… | |
… | |
2 | |
2 | |
3 | Guard - safe cleanup blocks |
3 | Guard - safe cleanup blocks |
4 | |
4 | |
5 | =head1 SYNOPSIS |
5 | =head1 SYNOPSIS |
6 | |
6 | |
7 | use Guard; |
7 | use Guard; |
|
|
8 | |
|
|
9 | # temporarily chdir to "/etc" directory, but make sure |
|
|
10 | # to go back to "/" no matter how myfun exits: |
|
|
11 | sub myfun { |
|
|
12 | scope_guard { chdir "/" }; |
|
|
13 | chdir "/etc"; |
|
|
14 | |
|
|
15 | code_that_might_die_or_does_other_fun_stuff; |
|
|
16 | } |
8 | |
17 | |
9 | =head1 DESCRIPTION |
18 | =head1 DESCRIPTION |
10 | |
19 | |
11 | This module implements so-called "guards". A guard is something (usually |
20 | This module implements so-called "guards". A guard is something (usually |
12 | an object) that "guards" a resource, ensuring that it is cleaned up when |
21 | an object) that "guards" a resource, ensuring that it is cleaned up when |
… | |
… | |
14 | |
23 | |
15 | Specifically, this module supports two different types of guards: guard |
24 | Specifically, this module supports two different types of guards: guard |
16 | objects, which execute a given code block when destroyed, and scoped |
25 | objects, which execute a given code block when destroyed, and scoped |
17 | guards, which are tied to the scope exit. |
26 | guards, which are tied to the scope exit. |
18 | |
27 | |
|
|
28 | =head1 FUNCTIONS |
|
|
29 | |
|
|
30 | This module currently exports the C<scope_guard> and C<guard> functions by |
|
|
31 | default. |
|
|
32 | |
19 | =over 4 |
33 | =over 4 |
20 | |
34 | |
21 | =cut |
35 | =cut |
22 | |
36 | |
23 | package Guard; |
37 | package Guard; |
24 | |
38 | |
|
|
39 | no warnings; |
|
|
40 | |
25 | BEGIN { |
41 | BEGIN { |
26 | $VERSION = '0.01'; |
42 | $VERSION = '0.5'; |
27 | @ISA = qw(Exporter); |
43 | @ISA = qw(Exporter); |
28 | @EXPORT = qw(guard scope_guard cancel); |
44 | @EXPORT = qw(guard scope_guard); |
29 | |
45 | |
30 | require Exporter; |
46 | require Exporter; |
31 | |
47 | |
32 | require XSLoader; |
48 | require XSLoader; |
33 | XSLoader::load Guard, $VERSION; |
49 | XSLoader::load Guard, $VERSION; |
… | |
… | |
38 | =item scope_guard BLOCK |
54 | =item scope_guard BLOCK |
39 | |
55 | |
40 | Registers a block that is executed when the current scope (block, |
56 | Registers a block that is executed when the current scope (block, |
41 | function, method, eval etc.) is exited. |
57 | function, method, eval etc.) is exited. |
42 | |
58 | |
|
|
59 | See the EXCEPTIONS section for an explanation of how exceptions |
|
|
60 | (i.e. C<die>) are handled inside guard blocks. |
|
|
61 | |
|
|
62 | The description below sounds a bit complicated, but that's just because |
|
|
63 | C<scope_guard> tries to get even corner cases "right": the goal is to |
|
|
64 | provide you with a rock solid clean up tool. |
|
|
65 | |
43 | This is similar to this code fragment: |
66 | The behaviour is similar to this code fragment: |
44 | |
67 | |
45 | eval ... code following scope_guard ... |
68 | eval ... code following scope_guard ... |
46 | { |
69 | { |
47 | local $@; |
70 | local $@; |
48 | eval BLOCK; |
71 | eval BLOCK; |
49 | eval { $Guard::DIED->() } if $@; |
72 | eval { $Guard::DIED->() } if $@; |
50 | } |
73 | } |
|
|
74 | die if $@; |
51 | |
75 | |
52 | Except it is much faster, and the whole thing gets executed even when the |
76 | Except it is much faster, and the whole thing gets executed even when the |
53 | BLOCK calls C<exit>, C<goto>, C<last> or escapes via other means. |
77 | BLOCK calls C<exit>, C<goto>, C<last> or escapes via other means. |
54 | |
78 | |
55 | See B<EXCEPTIONS>, below, for an explanation of exception handling |
79 | If multiple BLOCKs are registered to the same scope, they will be executed |
56 | (C<die>) within guard blocks. |
80 | in reverse order. Other scope-related things such as C<local> are managed |
|
|
81 | via the same mechanism, so variables C<local>ised I<after> calling |
|
|
82 | C<scope_guard> will be restored when the guard runs. |
57 | |
83 | |
58 | Example: Temporarily change the directory to F</etc> and make sure it's |
84 | Example: temporarily change the timezone for the current process, |
59 | set back to F</> when the function returns: |
85 | ensuring it will be reset when the C<if> scope is exited: |
60 | |
86 | |
61 | sub dosomething { |
87 | use Guard; |
62 | scope_guard { chdir "/" }; |
88 | use POSIX (); |
63 | chdir "/etc"; |
|
|
64 | |
89 | |
65 | ... |
90 | if ($need_to_switch_tz) { |
|
|
91 | # make sure we call tzset after $ENV{TZ} has been restored |
|
|
92 | scope_guard { POSIX::tzset }; |
|
|
93 | |
|
|
94 | # localise after the scope_guard, so it gets undone in time |
|
|
95 | local $ENV{TZ} = "Europe/London"; |
|
|
96 | POSIX::tzset; |
|
|
97 | |
|
|
98 | # do something with the new timezone |
66 | } |
99 | } |
67 | |
100 | |
68 | =item my $guard = guard BLOCK |
101 | =item my $guard = guard BLOCK |
69 | |
102 | |
70 | Behaves the same as C<scope_guard>, except that instead of executing |
103 | Behaves the same as C<scope_guard>, except that instead of executing |
… | |
… | |
72 | when the BLOCK gets executed: when the last reference to the object gets |
105 | when the BLOCK gets executed: when the last reference to the object gets |
73 | destroyed, the BLOCK gets executed as with C<scope_guard>. |
106 | destroyed, the BLOCK gets executed as with C<scope_guard>. |
74 | |
107 | |
75 | The returned object can be copied as many times as you want. |
108 | The returned object can be copied as many times as you want. |
76 | |
109 | |
77 | See B<EXCEPTIONS>, below, for an explanation of exception handling |
110 | See the EXCEPTIONS section for an explanation of how exceptions |
78 | (C<die>) within guard blocks. |
111 | (i.e. C<die>) are handled inside guard blocks. |
79 | |
112 | |
80 | Example: acquire a Coro::Semaphore for a second by registering a |
113 | Example: acquire a Coro::Semaphore for a second by registering a |
81 | timer. The timer callback references the guard used to unlock it again. |
114 | timer. The timer callback references the guard used to unlock it |
|
|
115 | again. (Please ignore the fact that C<Coro::Semaphore> has a C<guard> |
|
|
116 | method that does this already): |
82 | |
117 | |
|
|
118 | use Guard; |
83 | use AnyEvent; |
119 | use AnyEvent; |
84 | use Coro::Semaphore; |
120 | use Coro::Semaphore; |
85 | |
121 | |
86 | my $sem = new Coro::Semaphore; |
122 | my $sem = new Coro::Semaphore; |
87 | |
123 | |
88 | sub lock_1s { |
124 | sub lock_for_a_second { |
89 | $sem->down; |
125 | $sem->down; |
90 | my $guard = guard { $sem->up }; |
126 | my $guard = guard { $sem->up }; |
91 | |
127 | |
92 | my $timer; |
128 | my $timer; |
93 | $timer = AnyEvent->timer (after => 1, sub { |
129 | $timer = AnyEvent->timer (after => 1, sub { |
… | |
… | |
99 | |
135 | |
100 | The advantage of doing this with a guard instead of simply calling C<< |
136 | The advantage of doing this with a guard instead of simply calling C<< |
101 | $sem->down >> in the callback is that you can opt not to create the timer, |
137 | $sem->down >> in the callback is that you can opt not to create the timer, |
102 | or your code can throw an exception before it can create the timer, or you |
138 | or your code can throw an exception before it can create the timer, or you |
103 | can create multiple timers or other event watchers and only when the last |
139 | can create multiple timers or other event watchers and only when the last |
104 | one gets executed will the lock be unlocked. |
140 | one gets executed will the lock be unlocked. Using the C<guard>, you do |
|
|
141 | not have to worry about catching all the places where you have to unlock |
|
|
142 | the semaphore. |
105 | |
143 | |
106 | =item Guard::cancel $guard |
144 | =item $guard->cancel |
107 | |
145 | |
108 | Calling this function will "disable" the guard object returned by the |
146 | Calling this function will "disable" the guard object returned by the |
109 | C<guard> function, i.e. it will free the BLOCK originally passed to |
147 | C<guard> function, i.e. it will free the BLOCK originally passed to |
110 | C<guard >and will arrange for the BLOCK not to be executed. |
148 | C<guard >and will arrange for the BLOCK not to be executed. |
111 | |
149 | |
… | |
… | |
118 | |
156 | |
119 | =back |
157 | =back |
120 | |
158 | |
121 | =head1 EXCEPTIONS |
159 | =head1 EXCEPTIONS |
122 | |
160 | |
123 | Guard blocks should not normally throw exceptions (e.g. C<die>), after |
161 | Guard blocks should not normally throw exceptions (that is, C<die>). After |
124 | all, they are usually used to clean up after such exceptions. However, if |
162 | all, they are usually used to clean up after such exceptions. However, if |
125 | something truly exceptional is happening, a guard block should be allowed |
163 | something truly exceptional is happening, a guard block should be allowed |
126 | to die. Also, programming errors are a large source of exceptions, and the |
164 | to die. Also, programming errors are a large source of exceptions, and the |
127 | programmer certainly wants to know about those. |
165 | programmer certainly wants to know about those. |
128 | |
166 | |
129 | Since in most cases, the block executing when the guard gets executes does |
167 | Since in most cases, the block executing when the guard gets executed does |
130 | not know or does not care about the guard blocks, it makes little sense to |
168 | not know or does not care about the guard blocks, it makes little sense to |
131 | let containing code handle the exception. |
169 | let containing code handle the exception. |
132 | |
170 | |
133 | Therefore, whenever a guard block throws an exception, it will be caught, |
171 | Therefore, whenever a guard block throws an exception, it will be caught, |
134 | and this module will call the code reference stored in C<$Guard::DIED> |
172 | followed by calling the code reference stored in C<$Guard::DIED> (with |
135 | (with C<$@> set to the actual exception), which is similar to how most |
173 | C<$@> set to the actual exception), which is similar to how most event |
136 | event loops handle this case. |
174 | loops handle this case. |
|
|
175 | |
|
|
176 | The default for C<$Guard::DIED> is to call C<warn "$@">. |
|
|
177 | |
|
|
178 | The C<$@> variable will be restored to its value before the guard call in |
|
|
179 | all cases, so guards will not disturb C<$@> in any way. |
137 | |
180 | |
138 | The code reference stored in C<$Guard::DIED> should not die (behaviour is |
181 | The code reference stored in C<$Guard::DIED> should not die (behaviour is |
139 | not guaranteed, but right now, the exception will simply be ignored). |
182 | not guaranteed, but right now, the exception will simply be ignored). |
140 | |
|
|
141 | The default for C<$Guard::DIED> is to call C<warn "$@">. |
|
|
142 | |
183 | |
143 | =head1 AUTHOR |
184 | =head1 AUTHOR |
144 | |
185 | |
145 | Marc Lehmann <schmorp@schmorp.de> |
186 | Marc Lehmann <schmorp@schmorp.de> |
146 | http://home.schmorp.de/ |
187 | http://home.schmorp.de/ |
147 | |
188 | |
148 | =head1 THANKS |
189 | =head1 THANKS |
149 | |
190 | |
150 | To Marco Maisenhelder, who reminded me of the C<$Guard::DIED> solution to |
191 | Thanks to Marco Maisenhelder, who reminded me of the C<$Guard::DIED> |
151 | the problem of exceptions. |
192 | solution to the problem of exceptions. |
152 | |
193 | |
153 | =cut |
194 | =cut |
154 | |
195 | |