* spot/twaalgos/degen.cc, spot/twaalgos/degen.hh: Implement this.
Also throw a runtime error in case were todegen does not match any
subformula.
* tests/python/pdegen.py: Add tests.
On these 4 cases, added to pdegen.py, and supplied by Florian Renkin,
partial_degeneralize() is now at least as good as degeneralize_tba(),
and sometimes better. This is achieved as follows: (1) a
propagate_marks procedure is introduced to propagate marks as far as
possible on the automaton (e.g., common outgoing marks can be push
onto the incoming transitions and vice-versa), (2) the
degeneralization order is compute dynamically, and (3) whenever and
fully-accepting transition is taken in the original automaton, the
destination level is chosen to be the highest existing level.
* spot/twaalgos/degen.cc,
spot/twaalgos/degen.hh (propagate_marks_vector, propagate_marks_here):
New functions.
(partial_degeneralize): Improve, as described above.
* tests/python/pdegen.py: Add test cases.
Reported by František Blahoudek.
* spot/twa/twagraph.cc: Reset prop_universal() if edges are merged in
a non-deterministic automaton.
* tests/core/det.test: Add test case.
* NEWS: Mention the issue.
Reported by František Blahoudek.
* spot/twaalgos/postproc.cc: Turn "t" into "Inf(0)" for BA.
* tests/core/ltl2tgba.test: Add test case.
* NEWS: Mention the bug.
* spot/misc/tmpfile.cc: Display strerror(errno) plus some suggestions
that depend on the error. Based on a report from Shengping Shaw.
* THANKS: Add reporter.
* tests/core/ltlcross5.test: New file.
* tests/Makefile.am: Add it.
* spot/twaalgos/degen.hh, spot/twaalgos/degen.cc: Implement it.
* tests/python/pdegen.py: New file.
* tests/Makefile.am: Add it.
* NEWS: Mention the new function.
* spot/twaalgos/remfin.cc (tra_to_tba): Remove the SCC-emptiness
check that was done by creating a temporary automaton. Use the
scc_info::check_scc_emptiness() function instead.
* spot/twaalgos/sccinfo.cc,
spot/twaalgos/sccinfo.hh (scc_info::check_scc_emptiness): Mark
this function as const.
(scc_and_mark_filter): Make sure we only combine with a lower filter
of the same type.
* spot/twaalgos/genem.cc: Remove one stray debugging statement.
* tests/python/genem.py: Remove a duplicate test.
This fixes a complementation bug reported by Juraj Major and Tereza
Šťastná.
* spot/twaalgos/remfin.cc (is_scc_tba_type): Fix the condition for
handling Fin-alone pairs.
* tests/core/complement.test: Add Juraj & Tereza's test case.
* NEWS: Mention it.
This is to workaround differences in minidom's pretty-printing that
occurred between Python 3.7 and 3.8.
* python/spot/jupyter.py (SVG): New class.
* python/spot/__init__.py: Use it.
* tests/python/_altscc.ipynb, tests/python/alternation.ipynb,
tests/python/automata.ipynb, tests/python/formulas.ipynb,
tests/python/gen.ipynb, tests/python/highlighting.ipynb,
tests/python/ltsmin-dve.ipynb, tests/python/ltsmin-pml.ipynb,
tests/python/product.ipynb, tests/python/randaut.ipynb,
tests/python/testingaut.ipynb, tests/python/twagraph-internals.ipynb,
tests/python/word.ipynb: Adjust.
* spot/twa/acc.hh (acc_cond::mark_t): Diagnose mark_t with set numbers
that are too larges.
* tests/python/except.py: Adjust.
* tests/core/acc.cc: Remove most of asserts, as those can be disabled,
and adjust expected exception.
* spot/twaalgos/determinize.cc: Implement various micro
optimizations. Use robin-hood hashing in a few places.
* spot/priv/robin_hood.hh: New file.
* spot/priv/Makefile.am: Add it.
* tests/sanity/80columns.test, tests/sanity/includes.test,
tests/sanity/style.test: Adjust to skip robin_hood.hh.
* spot/twa/acc.hh: Introduce parity_min_odd(n) and friends.
* spot/twaalgos/determinize.cc, spot/twaalgos/rabin2parity.cc,
spot/twaalgos/toparity.cc: Use them.
* tests/python/parity.py: Call each function exhaustively.
* NEWS: Mention the new functions.
* spot/twaalgos/aiger.cc: Simplify some bit operatitions. Force the
ordering of operations in aig_and, that was causing a test case to
fail on ARM, and greatly simplify the code and data structures used in
remove_unused().
* tests/core/ltlsynt.test: Adjust expected output.
* spot/twaalgos/aiger.cc: Simplify some bit operatitions. Force the
ordering of operations in aig_and, that was causing a test case to
fail on ARM, and greatly simplify the code and data structures used in
remove_unused().
* tests/core/ltlsynt.test: Adjust expected output.
Suggested by Victor Khomenko.
* bin/ltl2tgba.cc, bin/ltldo.cc: Implement it.
* doc/org/hierarchy.org: Use it.
* tests/core/ltldo2.test: Test it.
* bin/common_output.cc: Typo.
* NEWS: Mention the new option.
Suggested by Victor Khomenko.
* bin/ltl2tgba.cc, bin/ltldo.cc: Implement it.
* doc/org/hierarchy.org: Use it.
* tests/core/ltldo2.test: Test it.
* bin/common_output.cc: Typo.
* NEWS: Mention the new option.
This was prompted by reports by Andrew Wells and Yong Li.
* NEWS, doc/tl/tl.tex: Document the changes.
* THANKS: Add Andrew.
* bin/ltlfilt.cc: Match --ltl before --from-ltlf if needed.
* spot/parsetl/parsedecl.hh, spot/parsetl/parsetl.yy,
spot/parsetl/scantl.ll: Parse X[!].
* spot/tl/formula.cc, spot/tl/formula.hh: Declare the new operator.
* spot/tl/ltlf.cc: Adjust to handle op::X and op::strong_X correctly.
* spot/tl/dot.cc, spot/tl/mark.cc, spot/tl/mutation.cc,
spot/tl/print.cc, spot/tl/simplify.cc, spot/tl/snf.cc,
spot/tl/unabbrev.cc, spot/twa/formula2bdd.cc,
spot/twaalgos/ltl2taa.cc, spot/twaalgos/ltl2tgba_fm.cc,
tests/core/ltlgrind.test, tests/core/rand.test,
tests/core/sugar.test, tests/python/randltl.ipynb: Adjust.
* tests/core/ltlfilt.test, tests/core/sugar.test,
tests/core/utf8.test: More tests.
* tests/Makefile.am (ltsmin_modelcheck_LDADD): Add -lpthread as it
seems Debian's libtool does not carries this dependency over from
libspotltsmin.la. Also using $(LTLIBMULTITHREAD) does not work,
because that would add -pthread which is currently ignored when
linking shared libraries (because libtool adds -nostdlib for some
reason).
std::cerr will flush after each operator<< by default, so it's simpler
to use \n instead of std::endl, especially if we can merge \n into the
previous string. Ideally we should prefer \n for std::cout as well,
but there are reasonable cases where we want to call std::endl there,
so it's hard to enforce.
* tests/sanity/style.test: Diagnose occurrences of cerr.*<<.*endl.
* bin/autcross.cc, bin/autfilt.cc, bin/ltlcross.cc, bin/ltlsynt.cc,
spot/tl/formula.cc, spot/twa/bdddict.cc, tests/core/checkpsl.cc,
tests/core/checkta.cc, tests/core/consterm.cc, tests/core/emptchk.cc,
tests/core/equalsf.cc, tests/core/ikwiad.cc, tests/core/kind.cc,
tests/core/length.cc, tests/core/ltlrel.cc, tests/core/parity.cc,
tests/core/randtgba.cc, tests/core/reduc.cc, tests/core/syntimpl.cc,
tests/ltsmin/modelcheck.cc: Fix them.
Reported by Salomon Sickert.
* spot/twaalgos/complement.cc: Make sure the output of postproc
is deterministic.
* tests/core/ltlcross.test: Add test case.
* NEWS: Mention the bug.
* spot/twaalgos/product.cc: The res pointer should be
passed by reference since we reset it to nullptr when
output_aborter says "too large".
* python/spot/impl.i: Add binding for powerset.hh,
so we can use output_aborter in Python.
* tests/python/prodexpt.py: Test it.
The previous patch triggered this issue again, failing
core/ltl2tgba2.test.
* spot/twaalgos/gfguarantee.cc: Separate the replaying of history from
the modification of the automaton.
* NEWS: Mention the bug.
* tests/python/twagraph-internals.ipynb, tests/python/automata.ipynb:
Adjust.
These rules come from Delag's paper, and help some cases
in issue #385.
* spot/tl/simplify.cc: Implement the simplification.
* doc/tl/tl.tex, NEWS: Document it.
* tests/core/385.test: New file.
* tests/Makefile.am: Add it.
* tests/core/reduccmp.test: More tests.
* tests/core/ltl2tgba2.test: Adjust one improved case.
* tests/python/automata.ipynb, tests/python/twagraph-internals.ipynb:
Adjust expected output, as the cnf/dnf reorder some subformulas.
* bin/autcross.cc: Let complement() decide how to complement automata.
Do not apply remove_fin(), because we have a generic emptiness check
now. Use intersecting_word() instead of product()+accepting_word() so
that the former can maybe be optimized in the future.
* tests/core/autcross2.test: Adjust test case to use TGBA instead
of monitors, as calling complement() had a side-effect of setting
the "weak" property on the input.
