Guard/Guard.pm

=head1 NAME

Guard - safe cleanup blocks

=head1 SYNOPSIS

 use Guard;

 # temporarily chdir to "/etc" directory, but make sure
 # to go back to "/" no matter how myfun exits:
 sub dosomething {
    scope_guard { chdir "/" };
    chdir "/etc";

    call_function_that_might_die_or_other_fun_stuff;
 }

=head1 DESCRIPTION

This module implements so-called "guards". A guard is something (usually
an object) that "guards" a resource, ensuring that it is cleaned up when
expected.

Specifically, this module supports two different types of guards: guard
objects, which execute a given code block when destroyed, and scoped
guards, which are tied to the scope exit.

=head1 FUNCTIONS

This module currently exports the C<scope_guard> and C<guard> functions by
default.

=over 4

=cut

package Guard;

BEGIN {
   $VERSION = '0.1';
   @ISA = qw(Exporter);
   @EXPORT = qw(guard scope_guard);

   require Exporter;

   require XSLoader;
   XSLoader::load Guard, $VERSION;
}

our $DIED = sub { warn "$@" };

=item scope_guard BLOCK

Registers a block that is executed when the current scope (block,
function, method, eval etc.) is exited.

The description below sounds a bit complicated, but that's just because
C<scope_guard> tries to get even corner cases "right": the goal is to
provide you with a rock solid clean up tool.

This is similar to this code fragment:

   eval ... code following scope_guard ...
   {
      local $@;
      eval BLOCK;
      eval { $Guard::DIED->() } if $@;
   }
   die if $@;

Except it is much faster, and the whole thing gets executed even when the
BLOCK calls C<exit>, C<goto>, C<last> or escapes via other means.

If multiple BLOCKs are registered to the same scope, they will be executed
in reverse order. Stuff like C<local> is managed via the same mechanism,
so variables C<local>ised after calling C<scope_guard> will be restored
when the guard runs.

See B<EXCEPTIONS>, below, for an explanation of exception handling
(C<die>) within guard blocks.

Example: temporarily change the timezone for the current process,
ensuring it will be reset when the C<if> scope is exited:

   use Guard;
   use POSIX ();

   if ($need_to_switch_tz) {
      # make sure we call tzset after $ENV{TZ} has been restored
      scope_guard { POSIX::tzset };

      # localise after the scope_guard, so it gets undone in time
      local $ENV{TZ} = "Europe/London";
      POSIX::tzset;

      # do something with the new timezone
   }

=item my $guard = guard BLOCK

Behaves the same as C<scope_guard>, except that instead of executing
the block on scope exit, it returns an object whose lifetime determines
when the BLOCK gets executed: when the last reference to the object gets
destroyed, the BLOCK gets executed as with C<scope_guard>.

The returned object can be copied as many times as you want.

See B<EXCEPTIONS>, below, for an explanation of exception handling
(C<die>) within guard blocks.

Example: acquire a Coro::Semaphore for a second by registering a
timer. The timer callback references the guard used to unlock it again.

   use AnyEvent;
   use Coro::Semaphore;

   my $sem = new Coro::Semaphore;

   sub lock_1s {
      $sem->down;
      my $guard = guard { $sem->up };

      my $timer;
      $timer = AnyEvent->timer (after => 1, sub {
         # do something
         undef $sem;
         undef $timer;
      });
   }

The advantage of doing this with a guard instead of simply calling C<<
$sem->down >> in the callback is that you can opt not to create the timer,
or your code can throw an exception before it can create the timer, or you
can create multiple timers or other event watchers and only when the last
one gets executed will the lock be unlocked.

=item Guard::cancel $guard

Calling this function will "disable" the guard object returned by the
C<guard> function, i.e. it will free the BLOCK originally passed to
C<guard >and will arrange for the BLOCK not to be executed.

This can be useful when you use C<guard> to create a fatal cleanup handler
and later decide it is no longer needed.

=cut

1;

=back

=head1 EXCEPTIONS

Guard blocks should not normally throw exceptions (that is, C<die>). After
all, they are usually used to clean up after such exceptions. However, if
something truly exceptional is happening, a guard block should be allowed
to die. Also, programming errors are a large source of exceptions, and the
programmer certainly wants to know about those.

Since in most cases, the block executing when the guard gets executes does
not know or does not care about the guard blocks, it makes little sense to
let containing code handle the exception.

Therefore, whenever a guard block throws an exception, it will be caught,
and this module will call the code reference stored in C<$Guard::DIED>
(with C<$@> set to the actual exception), which is similar to how most
event loops handle this case.

The code reference stored in C<$Guard::DIED> should not die (behaviour is
not guaranteed, but right now, the exception will simply be ignored).

The default for C<$Guard::DIED> is to call C<warn "$@">.

=head1 AUTHOR

 Marc Lehmann <schmorp@schmorp.de>
 http://home.schmorp.de/

=head1 THANKS

To Marco Maisenhelder, who reminded me of the C<$Guard::DIED> solution to
the problem of exceptions.

=cut

Revision:	1.5
Committed:	Sat Dec 13 18:42:45 2008 UTC (15 years, 5 months ago) by root
Branch:	MAIN
Changes since 1.4:	+1 -1 lines
Log Message:	* empty log message *
#	User	Rev	Content
1	root	1.1	=head1 NAME
2
3			Guard - safe cleanup blocks
4
5			=head1 SYNOPSIS
6
7			use Guard;
8
9	root	1.4	# temporarily chdir to "/etc" directory, but make sure
10			# to go back to "/" no matter how myfun exits:
11			sub dosomething {
12			scope_guard { chdir "/" };
13			chdir "/etc";
14
15			call_function_that_might_die_or_other_fun_stuff;
16			}
17
18	root	1.1	=head1 DESCRIPTION
19
20			This module implements so-called "guards". A guard is something (usually
21			an object) that "guards" a resource, ensuring that it is cleaned up when
22			expected.
23
24			Specifically, this module supports two different types of guards: guard
25			objects, which execute a given code block when destroyed, and scoped
26			guards, which are tied to the scope exit.
27
28	root	1.2	=head1 FUNCTIONS
29
30			This module currently exports the C<scope_guard> and C<guard> functions by
31			default.
32
33	root	1.1	=over 4
34
35			=cut
36
37			package Guard;
38
39			BEGIN {
40	root	1.3	$VERSION = '0.1';
41	root	1.1	@ISA = qw(Exporter);
42	root	1.3	@EXPORT = qw(guard scope_guard);
43	root	1.1
44			require Exporter;
45
46			require XSLoader;
47			XSLoader::load Guard, $VERSION;
48			}
49
50			our $DIED = sub { warn "$@" };
51
52			=item scope_guard BLOCK
53
54			Registers a block that is executed when the current scope (block,
55			function, method, eval etc.) is exited.
56
57	root	1.2	The description below sounds a bit complicated, but that's just because
58			C<scope_guard> tries to get even corner cases "right": the goal is to
59			provide you with a rock solid clean up tool.
60
61	root	1.1	This is similar to this code fragment:
62
63			eval ... code following scope_guard ...
64			{
65			local $@;
66			eval BLOCK;
67			eval { $Guard::DIED->() } if $@;
68			}
69	root	1.2	die if $@;
70	root	1.1
71			Except it is much faster, and the whole thing gets executed even when the
72			BLOCK calls C<exit>, C<goto>, C<last> or escapes via other means.
73
74	root	1.4	If multiple BLOCKs are registered to the same scope, they will be executed
75			in reverse order. Stuff like C<local> is managed via the same mechanism,
76			so variables C<local>ised after calling C<scope_guard> will be restored
77			when the guard runs.
78
79	root	1.1	See B<EXCEPTIONS>, below, for an explanation of exception handling
80			(C<die>) within guard blocks.
81
82	root	1.4	Example: temporarily change the timezone for the current process,
83			ensuring it will be reset when the C<if> scope is exited:
84
85			use Guard;
86			use POSIX ();
87
88			if ($need_to_switch_tz) {
89			# make sure we call tzset after $ENV{TZ} has been restored
90			scope_guard { POSIX::tzset };
91	root	1.1
92	root	1.4	# localise after the scope_guard, so it gets undone in time
93			local $ENV{TZ} = "Europe/London";
94			POSIX::tzset;
95	root	1.1
96	root	1.4	# do something with the new timezone
97	root	1.1	}
98
99			=item my $guard = guard BLOCK
100
101			Behaves the same as C<scope_guard>, except that instead of executing
102			the block on scope exit, it returns an object whose lifetime determines
103			when the BLOCK gets executed: when the last reference to the object gets
104			destroyed, the BLOCK gets executed as with C<scope_guard>.
105
106			The returned object can be copied as many times as you want.
107
108			See B<EXCEPTIONS>, below, for an explanation of exception handling
109			(C<die>) within guard blocks.
110
111			Example: acquire a Coro::Semaphore for a second by registering a
112			timer. The timer callback references the guard used to unlock it again.
113
114			use AnyEvent;
115			use Coro::Semaphore;
116
117			my $sem = new Coro::Semaphore;
118
119			sub lock_1s {
120			$sem->down;
121			my $guard = guard { $sem->up };
122
123			my $timer;
124			$timer = AnyEvent->timer (after => 1, sub {
125			# do something
126			undef $sem;
127			undef $timer;
128			});
129			}
130
131			The advantage of doing this with a guard instead of simply calling C<<
132			$sem->down >> in the callback is that you can opt not to create the timer,
133			or your code can throw an exception before it can create the timer, or you
134			can create multiple timers or other event watchers and only when the last
135			one gets executed will the lock be unlocked.
136
137			=item Guard::cancel $guard
138
139			Calling this function will "disable" the guard object returned by the
140			C<guard> function, i.e. it will free the BLOCK originally passed to
141			C<guard >and will arrange for the BLOCK not to be executed.
142
143			This can be useful when you use C<guard> to create a fatal cleanup handler
144			and later decide it is no longer needed.
145
146			=cut
147
148			1;
149
150			=back
151
152			=head1 EXCEPTIONS
153
154	root	1.5	Guard blocks should not normally throw exceptions (that is, C<die>). After
155	root	1.1	all, they are usually used to clean up after such exceptions. However, if
156			something truly exceptional is happening, a guard block should be allowed
157			to die. Also, programming errors are a large source of exceptions, and the
158			programmer certainly wants to know about those.
159
160			Since in most cases, the block executing when the guard gets executes does
161			not know or does not care about the guard blocks, it makes little sense to
162			let containing code handle the exception.
163
164			Therefore, whenever a guard block throws an exception, it will be caught,
165			and this module will call the code reference stored in C<$Guard::DIED>
166			(with C<$@> set to the actual exception), which is similar to how most
167			event loops handle this case.
168
169			The code reference stored in C<$Guard::DIED> should not die (behaviour is
170			not guaranteed, but right now, the exception will simply be ignored).
171
172			The default for C<$Guard::DIED> is to call C<warn "$@">.
173
174			=head1 AUTHOR
175
176			Marc Lehmann <schmorp@schmorp.de>
177			http://home.schmorp.de/
178
179			=head1 THANKS
180
181			To Marco Maisenhelder, who reminded me of the C<$Guard::DIED> solution to
182			the problem of exceptions.
183
184			=cut
185