mirror of
https://github.com/svg/svgo.git
synced 2026-01-25 18:41:39 +03:00
50c326bbff
Small group of changes that at reduce the regression test runtime by about 14%. Motivation for this change in particular were the large PNGs usually allocated for even the smallest SVGs in the test data which both increases the time spent allocating as well as the time pixelmatch would take for the comparison. The other changes improve not only the test runtime but also the normal SVG processing given the input data has cases caught by removeHiddenElems which also has the by far most significant of the 3 changes. All the changes done for baseline performance optimization are very unlikely to cause any harm and would need specific sets of input SVGs mostly consisting of visibility checks regarding the element style to impact performance in a negative way. In my local tests these changes gave an additional ~7-8% time reduction with the svgo data set.