* configure.ac: Update Vaucanson version to 1.3.9.

* include/vaucanson/algorithms/sub_normalized.hxx: Downgrade (check
error for fmp).

* taf-kit/src/Makefile.am: add
taf-kit/src/star_alphabet_command.hh, and
taf-kit/src/star_alphabet_command.hx

* include/vaucanson/algorithms/infiltration.hxx,
* include/vaucanson/algorithms/sub_normalize.hxx,
* include/vaucanson/algorithms/shuffle.hxx: Correct
bug.

* include/vaucanson/algebra/implementation/semiring/q_number.hh,
* include/vaucanson/algebra/implementation/semiring/q_number.hh:
Chagen type for RationalNumber.

* include/vaucanson/algebra/implementation/semiring/q_number.hxx:
Correct rational number operator.

* taf-kit/src/cmd_wt.cc,
* taf-kit/src/star_alphabet_command.hh,
* taf-kit/src/star_alphabet_command.hxx:
Add star_alphabet in taf-kit factory.

* configure.ac,
* data/automata/Makefile.am,
* data/automata/char-b/Makefile.am,
* data/automata/char-b/evena.cc,
* data/automata/char-b/oddb.cc,
* data/automata/char-char-b/Makefile.am,
* data/automata/char-f2/Makefile.am,
* data/automata/char-f2/ring.cc,
* data/automata/char-fmp-z/Makefile.am,
* ata/automata/char-int-b/Makefile.am,
* data/automata/char-q/Makefile.am,
* data/automata/char-r/Makefile.am,
* data/automata/char-zmax/Makefile.am,
* data/automata/char-zmax/gag.cc,
* data/automata/char-zmax/maxab.cc,
* data/automata/char-zmin/Makefile.am,
* data/automata/char-zmin/minab.cc,
* data/automata/char-zmin/sag.cc,
* data/automata/int-b/Makefile.am,
* data/automata/int-z/Makefile.am: Create demonstration
in data/automata for int-z, int-b, char-zmin, char-zmax,
char-r, char-q, char-int-b, char-fmp-z, char-f2, char-char-b,
and char-b. Create Makefile.am for demonstration, and update
main Makefile.am

* taf-kit/src/library_command.hh: Correct three letters file
bug.

* include/vaucanson/algebra/implementation/q_number.hh,
* include/vaucanson/algebra/implementation/q_number.hxx: Change
data type to 'long long' instead of int. Correct operator== for
null numbers.
* include/vaucanson/misc/algebra.hh,
* include/vaucanson/misc/algebra.hxx: Change type for lcm pgm
and is_coprime functions (unsigned long long).

* include/vaucanson/algorithms/eps_removal.hxx
(positive_path_covering): Fix iteration on supp().

* include/vaucanson/xml/builders.hxx (get_semiring_operations):
Add specialization to output CyclicSemirings as "field".
* data/vaucanson.xsd: Update it.

* taf-kit/src/pipe.hh: Declare OUTPUT_TYPE_FPEXP.
* taf-kit/src/pipe_writers.hxx: Handle it.
* taf-kit/src/main.cc: Handle "-o EXP" and "-o FPEXP".

* include/vaucanson/algorithms/eps_removal.hxx
(EpsilonRemover::test_for_non_positive_semiring): Move ...
(test_for_non_positive_semiring::run): ... here in a separate
struct that we can specialize if the semiring is positive.

on FMP).

* include/vaucanson/algorithms/eps_removal.hxx
(positive_path_covering): Use "weight < semiring_elt_zero" instead
of "weight.value() < 0".

* include/vaucanson/algebra/concept/semiring_base.hh
(semiring_traits<T,U>): New trait.
* include/vaucanson/algebra/implementation/semiring/numerical_semiring.hh,
include/vaucanson/algebra/implementation/semiring/numerical_semiring.hxx,
include/vaucanson/algebra/implementation/semiring/tropical_semiring.hh,
include/vaucanson/algebra/implementation/semiring/tropical_semiring.hxx,
include/vaucanson/algebra/implementation/series/krat.hh,
include/vaucanson/algebra/implementation/series/krat.hxx
(is_positive_semiring): Remove and...
(semiring_traits<T,U>): ... specialize.
* taf-kit/src/cmd_gen.cc (semiring-info): Adjust caller.

* include/vaucanson/algorithms/factor.hh,
include/vaucanson/algorithms/factor.hxx,
include/vaucanson/algorithms/prefix.hh,
include/vaucanson/algorithms/prefix.hxx,
include/vaucanson/algorithms/suffix.hh,
include/vaucanson/algorithms/suffix.hxx: Add is_trim precondition
as reported by Jacques Sakarovitch.
* src/tests/sanity/hhck.pl: Add support from checking is_trim()
preconditions.

* include/vaucanson/algorithms/krat_exp_expand.hxx (list_concat):
Handle ONE and ZERO in lists.
(MATCH__(Product, lhs, rhs)): Sum weights of identical element
generated while expansing the product.
* taf-kit/tests/vcsn-char-z.test: Add some test cases based on
a report by Jacques Sakarovitch.

* include/vaucanson/algebra/implementation/semiring/numerical_semiring.hh,
include/vaucanson/algebra/implementation/semiring/numerical_semiring.hxx
(op_parse): Remove this overloading for RationalNumber.  Its
incorrect usage of tellg() broke with g++-4.6, and the general
definition of op_parse() in
include/vaucanson/algebra/concept/semiring_base.hxx will call the
operator<< code that is specialized for RationalNumber already.
* include/vaucanson/algebra/implementation/semiring/q_number.hh,
include/vaucanson/algebra/implementation/semiring/q_number.hxx
(operator<<): Move into the vcsn::algebra namespace, so that Koening
lookup works.  Also to not try to read past the end of the stream.

* taf-kit/src/cmd_wt.cc (quotient_command): Do not call realtime.
* taf-kit/tests/vcsn-int-z.test, taf-kit/tests/vcsn-int-b.test: Adjust.
Do not call realtime in quotient.

* taf-kit/src/cmd_wt.cc (eval_command): Rename as ...
(eval_S_command): ... this
(eval_command): New version that does not call realtime.
* taf-kit/tests/vcsn-char-b.test,
taf-kit/tests/vcsn-char-fmp-b.test,
taf-kit/tests/vcsn-char-z.test, taf-kit/tests/vcsn-int-b.test,
taf-kit/tests/vcsn-int-z.test: Adjust.

Angrand <pangrand@gmail.com>.

* include/vaucanson/algorithms/infiltration.hh,
include/vaucanson/algorithms/infiltration.hxx,
include/vaucanson/algorithms/shuffle.hh,
include/vaucanson/algorithms/shuffle.hxx: New files.
* include/Makefile.am: Add them.
* lib/generate_libvcsn.rb (create?): Do not compile these
in FMP contexts.
* taf-kit/src/cmd_wt.cc (shuffle_command, infiltration_command):
New functions.

* taf-kit/src/cmd_fmp.cc (u_composition): Rename as ...
(composition_R): ... this.
(composition): New function similar to composition_R, but
without subnormalization.
* taf-kit/tests/vcsn-char-fmp-b.test: Adjust.

* include/vaucanson/algorithms/normalized_composition.hxx
(do_compose): Call eps_removal_here() after composing two
Boolean transducers.

* taf-kit/src/cmd_fmp.cc (composition, composition_R): Rename as ...
(b_composition, b_composition_R): ... these.  And compile
only for Boolean weights.
* taf-kit/tests/vcsn-char-fmp-b.test: Adjust.

* include/vaucanson/algorithms/characteristic.hxx,
include/vaucanson/algorithms/characteristic.hh: New files.
* include/Makefile.am: Add them.
* taf-kit/src/Makefile.am: Define BOOL_CONTEXT macros for
all instances.
* taf-kit/src/cmd_wt.cc (characteristic_command): New function.
* taf-kit/src/common.hh (BOOL_CONTEXT_HEADER): Define.
* taf-kit/src/getters.cc, taf-kit/src/getters.hh
(get_boolean_aut): Rename as ...
(get_single_band_aut): ... this because the automaton we
read with this function is not always Boolean.
(get_boolean_aut, locate_boolean_file): New functions.
* taf-kit/src/cmd_fmp.cc (evaluation_command): Adjust call.
* taf-kit/src/library_commands.hh (get_base_path): Define
base_path as static so that it is computed only once.
* taf-kit/src/pipe.hh, taf-kit/src/pipe_getters.hh,
taf-kit/src/pipe_getters.hxx, taf-kit/src/pipe_writers.hxx:
Implement I/O for BOOL_CONTEXT.

* include/vaucanson/algorithms/aut_to_exp.hh: Add missing
interface for the two-argument version of aut_to_exp when the
second argument is vcsn::DefaultChooser.  G++ 4.6 complained
today, but I'm surprised this issue hasn't come up before.

* taf-kit/src/cmd_gen.cc (command_chain): New function.

* taf-kit/src/cmd_wt.cc (prefix_command, suffix_command,
factor_command): Move these functions ...
* taf-kit/src/cmd_b.cc (prefix_command, suffix_command,
factor_command): ... here, on Jacques' request.

* lib/timer.cc: Use vertex_descriptor instead of vertex_iterator
to fix compilation with Boost 1.46.1.

* include/vaucanson/algebra/implementation/semiring/cyclic_semiring.hxx:
remove "unit" for "unsigned int".

* taf-kit/tests/vcsn-char-f2.test: Add tests for f2.

* taf-kit/src/cmd_r.cc: Fix bug in are_equivalent_command
for compatibility with the f2 semiring.
* include/vaucanson/algebra/implementation/semiring/cyclic_semiring.hxx:
Remove useless inlines.
* include/vaucanson/algebra/implementation/semiring/q_number.hh,
include/vaucanson/algebra/implementation/semiring/q_number.hxx:
Fix the >> operator.

* taf-kit/src/cmd_gen.cc: Include is_trim algorithm.

* include/vaucanson/automata/generic_contexts/f2_automaton.hh,
include/vaucanson/automata/generic_contexts/f2_automaton_structures.hh,
src/tests/test-suites/generate-all.sh,
taf-kit/src/Makefile.am,
taf-kit/tests/Makefile.am,
lib/Makefile.am,
lib/generate_libvcsn.rb,
include/vaucanson/contexts/f2_semiring.thh,
configure.ac: Change the context name.
* include/Makefile.am: Add files to distcheck

    algebra/implementation/semiring/cyclic_semiring.hh: Specialized
    the cyclic semiring implementation for Z/2Z (Z/nZ with n = 2)
    using booleans
  * include/vaucanson/contexts/z2z_semiring.thh: Modified the Z/2Z
    context accordingly.

    algorithm so it does not add transitions labeled by zero.
    This can happenens while working on a Z/2Z semiring.
  * xml/builders.hxx: Added a FIXME in the writer.