With the patch in SF bug 1952642, you can use the fourier_elim preprocessor
this way:
(%i16) load("fourier_elim.lisp")$
(%i17) standardize_ineq(e) := ?standardize\-inequality(e)$
(%i18) standardize_ineq(abs(x-6) > 7);
(%o18) x-13>0 or -(x+1)>0
(%i20) standardize_ineq(abs(x-6) < 7);
(%o20) -(x-13)>0 and x+1>0
(%i23) standardize_ineq((x-6) # 0);
(%o23) x-6>0 or -(x-6)>0
The output is hard for me to read---it needs parens:
(%i24) standardize_ineq(max(x,-x) > 42);
(%o24) -(x+42)>0 and -x>0 or x-42>0 and (x>0 or x=0)
(%i25) args(%);
(%o25) [-(x+42)>0 and -x>0,x-42>0 and (x>0 or x=0)]
Of course, x-42>0 and (x>0 or x=0) simplifies to x - 42 > 0. But
standardize-inequality
doesn't do that.
Barton