Summary: Normalize conditional formulas to ensure that their "condition" formula is not "negative". This avoids redundant formulas such as `(x = 0 ? p : q)` and `(x ≠ 0 ? q : p)`. The choice of which formulas are "negative" for this purpose is mostly arbitrary, with the only real constraint being that negating a negative formula should produce a positive one. Note that conditional formulas themselves are considered to be "positive" since negating them produces another conditional formula with the same condition formula. Reviewed By: ngorogiannis Differential Revision: D23487502 fbshipit-source-id: 63606d89cmaster
parent
74086e926a
commit
d8114b4688
Loading…
Reference in new issue