[SOLVED] Looking for documentation on Reportlab's canvas.transform() function

[NeilUK]

I've looked 'reportlab-userguide_2022_v3_5.pdf', on page 16 under Changing Geometry, and there's no documentation beyond the mere mention of transform(a, b, c, d, e, f). It does go into more detail on scale() and translate(), the more common functions for beginners.

I've looked in 'reportlab-reference.pdf' on page 11, and get

adjoin a mathematical transform to the current graphics state matrix.
Not recommended for beginners.

Yeah, thanks!

If I open a python REPL, import reportlab, create a canvas.Canvas instance, and help(canvas.transform), I get exactly the same message as is in the reportlab-reference documentation, obviously rendered directly from the docstrings.

My general google fu is not that great, and hasn't thrown anything up yet.

What I'm trying to do is correct a small distortion of my printer. Maybe the simpler canvas.skew() would be sufficient for that. It's even more poorly documented, but should be easier to hope'n'poke/test.

Is there a graphics 'standard' set of six (a, b, c, d, e, f) parameters for transforms that Reportlab is 'obviously' using, that will be documented in other more thorough packages?

[NeilUK]

To answer my own question, I finally found something via a net search

docs.reportlab.com/rml/useguide/Chapter_10_More_graphics

It's in the reportlab markup language guide, which I'd ignored as I don't use the markup way of driving it.

However, transform(a, b, c, d, e, f) does this

x' = ax + cy + e
y' = bx + dy + f

Still no proper documentation on skew(alpha, beta), but at least it confirms that they are angles