.05-248.36,70.51-84.25,300.09-84.25,370.59,0,44.92,53.68,22.28,166.41,18.06,248.36C999.2,680.6,987.39,757.77,918.94,831.44Z" transform="translate(-587.72 -278.69)"/></svg>'

  let svgElement = new DOMParser().parseFromString(svgString, 'text/xml').documentElement

  $('#box').html(svgElement)

  $('#box').css({
    transform: 'rotate(60deg)'
  })

  $('#rect').click(function() {

    let rect = svgElement.getBoundingClientRect()

    $('#bounding-border').css({
      left: rect.left + 'px',
      top: rect.top + 'px',
      width: rect.width + 'px',
      height: rect.height + 'px',
    })
  })

  $('#bbox').click(function() {

    let bbox = svgElement.getBBox()

    $('#bounding-border').css({
      left: bbox.left + 'px',
      top: bbox.top + 'px',
      width: bbox.width + 'px',
      height: bbox.height + 'px',
    })
  })
})

#box {
  position: absolute;
  left: 100px;
  top: 100px;
  width: 100px;
  height: 150px;
}

svg {
  path:hover {
    fill: red;
  }
}

#bounding-border {
  position: absolute;
  border: 2px solid red;
  z-index: -1;
}

<script src="https://cdnjs.cloudflare.com/ajax/libs/jquery/3.7.1/jquery.min.js"></script>

<button id="rect">getBoundingClientRect</button>
<button id="bbox">getBBox</button>
<div id="box"></div>
<div id="bounding-border"></div>

codepen

Solution

To this date you can't tweak native methods to get a tight bounding box as expected by the visual boundaries after transformations.

The best you can do is

rotate SVG child elements instead of the parent SVG
convert the transformed Geometry elements to paths with "hard-coded" pathData coordinates

$(function() {
  let svgString =
    `<svg id="Layer_1" data-name="Layer 1" xmlns="http://www.w3.org/2000/svg" viewBox="0 0 425.58 635.88"><defs><style>.cls-1{fill:#000000;fill-rule:evenodd;}</style></defs><title>1</title><path id="path" class="cls-1" d="M918.94,831.44c-103,110.83-133.9,110.83-236.86,0-68.44-73.68-80.26-150.84-84.92-241.2-4.22-81.95-26.86-194.68,18.05-248.36,70.51-84.25,300.09-84.25,370.59,0,44.92,53.68,22.28,166.41,18.06,248.36C999.2,680.6,987.39,757.77,918.94,831.44Z" 
    transform-origin="center" 
  transform="rotate(60) translate(-587.72 -278.69)"/></svg>`;

  let svgElement = new DOMParser().parseFromString(svgString, "text/xml")
    .documentElement;
  let path = svgElement.getElementById("path");

  // append svg
  $("#box").html(svgElement);


  /**
   * "flatten" transformations
   * for hardcoded path data cordinates
   */
  flattenSVGTransformations(svgElement);


  // get bounding box of flattened svg
  let bb = svgElement.getBBox();


  // adjust viewBox according to bounding box
  svgElement.setAttribute('viewBox', [bb.x, bb.y, bb.width, bb.height].join())

  // get DOM bounding box
  let rect = svgElement.getBoundingClientRect();

  // adjust absolutely positioned wrapper
  $("#bounding-border").css({
    left: rect.x + "px",
    top: rect.y + "px",
    width: rect.width + "px",
    height: rect.height + "px"
  });

});


/**
 * flattening/detransform helpers
 */

function flattenSVGTransformations(svg) {
  let els = svg.querySelectorAll('text, path, polyline, polygon, line, rect, circle, ellipse');
  els.forEach(el => {
    // convert primitives to paths
    if (el instanceof SVGGeometryElement && el.nodeName !== 'path') {
      let pathData = el.getPathData({
        normalize: true
      });
      let pathNew = document.createElementNS('http://www.w3.org/2000/svg', 'path');
      pathNew.setPathData(pathData);
      copyAttributes(el, pathNew);
      el.replaceWith(pathNew)
      el = pathNew;
    }
    reduceElementTransforms(el);
  });
  // remove group transforms
  let groups = svg.querySelectorAll('g');
  groups.forEach(g => {
    g.removeAttribute('transform');
    g.removeAttribute('transform-origin');
    g.style.removeProperty('transform');
    g.style.removeProperty('transform-origin');
  });
}

function reduceElementTransforms(el, decimals = 3) {
  let parent = el.farthestViewportElement;
  // check elements transformations
  let matrix = parent.getScreenCTM().inverse().multiply(el.getScreenCTM());
  let {
    a,
    b,
    c,
    d,
    e,
    f
  } = matrix;
  // round matrix
  [a, b, c, d, e, f] = [a, b, c, d, e, f].map(val => {
    return +val.toFixed(3)
  });
  let matrixStr = [a, b, c, d, e, f].join('');
  let isTransformed = matrixStr !== "100100" ? true : false;
  if (isTransformed) {


    // if text element: consolidate all applied transforms 
    if (el instanceof SVGGeometryElement === false) {
      if (isTransformed) {
        el.setAttribute('transform', transObj.svgTransform);
        el.removeAttribute('transform-origin');
        el.style.removeProperty('transform');
        el.style.removeProperty('transform-origin');
      }
      return false
    }
    /**
     * is geometry elements: 
     * recalculate pathdata
     * according to transforms
     * by matrix transform
     */
    let pathData = el.getPathData({
      normalize: true
    });
    let svg = el.closest("svg");
    pathData.forEach((com, i) => {
      let values = com.values;
      for (let v = 0; v < values.length - 1; v += 2) {
        let [x, y] = [values[v], values[v + 1]];
        let pt = new DOMPoint(x, y);
        let pTrans = pt.matrixTransform(matrix);
        // update coordinates in pathdata array
        pathData[i]["values"][v] = +(pTrans.x).toFixed(decimals);
        pathData[i]["values"][v + 1] = +(pTrans.y).toFixed(decimals);
      }
    });
    // apply pathdata - remove transform
    el.setPathData(pathData);
    el.removeAttribute('transform');
    el.style.removeProperty('transform');
    return pathData;
  }
}


/**
 * get element transforms
 */
function getElementTransform(el, parent, precision = 6) {
  let matrix = parent.getScreenCTM().inverse().multiply(el.getScreenCTM());
  let matrixVals = [matrix.a, matrix.b, matrix.c, matrix.d, matrix.e, matrix.f].map(val => {
    return +val.toFixed(precision)
  });
  return matrixVals;
}


/**
 * copy attributes:
 * used for primitive to path conversions
 */
function copyAttributes(el, newEl) {
  let atts = [...el.attributes];
  let excludedAtts = ['d', 'x', 'y', 'x1', 'y1', 'x2', 'y2', 'cx', 'cy', 'r', 'rx',
    'ry', 'points', 'height', 'width'
  ];
  for (let a = 0; a < atts.length; a++) {
    let att = atts[a];
    if (excludedAtts.indexOf(att.nodeName) === -1) {
      let attrName = att.nodeName;
      let attrValue = att.nodeValue;
      newEl.setAttribute(attrName, attrValue + '');
    }
  }
}

* {
  box-sizing: border-box;
}

svg {
  display: block;
  outline: 1px solid #ccc;
  overflow: visible;
}

#box {
  position: absolute;
  width: 75%;
  outline: 1px solid #ccc;
}


#bounding-border {
  position: absolute;
  border: 2px solid rgba(255, 0, 0, 0.5);
}

<script src="https://cdnjs.cloudflare.com/ajax/libs/jquery/3.7.1/jquery.min.js"></script>

<!-- path data parser -->
<script src="https://cdn.jsdelivr.net/npm/path-data-polyfill@1.0.4/path-data-polyfill.min.js"></script>


<div id="box"></div>
<div id="bounding-border"></div>

Obviously, the detransformation/flattening process will change the current transformation.
We also need a path data parser to get calculable absolute command coordinates. I'm using Jarek Foksa's getpathData() polyfill (which may soon become obsolete at least for Firefox =).

Worth noting, we need to adjust the svg's viewBox after flattening – otherwise we won't get the correct layout offsets via getBoundingClientRect() as this method respects the SVG's viewBox and thus returns a 'clipped' bounding box.