Acme-FSM
view release on metacpan or search on metacpan
=item I<$namespace>
B<A::F> uses elaborate schema to reach various callbacks
(three of them, at the time of writing).
This optional parameter is in use by this schema.
L<B<query()> method|/query()> has more.
=item I<$rule>
A scalar identifing a I<[turn]>.
One of opaque scalars returned by B<switch()> callback
(the other is processed (modified or not) I<$item>).
L<B<process()> method|/process()> description has more.
=item B<source()>
Special piece of user-code
(it's required at construction (L<B<connect()> method|/connect()>),
L<B<query_source()> method|/query_source()> describes how FSM reaches it).
Whatever it returns (in explicit scalar context) becomes I<$item>.
=item I<$state>
A scalar identifing a I<{state}> (see below) or parameter of I<{fsm}> (see
above).
=item state flow
Just like control flow but for I<$state>s.
=item state record
=item I<{state}>
One record in I<{fst}>.
The record consists of entries (see above).
L<B<process()>|/process()> method description has more.
Should be noted, in most cases "I<{state}>" should be read as
"I<$state> I<{state}>" instead,
but that starts to smell bufallo.
=item B<switch()>
A mandatory callback associated with every I<{state}>.
L<B<process()> method|/process()> and
L<B<query_switch()> method|/query_switch()>
descriptions have more.
=item I<$turn>
No such thing.
It's I<$rule> instead (see above).
=item I<[turn]>
Specially crafted entry in I<{state}>
(covered in details in L<B<process()> method|/process()> description).
Such entry describes what next I<$state> should be picked in state flow
and what to do with I<$item>.
=item turn map
This idiom is used in place of "C<turns> I<$rule> of I<[turn]>".
=back
=cut
=head1 B<connect()>
$bb1 = Acme::FSM->connect( { %options1 }, %fst1 );
$bb2 = Acme::FSM->connect( { %options2 }, { %fst2 } );
$bb3 = $bb2->connect( { %options3 } );
Creates a blackboard object.
Blackboard isa HASH, it's free to use except special C<_> key;
that key is for I<{fsm}> exclusively.
First parameter isa I<%$options>, it's required
(pass empty HASH if nothing to say).
Defined keys are:
=over
=item I<diag_level>
(positive integer)
Sets a diagnostic threshold.
It's meaning is covered in L<B<diag()> method|/diag()> documentation.
If C<undef> then set to C<1> (C<0> is B<defined>).
=item I<dumper>
(scalar or C<CODE>)
B<A::F> operates on arbitrary items and there's a diagnostic service that
sometimes insists on somehow showing those arbitrary items.
It's up to user's code to process that arbitrary data and yield some scalar
represantation.
Refer to L<B<query_dumper()> method|/query_dumper()> documentation for
details.
Optional.
Simple stringifier is provided by L<B<query_dumper()> method|/query_dumper()>
itself.
=item I<namespace>
(scalar or object(!) B<ref>)
Sets a context for various parts of I<{fsm}> and services would be resolved.
No defaults.
Refer to L<B<query()> method|/query()> documentation for details.
=item I<source>
(scalar or C<CODE>)
Sets a source of items to process to be queried.
Required.
Refer to L<B<query_source()> method|/query_source()> documentation for
details.
=back
Second is FST (Finite State Table).
return undef }
elsif( 2 == @_ && ref $_[1] eq q|HASH| ) {
my $backup = $self->fst( $_[0] );
$self->diag( 3, q|%s {%s} record|,
( $backup ? q|updating| : q|creating| ), $_[0] );
# XXX:202202150056:whynot: Copy is a side-effect instead.
$self->{_}{fst}{shift @_} = {%{ pop @_ }};
return $backup }
elsif( !exists $self->{_}{fst}{$_[0]} ) {
$self->carp( qq|($_[0]): no such {fst} record| );
return undef }
elsif( 1 == @_ ) {
return $self->{_}{fst}{shift @_} }
elsif( 2 == @_ ) {
return $self->{_}{fst}{shift @_}{shift @_} }
elsif( 3 == @_ ) {
my $backup = $self->fst( $_[0] => $_[1] );
$self->diag( 3, q|%s {%s}{%s} entry|,
( $backup ? q|updating| : q|creating| ), @_[0,1] );
$self->{_}{fst}{shift @_}{shift @_} = pop @_;
return $backup }
else {
$self->carp( sprintf q|too many args (%i)|, scalar @_ );
return undef }}
=item B<turn()>
$bb->turn( $state ) eq '' or die;
$bb->turn( $state => 'uturn' )->[1] eq 'NEXT' or die;
Queries I<[turn]>s of arbitrary I<{state}>s.
B<turn()> doesn't manipulate entries, use L<B<fst()> method|/fst()> instead
L<if you can|/turn() and fst()>.
Modes:
=over
=item query expected behaviour
This mode is entered if there is lone scalar.
Such scalar is believed to be I<$state>.
Returns something that describes what kind of least special states are
present.
Namely:
=over
=item *
C<undef> is returned if I<$state> isn't present in the I<{fst}>
(also B<carp>s).
Also see below.
=item *
Empty string is returned if there're I<tturn> and/or I<fturn> turns.
I<turns> hash is ignored in that case.
=item *
C<HASH> is returned if there's turn map
(and neither I<tturn> nor I<fturn> is present).
B<(note)> In that case, B<turn()> checks for I<turns> is indeed a HASH,
nothing more
(however B<croaks> if that's not the case);
It may as well be empty;
Design legacy.
=item *
Returns C<HASH> for C<STOP> and C<BREAK> I<$state>s without any further
processing
(For those I<$state>s any I<$rule> is ignored and C<HASH> enables I<switch()>
callbacks to give more informative logs
(while that information is mangled anyway);
Probably bad idea).
=item *
C<undef> is returned if there's nothing to say --
neither I<tturn>, nor I<fturn>, nor turn map --
this record is kind of void.
The record should be studied to find out why.
B<carp>s in that case.
=back
=item query specific I<[turn]>
Two scalars are I<$state> and specially encoded I<$rule>
(refer to L<B<query_switch()> method|/query_switch()> about encoding).
If I<$rule> can't be decoded then B<croak>s.
Returns (after verification) requested I<$rule> as ARRAY.
While straightforward I<[turn]>s (such as C<tturn>, C<fturn>, and such) could
be in fact queried through L<B<fst()> method|/fst()> turn map needs bit more
sophisticated handling;
and that's what B<turn()> does;
in fact asking for C<turns> will result in B<croak>.
I<$action> of C<START> and C<CONTINUE> special states suffer implicit
defaulting to empty string.
=item anything else
No arguments or more then two is an non-fatal error.
Returns C<undef> (with B<carp>).
=back
=cut
# TODO:202202172011:whynot: As soon as supported perl is young enough change it to smartmatch, plz.
my %special_turns = map { $_ => 1 } qw| eturn uturn tturn fturn |;
# TODO:202202162030:whynot: Consider more elaborate (informative) returns.
sub turn {
my $self = shift @_;
unless( @_ ) {
$self->carp( q|no args| ); return undef }
elsif( 1 == @_ && !exists $self->{_}{fst}{$_[0]} ) {
$self->carp( qq|($_[0]): no such {fst} record| );
return undef }
elsif( 1 == @_ ) {
my $state = shift @_;
my $entry = $self->verify(
$self->{_}{fst}{$state}, $state, '', q|entry|, q|HASH| );
# WORKAROUND:201305070051:whynot: Otherwise there will be spurious B<carp>s about anyway useless turns in those entries.
$state eq q|STOP| || $state eq q|BREAK| and return q|HASH|;
exists $entry->{tturn} || exists $entry->{fturn} and return '';
unless( exists $entry->{turns} ) {
# XXX:201305071531:whynot: Should just B<croak> instead, probably.
$self->carp( qq|{$state}: none supported turn| );
return undef }
$self->verify( $entry->{turns}, $state, q|turns|, q|turn|, q|HASH| );
return q|HASH| }
elsif( 2 == @_ ) {
my( $state, $turn ) = @_;
my $entry;
$self->verify( $turn, $state, $turn, q|turn|, '' );
if( exists $special_turns{$turn} ) {
$entry = $self->{_}{fst}{$state}{$turn} }
elsif( !index $turn, q|turn%| ) {
$entry = $self->{_}{fst}{$state}{turns}{substr $turn, 5} }
else {
croak sprintf q|[turn]: {%s}(%s): unknown turn|, $state, $turn }
$self->verify( $entry, $state, $turn, q|turn|, q|ARRAY| );
$self->verify( $entry->[0], $state, $turn, q|state|, '' );
# XXX:20121230140241:whynot: {START}{turn}{action} is ignored anyway.
# XXX:201305072006:whynot: {CONTINUE}{turn}{action} is ignored too.
$entry->[1] //= '' if $state eq q|START| || $state eq q|CONTINUE|;
$self->verify( $entry->[1], $state, $turn, q|action|, '' );
return $entry }
else {
$self->carp( sprintf q|too many args (%i)|, scalar @_ );
return undef }
}
=item B<action()>
$bb->action eq $action and die;
$action = $bb->action( $new_action );
Queries and sets I<$action> of B<A::F> instance.
Modes:
=over
=item query I<$action>
No arguments -- returns current I<$action> of the instance.
Note, Perl FALSE B<isa> parameter.
=item set I<$action>
( $rule, $item ) = $self->query_switch( $item );
Internal multitool.
That's the point where decisions about turns are made.
B<(note)>
B<query_switch()> converts I<$rule> (as returned by B<switch()>) to specially
encoded scalar;
it's caller's responcibility pick correct I<[turn]> later.
Strategy:
=over
=item no arguments
Special-state mode:
invoke B<switch()> with no arguments;
ignore whatever I<$item> has been possibly returned;
return I<$rule> alone.
=item I<$item> is C<undef>
EOF mode:
ignore B<switch()> completely;
return C<eturn> and C<undef>.
=item I<$item> is not C<undef>
King-size mode:
invoke B<switch()>, pass I<$item> as single argument.
return I<$rule> and I<$item>
(whatever it became after going through B<switch()>).
=back
I<$rule>, as it was returned by B<switch()>, is encoded like this:
=over
=item I<$rule> is C<undef>
Return C<uturn>.
B<(note)>
Don't verify if C<uturn> I<[turn]> exists.
=item I<$rule> is Perl TRUE and C<tturn> and/or C<fturn> are present
Return C<tturn>
B<(note)>
Don't verify if C<tturn> I<[turn]> exists.
=item I<$rule> is Perl FALSE and C<tturn> and/or C<fturn> are present
Return C<fturn>
B<(note)>
Don't verify if C<fturn> I<[turn]> exists.
=item neither C<tturn> or C<fturn> are present
Encode I<$rule> like this C<'turn%' . $rule> and return that.
B((note)>
Don't verify if turn map exists.
B<(note)>
Don't verify if C<"turn%$rule"> exists in turn map.
=back
B<switch()> is always invoked in list context even if I<$item> would be
ignored.
If I<$rule> shouldn't be paired with I<$item> it won't be
(it's safe to call B<query_switch()> in scalar context then and
there won't be any trailing C<undef>s).
=cut
sub query_switch {
my $self = shift @_;
my @turn;
# WORKAROUND:20121229000801:whynot: No B<verify()>, B<query()> does its checks by itself.
@turn = $self->query(
$self->fst( $self->state, q|switch| ),
sprintf( q|{%s}{switch}|, $self->state ),
@_ ) if !@_ || defined $_[0];
my $kind = $self->turn( $self->state );
$turn[0] =
@_ && !defined $_[0] ? q|eturn| :
# TODO:202201071700:whynot: Make C<undef> special only when C<uturn> is present, plz.
!defined $turn[0] ? q|uturn| :
# FIXME:201304230145:whynot: Defaulting to basics here looks as bad as B<croak>ing.
# TODO:202212202039:whynot: L<Default For Turn Map>.
$kind ? qq|turn%$turn[0]| :
$turn[0] ? q|tturn| : q|fturn|;
return @_ ? @turn : $turn[0] }
=item B<query_source()>
( $item, $dump ) = $self->query_source;
Seeks B<source()> callback and acquires whatever it returns.
The callback is called in scalar context.
As useful feature, also feeds I<$item> to L<dumper callback|/query_dumper()>.
L<B<query()> method|/query()> has detailed description how B<source()>
callback is acquired.
Returns I<$item> and result of L<I<dumper> callback|/dumper>.
=cut
sub query_source {
my $self = shift @_;
# WORKAROUND:20121229001530:whynot: No B<verify()>, I<{source}> can return anything.
my $item = $self->query( $self->{_}{source}, q|{source}|, @_ );
return $item, $self->query_dumper( $item ) }
=item B<query_dumper()>
$dump = $self->query_dumper( $item );
Seeks I<dumper> callback (L<configured at construction time|/dumper>).
If the callback wasn't configured uses simple hopefully informative and
C<undef> proof substitution.
Whatever the callback returns is checked to be B<defined>
(C<undef> is changed to C<"(unclear)">)
and then returned.
$self->query(
# TODO:202202210258:whynot: This is inefficient, defaulting should happen in B<connect()> instead.
$self->{_}{dumper} // sub { sprintf q|(%s)|, $_[1] // q|undef| },
q|{dumper}|, @_ ) // q|(unclear)|,
# XXX:202202210304:whynot: 'source' looks like remnants of refactoring. Should investigate it deeper.
$self->state, qw| source source |, '' ) }
=item B<diag()>
$bb->diag( 3, 'going to die at %i.', __LINE__ );
Internal.
Provides unified and single-point-of-failure way to output diagnostics.
Intensity is under control of
L<I<diag_level> configuration parameter|/diag_level>.
Each object has it's own,
however it's inherited when objects are copied.
Defined levels are:
=over
=item C<0>
Nothing at all.
Even error reporting is suppressed.
=item C<1>
Default.
Errors of here-be-dragons type.
=item C<2>
Basic diagnostics for callbacks.
=item C<3>
Basic trace.
Construction, starting and leaving runs.
=item C<4>
Extended diagnostics for callbacks.
=item C<5>
Deep trace.
By the way diagnostics of I<switch> entry resolving.
=back
=cut
sub diag {
my $self = shift @_;
$self->{_}{diag_level} >= shift @_ or return $self;
# TODO:202212222141:whynot: Since something this B<sprintf> might emit warnings. And maybe it's appropriate.
printf STDERR sprintf( qq|[%s]: %s\n|,
( split m{::}, ( caller 1 )[3])[-1], shift @_ ),
map $_ // q|(undef)|, @_;
return $self }
=item B<carp()>
$bb->carp( 'something wrong...' );
Internal.
B<carp>s consistently if I<{_}{diag_level}> is B<gt> C<0>.
=back
=cut
sub carp {
my $self = shift @_;
$self->{_}{diag_level} >= 1 or return;
unshift @_, sprintf q|[%s]: |, ( split m{::}, ( caller 1 )[3])[-1];
&Carp::carp }
=head1 BUGS AND CAVEATS
=over
=item Default For Turn Map
B<(missing feature)>
It's not hard to imagine application of rather limited turn map that should
default on anything else deemed irrelevant.
Right now to achieve logic like this such defaulting ought to be merged into
B<switch()>.
That's insane.
=item Diagnostics
B<(misdesign)>
Mechanics behind diagnostics isa failure.
It's messy, fragile, misguided, and (honestly) premature.
At the moment it's useless.
=item Hash vs HASH vs Acme::FSM Ref Constructors
B<(messy, probably bug)>
L<B<connect()> method description|/connect()> _declares_ that list is copied.
Of course, it's not true deep copy
({fst} might contain CODE, it's not clear how to copy it).
It's possible, one day list trailer variant of initialization may be put to
sleep.
See also L<linter below|/Linter>.
=item Linter
B<(missing feature)>
It might be hard to imagine,
but FST might get out of hands
(ie check F<t/process/quadratic.t>).
Indeed, some kind of (limited!) linter would be much desired.
It's missing.
=item Perl FALSE, C<undef>, and C<uturn>
B<(caveat)>
Back then B<DMA::FSM> treated C<undef> as any other Perl FALSE.
C<uturn> I<$rule> mech has made C<undef> special
(if B<switch()> returns C<undef> and C<uturn> I<{turn}> isn't present then
it's a B<croak>).
Thus, at the moment, C<undef> isn't FALSE (for B<A::F>).
This is counter-intuitive, actually.
=item Returning C<undef> for misuse
B<(bug)>
Why B<A::F> screws with caller,
in case of API violations (by returning C<undef>),
is beyond my understanding now.
=item B<source()> and I<$state>
B<(bug)> (also see L<B<switch()> and I<$item>|/switch() and $item>)
By design and legacy,
there is single point of input -- B<source()>.
IRL, multiple sources are normal.
Implementation wise that leads to B<source()> that on the fly figures out
current I<$state> and then somehow branches (sometimes branches out).
Such B<source()>s look horrible because they are.
=item Special handling of C<START> and C<CONTINUE>
B<(bug)>
In contrary with C<STOP> and C<BREAK> special states,
special treatement of C<START> and C<CONTINUE> is a side effect of
L<B<process()> method|/process()> entry sequence.
That's why routinely enter C<START> or C<CONTINUE> by state flow is of
there-be-dragons type.
Totally should be rewritten.
=item B<switch()> and I<$item>
B<(bug)>
It might be surprising, but, from experience,
the state flow mostly doesn't care about I<$item>.
Mostly, I<{state}> entered with I<$action> C<NEXT> processes input
and just wonders ahead.
Still those pesky I<$item>s *must* be passed around (by design).
Still every B<switch()> *must* return I<$item> too
(even if it's just a placeholder).
Something must be done about it.
=item C<tturn>, C<fturn>, and B<switch()>
B<(misdesign)>
First, by design and legacy B<switch()> must report what turn the control flow
must make
(that's why it *must* return I<$rule>).
OTOH, there's some value in keeping B<switch()>es as simple as possible what
results in I<{state}> with alone C<tturn> I<{turn}> and B<switch()> that
always returns TRUE.
Changes are coming.
=item B<turn()> and B<fst()>
B<(misdesign)>
Encoding (so to speak) in use by B<turn()> (in prediction mode) is plain
stupid.
C<undef> signals two distinct conditions
(granted, both are manifest of broken I<{fst}>).
Empty string doesn't distinguish safe (both C<tturn> and C<fturn> are present)
and risky (C<tturn> or C<fturn> is missing) I<{state}>.
C<HASH> doesn't say if there's anything in turn map.
All that needs loads of workout.
=back
=cut
=head1 DIAGNOSTICS
=over
=item C<[action]: changing action: (%s) (%s)>
B<(deep trace)>, L<B<action()> method|/action()>.
Exposes change of I<$action> from previous (former I<%s>)
to current (latter I<%s>).
=item C<[action]: too many args (%i)>
B<(warning)>, L<B<action()> method|/action()>.
Obvious.
None or one argument is supposed.
=item C<[connect]: (%s): unknow option, ignored>
B<(warning)>, L<B<connect()> method|/connect()>.
Each option that's not known to B<A::F> is ignored and B<carp>ed.
There're no means for child class to force B<A::F> (as a parent class) to
accept something.
Child classes should process their options by itself.
=item C<[connect]: clean init with (%i) items in FST>
B<(basic trace)>, L<B<connect()> method|/connect()>.
During class construction FST has been found.
That FST consists of I<%i> items.
Those items are number of I<$state>s and not I<$state>s plus I<{state}>s.
=item C<[connect]: FST has no {START} state>
B<(warning)>, L<B<connect()> method|/connect()>.
I<START> I<{state}> is required.
It's missing.
This situation will definetely result in devastating failure later.
=item C<[connect]: FST has no {STOP} state>
B<(warning)>, L<B<connect()> method|/connect()>.
I<STOP> I<{state}> is required.
It's missing.
If FST is such that I<STOP> I<$state> can't be reached
(for whatever reasons)
this may be cosmetic.
=item C<[connect]: ignoring (%i) FST items in trailer>
B<(warning)>, L<B<connect()> method|/connect()>.
A trailer has been found and was ignored.
Beware, I<%i> could be misleading.
For HASH in trailer that HASH is considered and key number is reported.
For list in trailer an B<item> number in list is reported, even if it's twice
( run in 0.534 second using v1.01-cache-2.11-cpan-39bf76dae61 )