Skip to content

High-pass data compression filter for inline styling (embedded CSS).

Notifications You must be signed in to change notification settings

zm-cttae/dom-inline-style-filter

Repository files navigation

dom-inline-style-filter

dom-inline-style-filter library filters inline style declarations for a standalone DOM element tree by computed effect.

  • As web developers, we would like elements that ship only with inline styling to be light so that they can be included in NPM packages.
  • A main use case of this is SVG screenshots of HTML elements.
  • Even after a filter algorithm to filter out user agent styling when inlining the style, there is some way to go with data size.

Usage

dominlinestylefilter(node)

Parameter: node - a HTMLElement with all style rules embedded as inline style attributes or <style> tags.

Returns: a Promise that resolves to node. Within node, all inline styling has been filtered to the minimum declarations that produce the same computed style.

dominlinestylefilter.sync(node)

Synchronous version. Returns node when the styling compression is completed.

Optimizations

  1. When traversing DOM tree of node, group nodes by descending node depth.

    CSS inheritance is computed on the DOM tree via preorder traversal and is additive-cumulative (increases styling data).

    For the filter op which is subtractive, we want to traverse the tree in the opposite direction.

    The algorithm sorts elements in the node tree by descending node depth. (This is known as reverse level order traversal.)

    This gives us a 30% to 40% speed boost. This also ensures declarations are only removed when they really can be inherited.

  2. When filtering each inline style declaration by computed effect, go for the most hyphenated properties first.

    In CSS, shorthands consistently have less hyphens than their longhand.

    We want to filter out scenarios where a CSS property matches their shorthand, e.g. block-size -> height or border-color -> border.

    The algorithm does a radix sort with bitmasks for standard, custom and vendored proprties, then subsorts by descending hyphen count.

    In tests this filtered another 50% of inline styling. We also get a 20-40% speed boost because we're not setting as many properties back.

Performance

The underlying algorithm was determined to be a high-pass multi-pass - $N \approx 4$ - deterministic compression in two modes.

The data was collected from manual testing on the output of the domtoimage.toSvg function in the dom-to-image-more NPM package.

Large file inputs

$O(log(N))$ growth for inputs at large filesizes $|F| &gt;&gt; 1e6 \text{ bytes}$.

Wikipedia article demo Value
Number of nodes 5558 nodes
Initial declaration count 177818 (31.9 declarations / node)
Pre-compression bytes 3.63mb
Reductions [3058654, 98781, 16774, 0]
Processing time 10316.5ms (1.86 ms/node)
Total reduction 3.17mb
Output declaration count 33643 (6.05 / node)
Post-compression bytes 709.4kb
Compression quotients [0.9698, 0.9991, 0.9999, 1]
Compression ratio 5.117
Decay formula $1-exp(-7 / 2 \cdot N)$

Graph

Small file results

$O(c \cdot N), \space c \space \approx \space 4$ growth for inputs at small filesizes $|F| &lt;&lt; 1e6\space\text{ bytes}$.

Code screenshot demo Value
Number of nodes 420 nodes
Initial declaration count 14933 (35.5 declarations / node)
Pre-compression bytes 372430b
Reductions [275482, 40312, 0]
Processing time 604ms (1.6 ms / node)
Total reduction 315794b
Post-compression bytes 56636b
Output declaration count 2443 (5.82 / node)
Compression quotients [0.872, 0.999, 1]
Total quotient (compound) 6.575
Decay formula $1-exp(-9 / 4 \cdot N)$