Closed Jollywatt closed 2 days ago
Related discussion: cetz-package/cetz#604 and Jollywatt/typst-fletcher/issues/34.
This seems to apply to scale(reflow: true, …)
as well.
I experimented a bit with rotate(reflow: true)
. I have no idea to what extent this can be mapped to other things, but after the rotation of a rect(height: 100pt, width: 50pt)
, one of the lines appears to be the hypotenuse of the measured width and height. The "hypotenuse" is initially a vertical line of the original rectangle for 0 < α ≤ 90
and switches for every 90deg
(i.e. horizontal line — zero-height — for 90 < α ≤ 180
and so on).
Description
The dimensions of an object that has been rotated with
reflow: true
seem to be different according to 1) an enclosingrect
and 2) the contextualmeasure
function. Specifically, the bounding box reported by measure is that of a rotated zero-width object.MWE:
This is also observed for rotated objects other than text (see reproduction URL).
Perhaps related: the dimensions seem to be incorrect when the parent container sizes automatically to content:
Reproduction URL
https://typst.app/project/rUM1E564AvpxiKtmw50JMW
Operating system
Web app, Windows, Linux, macOS
Typst version