/*  Part of SWI-Prolog

    Author:        Jan Wielemaker
    E-mail:        J.Wielemaker@cs.vu.nl
    WWW:           http://www.swi-prolog.org
    Copyright (C): 1985-2014, University of Amsterdam
			      VU University Amsterdam

    This program is free software; you can redistribute it and/or
    modify it under the terms of the GNU General Public License
    as published by the Free Software Foundation; either version 2
    of the License, or (at your option) any later version.

    This program is distributed in the hope that it will be useful,
    but WITHOUT ANY WARRANTY; without even the implied warranty of
    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
    GNU General Public License for more details.

    You should have received a copy of the GNU Lesser General Public
    License along with this library; if not, write to the Free Software
    Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA

    As a special exception, if you link this library with other files,
    compiled with a Free Software compiler, to produce an executable, this
    library does not by itself cause the resulting executable to be covered
    by the GNU General Public License. This exception does not however
    invalidate any other reasons why the executable file might be covered by
    the GNU General Public License.
*/


:- module(prolog_statistics,
	  [ statistics/0,
	    statistics/1,		% -Stats
	    thread_statistics/2,	% ?Thread, -Stats
	    time/1,			% :Goal
	    profile/1,			% :Goal
	    profile/2,			% :Goal, +Options
	    show_profile/1		% +Options
	  ]).
:- use_module(library(lists)).
:- use_module(library(pairs)).
:- use_module(library(option)).

:- meta_predicate
	time(0),
	profile(0),
	profile(0, +).

/** <module> Get information about resource usage

This library provides predicates to   obtain  information about resource
usage by your program. The predicates of  this library are for human use
at the toplevel: information is _printed_.   All predicates obtain their
information using public low-level primitives.   These primitives can be
use to obtain selective statistics during execution.
*/

%%	statistics is det.
%
%	Print information about resource usage using print_message/2.
%
%	@see	All statistics printed are obtained through statistics/2.

statistics :-
	phrase(collect_stats, Stats),
	print_message(information, statistics(Stats)).

%%	statistics(-Stats:dict) is det.
%
%	Stats  is  a  dict   representing    the   same  information  as
%	statistics/0. This convience function is   primarily intended to
%	pass  statistical  information  to  e.g.,  a  web  client.  Time
%	critical code that wishes to   collect statistics typically only
%	need a small subset  and  should   use  statistics/2  to  obtain
%	exactly the data they need.

statistics(Stats) :-
	phrase(collect_stats, [CoreStats|StatList]),
	dict_pairs(CoreStats, _, CorePairs),
	map_list_to_pairs(dict_key, StatList, ExtraPairs),
	append(CorePairs, ExtraPairs, Pairs),
	dict_pairs(Stats, statistics, Pairs).

dict_key(Dict, Key) :-
	gc{type:atom} :< Dict, !,
	Key = agc.
dict_key(Dict, Key) :-
	is_dict(Dict, Key).

collect_stats -->
	core_statistics,
	gc_statistics,
	agc_statistics,
	shift_statistics,
	thread_counts.

core_statistics -->
	{ statistics(process_cputime, Cputime),
	  statistics(process_epoch, Epoch),
	  statistics(inferences, Inferences),
	  statistics(atoms, Atoms),
	  statistics(functors, Functors),
	  statistics(predicates, Predicates),
	  statistics(modules, Modules),
	  statistics(codes, Codes),
	  thread_self(Me),
	  thread_stack_statistics(Me, Stacks)
	},
	[ core{ time:time{cpu:Cputime, inferences:Inferences, epoch:Epoch},
		data:counts{atoms:Atoms, functors:Functors,
			    predicates:Predicates, modules:Modules,
			    vm_codes:Codes},
		stacks:Stacks
	      }
	].

thread_stack_statistics(Thread,
		  stacks{local:stack{name:local,
				    limit:LocalLimit,
				     allocated:Local,
				     usage:LocalUsed},
			 global:stack{name:global,
				      limit:GlobalLimit,
				      allocated:Global,
				      usage:GlobalUsed},
			 trail:stack{name:trail,
				     limit:TrailLimit,
				     allocated:Trail,
				     usage:TrailUsed},
			 total:stack{name:stacks,
				     limit:StackLimit,
				     allocated:StackAllocated,
				     usage:StackUsed}
			}) :-
	thread_statistics(Thread, trail,       Trail),
	thread_statistics(Thread, trailused,   TrailUsed),
	thread_statistics(Thread, local,       Local),
	thread_statistics(Thread, localused,   LocalUsed),
	thread_statistics(Thread, global,      Global),
	thread_statistics(Thread, globalused,  GlobalUsed),
	thread_statistics(Thread, locallimit,  LocalLimit),
	thread_statistics(Thread, globallimit, GlobalLimit),
	thread_statistics(Thread, traillimit,  TrailLimit),
	StackUsed is LocalUsed+GlobalUsed+TrailUsed,
	StackAllocated is Local+Global+Trail,
	StackLimit is LocalLimit+GlobalLimit+TrailLimit.

gc_statistics -->
	{ statistics(collections, Collections),
	  Collections > 0, !,
	  statistics(collected, Collected),
	  statistics(gctime, GcTime)
	},
	[ gc{type:stack, unit:byte,
	     count:Collections, time:GcTime, gained:Collected } ].
gc_statistics --> [].

agc_statistics -->
	{ catch(statistics(agc, Agc), _, fail),
	  Agc > 0, !,
	  statistics(agc_gained, Gained),
	  statistics(agc_time, Time)
	},
	[ gc{type:atom, unit:atom,
	     count:Agc, time:Time, gained:Gained} ].
agc_statistics --> [].

shift_statistics -->
	{ statistics(local_shifts, LS),
	  statistics(global_shifts, GS),
	  statistics(trail_shifts, TS),
	  (   LS > 0
	  ;   GS > 0
	  ;   TS > 0
	  ), !,
	  statistics(shift_time, Time)
	},
	[ shift{local:LS, global:GS, trail:TS, time:Time} ].
shift_statistics --> [].

thread_counts -->
	{ current_prolog_flag(threads, true), !,
	  statistics(threads, Active),
	  statistics(threads_created, Created),
	  statistics(thread_cputime, CpuTime),
	  Finished is Created - Active
	},
	[ thread{count:Active, finished:Finished, time:CpuTime} ].
thread_counts --> [].

%%	thread_statistics(?Thread, -Stats:dict) is nondet.
%
%	Obtain statistical information about a single thread.  Fails
%	silently of the Thread is no longer alive.
%
%	@arg	Stats is a dict containing status, time and stack-size
%		information about Thread.

thread_statistics(Thread, Stats) :-
	thread_property(Thread, status(Status)),
	(   catch(thread_stats(Thread, Stacks, Time), _, fail)
	->  Stats = thread{id:Thread,
			   status:Status,
			   time:Time,
			   stacks:Stacks}
	;   Stats = thread{id:Thread,
			   status:Status}
	).

thread_stats(Thread, Stacks,
	     time{cpu:CpuTime,
		  inferences:Inferences,
		  epoch:Epoch
		 }) :-
	thread_statistics(Thread, cputime, CpuTime),
	thread_statistics(Thread, inferences, Inferences),
	thread_statistics(Thread, epoch, Epoch),
	thread_stack_statistics(Thread, Stacks).


%%	time(:Goal) is nondet.
%
%	Execute Goal, reporting statistics to the user. If Goal succeeds
%	non-deterministically,  retrying  reports  the   statistics  for
%	providing the next answer.
%
%	Statistics  are  retrieved  using   thread_statistics/3  on  the
%	calling   thread.   Note   that   not    all   systems   support
%	thread-specific CPU time. Notable, this is lacking on MacOS X.
%
%	@bug Inference statistics are often a few off.
%	@see statistics/2 for obtaining statistics in your program and
%	     understanding the reported values.

time(Goal) :-
	time_state(State0),
	(   call_cleanup(catch(Goal, E, (report(State0,10), throw(E))),
			 Det = true),
	    time_true(State0),
	    (	Det == true
	    ->	!
	    ;	true
	    )
	;   report(State0, 11),
	    fail
	).

report(t(OldWall, OldTime, OldInferences), Sub) :-
	time_state(t(NewWall, NewTime, NewInferences)),
	UsedTime is NewTime - OldTime,
	UsedInf  is NewInferences - OldInferences - Sub,
	Wall     is NewWall - OldWall,
	(   UsedTime =:= 0
	->  Lips = 'Infinite'
	;   Lips is integer(UsedInf / UsedTime)
	),
	print_message(information, time(UsedInf, UsedTime, Wall, Lips)).

time_state(t(Wall, Time, Inferences)) :-
	get_time(Wall),
	statistics(cputime, Time),
	statistics(inferences, Inferences).

time_true(State0) :-
	report(State0, 12).		% leave choice-point
time_true(State) :-
	get_time(Wall),
	statistics(cputime, Time),
	statistics(inferences, Inferences0),
	plus(Inferences0, -3, Inferences),
	nb_setarg(1, State, Wall),
	nb_setarg(2, State, Time),
	nb_setarg(3, State, Inferences),
	fail.


		 /*******************************
		 *     EXECUTION PROFILING	*
		 *******************************/

/* - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
This module provides a simple backward compatibility frontend on the new
(in version 5.1.10) execution profiler  with  a   hook  to  the  new GUI
visualiser for profiling results defined in library('swi/pce_profile').

Later we will add a proper textual report-generator.
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - */

:- multifile
	prolog:show_profile_hook/1.

%%	profile(:Goal).
%%	profile(:Goal, +Options).
%
%	Run Goal under the execution profiler.  Defined options are:
%
%	  * time(Which)
%	  Profile =cpu= or =wall= time.  The default is CPU time.
%	  * top(N)
%	  When generating a textual report, show the top N predicates.
%	  * cummulative(Bool)
%	  If =true= (default =false=), show cummulative output in
%	  a textual report.

profile(Goal) :-
	profile(Goal, []).

profile(Goal0, Options) :-
	option(time(Which), Options, cpu),
	time_name(Which, How),
	expand_goal(Goal0, Goal),
	call_cleanup('$profile'(Goal, How),
		     prolog_statistics:show_profile(Options)).

time_name(cpu,	    cputime)  :- !.
time_name(wall,	    walltime) :- !.
time_name(cputime,  cputime)  :- !.
time_name(walltime, walltime) :- !.
time_name(Time, _) :-
	must_be(oneof([cpu,wall]), Time).

%%	show_profile(+Options)
%
%	Display last collected profiling data.  Options are
%
%	  * top(N)
%	  When generating a textual report, show the top N predicates.
%	  * cummulative(Bool)
%	  If =true= (default =false=), show cummulative output in
%	  a textual report.

show_profile(N) :-
	integer(N), !,
	show_profile([top(N)]).
show_profile(Options) :-
	profiler(Old, false),
	show_profile_(Options),
	profiler(_, Old).

show_profile_(Options) :-
	prolog:show_profile_hook(Options), !.
show_profile_(Options) :-
	prof_statistics(Stat),
	prof_statistics(time, Stat, Time),
	sort_on(Options, SortKey),
	findall(KeyedNode, prof_node(SortKey, KeyedNode), Nodes),
	keysort(Nodes, Sorted),
	reverse(Sorted, HighFirst),
	format('~61t~69|~n'),
	format('Total time: ~2f seconds~n', [Time]),
	format('~61t~69|~n'),
	format('~w~t~w =~45|~t~w~60|~t~w~69|~n',
	       [ 'Predicate', 'Box Entries', 'Calls+Redos', 'Time'
	       ]),
	format('~61t~69|~n'),
	option(top(N), Options, 25),
	show_plain(HighFirst, N, Stat, SortKey).

sort_on(Options, ticks_self) :-
	option(cummulative(false), Options, false), !.
sort_on(_, ticks).

show_plain([], _, _, _).
show_plain(_, 0, _, _) :- !.
show_plain([_-H|T], N, Stat, Key) :-
	show_plain(H, Stat, Key),
	N2 is N - 1,
	show_plain(T, N2, Stat, Key).

show_plain(Node, Stat, Key) :-
	value(label,			   Node, Pred),
	value(call,			   Node, Call),
	value(redo,			   Node, Redo),
	value(time(Key, percentage, Stat), Node, Percent),
	IntPercent is round(Percent*10),
	Entry is Call + Redo,
	format('~w~t~D =~45|~t~D+~55|~D ~t~1d%~69|~n',
	       [Pred, Entry, Call, Redo, IntPercent]).


		 /*******************************
		 *	   DATA GATHERING	*
		 *******************************/

%%	prof_statistics(-Node) is det.
%
%	Get overall statistics
%
%	@param Node	term of the format prof(Ticks, Account, Time, Nodes)

prof_statistics(prof(Samples, Ticks, Account, Time, Nodes)) :-
	'$prof_statistics'(Samples, Ticks, Account, Time, Nodes).

prof_statistics(samples, Term, Samples) :-
	arg(1, Term, Samples).
prof_statistics(ticks, Term, Ticks) :-
	arg(2, Term, Ticks).
prof_statistics(accounting, Term, Ticks) :-
	arg(3, Term, Ticks).
prof_statistics(time, Term, Ticks) :-
	arg(4, Term, Ticks).
prof_statistics(nodes, Term, Ticks) :-
	arg(5, Term, Ticks).


%%	prof_node(+Field, -Pairs) is nondet.
%
%	Collect data for each of the interesting predicate.
%
%	@param Field specifies the field to use as key in each pair.
%	@param Pair is a term of the following format:
%
%	  ==
%	  KeyValue-node(Pred,
%		        TimeSelf, TimeSiblings,
%			Calls, Redo, Recursive,
%		        Parents)
%	  ==
%

prof_node(KeyOn, Node) :-
	setup_call_cleanup(
	    ( current_prolog_flag(access_level, Old),
	      set_prolog_flag(access_level, system)
	    ),
	    get_prof_node(KeyOn, Node),
	    set_prolog_flag(access_level, Old)).

get_prof_node(KeyOn, Key-Node) :-
	Node = node(M:H,
		    TicksSelf, TicksSiblings,
		    Call, Redo,
		    Parents, Siblings),
	current_predicate(_, M:H),
	\+ predicate_property(M:H, imported_from(_)),
	'$prof_procedure_data'(M:H,
			       TicksSelf, TicksSiblings,
			       Call, Redo,
			       Parents, Siblings),
	value(KeyOn, Node, Key).

key(predicate,	    1).
key(ticks_self,	    2).
key(ticks_siblings, 3).
key(call,	    4).
key(redo,	    5).
key(callers,	    6).
key(callees,	    7).

value(name, Data, Name) :- !,
	arg(1, Data, Pred),
	predicate_functor_name(Pred, Name).
value(label, Data, Label) :- !,
	arg(1, Data, Pred),
	predicate_label(Pred, Label).
value(ticks, Data, Ticks) :- !,
	arg(2, Data, Self),
	arg(3, Data, Siblings),
	Ticks is Self + Siblings.
value(time(Key, percentage, Stat), Data, Percent) :- !,
	value(Key, Data, Ticks),
	prof_statistics(ticks, Stat, Total),
	prof_statistics(accounting, Stat, Account),
	(   Total-Account > 0
	->  Percent is 100 * (Ticks/(Total-Account))
	;   Percent is 0.0
	).
value(Name, Data, Value) :-
	key(Name, Arg),
	arg(Arg, Data, Value).

%%	predicate_label(+Head, -Label)
%
%	Create a human-readable label for the given head

predicate_label(M:H, Label) :- !,
	functor(H, Name, Arity),
	(   hidden_module(M, H)
	->  atomic_list_concat([Name, /, Arity], Label)
	;   atomic_list_concat([M, :, Name, /, Arity], Label)
	).
predicate_label(H, Label) :- !,
	functor(H, Name, Arity),
	atomic_list_concat([Name, /, Arity], Label).

hidden_module(system, _).
hidden_module(user, _).
hidden_module(M, H) :-
	predicate_property(system:H, imported_from(M)).

%%	predicate_functor_name(+Head, -Name)
%
%	Return the (module-free) name of the predicate for sorting
%	purposes.

predicate_functor_name(_:H, Name) :- !,
	predicate_functor_name(H, Name).
predicate_functor_name(H, Name) :-
	functor(H, Name, _Arity).


		 /*******************************
		 *	      MESSAGES		*
		 *******************************/

:- multifile
	prolog:message/3.

prolog:message(time(UsedInf, UsedTime, Wall, Lips)) -->
	[ '~D inferences, ~3f CPU in ~3f seconds (~w% CPU, ~w Lips)'-
	  [UsedInf, UsedTime, Wall, Perc, Lips] ],
	{   Wall > 0
	->  Perc is round(100*UsedTime/Wall)
	;   Perc = ?
	}.
prolog:message(statistics(List)) -->
	msg_statistics(List).

msg_statistics([]) --> [].
msg_statistics([H|T]) -->
	{ is_dict(H, Tag) },
	msg_statistics(Tag, H),
	(   { T == [] }
	->  []
	;   [nl], msg_statistics(T)
	).

msg_statistics(core, S) -->
	time_stats(S.time), [nl],
	data_stats(S.data), [nl,nl],
	stacks_stats(S.stacks).
msg_statistics(gc, S) -->
	{ S.type == stack -> Label = '' ; Label = 'atom ' },
	[ '~D ~wgarbage collections gained ~D ~ws in ~3f seconds.'-
	  [ S.count, Label, S.gained, S.unit, S.time]
	].
msg_statistics(shift, S) -->
	[ 'Stack shifts: ~D local, ~D global, ~D trail in ~3f seconds'-
	  [ S.local, S.global, S.trail, S.time ]
	].
msg_statistics(thread, S) -->
	[ '~D threads, ~D finished threads used ~3f seconds'-
	  [S.count, S.finished, S.time]
	].

time_stats(T) -->
	{ format_time(string(Epoch), '%+', T.epoch) },
	[ 'Started at ~s'-[Epoch], nl,
	  '~3f seconds cpu time for ~D inferences'-
	  [ T.cpu, T.inferences ]
	].
data_stats(C) -->
	[ '~D atoms, ~D functors, ~D predicates, ~D modules, ~D VM-codes'-
	  [ C.atoms, C.functors, C.predicates, C.modules, C.vm_codes]
	].
stacks_stats(S) -->
	[ '~|~tLimit~28+~tAllocated~13+~tIn use~13+'-[], nl ],
	stack_stats('Local ', S.local),  [nl],
	stack_stats('Global', S.global), [nl],
	stack_stats('Trail ', S.trail),  [nl].

stack_stats(Stack, S) -->
	[ '~|~w stack:~t~D~28+ ~t~D~13+ ~t~D~13+ Bytes'-
	  [Stack, S.limit, S.allocated, S.usage]
	].

:- multifile sandbox:safe_primitive/1.

sandbox:safe_primitive(prolog_statistics:statistics(_)).
sandbox:safe_primitive(prolog_statistics:statistics).