replace bdd_satoneset(x,y,bddtrue) loops by minterms_of(x,y)

Because the loops iterate in the opposite order, multiple
test cases need to be adjusted.

* spot/taalgos/tgba2ta.cc, spot/twaalgos/alternation.cc,
spot/twaalgos/dualize.cc, spot/twaalgos/ltl2tgba_fm.cc,
spot/twaalgos/simulation.cc, spot/twaalgos/stutter.cc,
spot/twaalgos/toweak.cc: Replace loops based on
bdd_satonest(x,y,bddtrue) by loops based on minterms_of(x,y).
* tests/core/degenscc.test, tests/core/dualize.test,
tests/core/genltl.test, tests/core/readsave.test,
tests/python/alternation.ipynb, tests/python/automata.ipynb,
tests/python/decompose.ipynb, tests/python/decompose_scc.py,
tests/python/dualize.py, tests/python/sccinfo.py,
tests/python/simstate.py, tests/python/testingaut.ipynb,
tests/python/word.ipynb: Adjust expected test cases.  The
only regression is in genltl.test, but the worsened case
should eventually be fixed as discussed in issue #425 anyway.

tests/python/decompose_scc.py

@@ -1,5 +1,5 @@
 # -*- mode: python; coding: utf-8 -*-
-# Copyright (C) 2017 Laboratoire de Recherche et
+# Copyright (C) 2017, 2021 Laboratoire de Recherche et
 # Développement de l'Epita
 #
 # This file is part of Spot, a model checking library.
@@ -22,7 +22,14 @@ import spot
 aut = spot.translate('(Ga -> Gb) W c')
 si = spot.scc_info(aut)
 
-assert (spot.decompose_scc(si, 2).to_str('hoa', '1.1') == """HOA: v1.1
+# Extract the only rejecting SCC.  Its number might differ
+# if the generation of the automaton changes, so just scan
+# for it.
+rej = [j for j in range(si.scc_count()) if si.is_rejecting_scc(j)]
+assert len(rej) == 1
+s = spot.decompose_scc(si, rej[0]).to_str('hoa', '1.1')
+
+assert (s == """HOA: v1.1
 States: 3
 Start: 0
 AP: 3 "b" "a" "c"