joelpt
commented
12 years ago joe...@gmail.com - 2012-03-22
Good point. 4d/4u type transforms basically are currently executed after all the other transforms separately, during a later stage. I probably will not fix this real soon but I'll keep it on the radar. It's a non trivial change and I need to think through how this would be impacted by the "two buckets" logic of the existing transform system; e.g. what should happen if we say "fliph 2d flipv 2d".
For a somewhat clunky workaround you can have qfconvert output a transformed csv file and then manually tack it onto your original csv file. Here's an example of how you could create a "flipped" csv which you could then combine as desired with your original csv.
qfconvert --transform="fliph" --mode=csv original.csv > flipped.csv
Not sure if this is the correct behaviour. I found it very unintuitive.
What steps will reproduce the problem?
What is the expected output? What do you see instead?
Expected output is the pattern being horizontally flipped for every level that it's printed on. Like "fliph 4e" behaves.
What version of the product are you using? On what operating system?
Windows XP, DF 0.31.25, QF 2.03
Please provide any additional information below.
"fliph 4d" behaves in exactly the same way as "fliph ! 4d" meaning there's no way to rotate a pattern while repeating downwards.
Original issue: http://code.google.com/p/quickfort/issues/detail?id=72 (2012-02-14)