… | |
… | |
107 | |
107 | |
108 | See the EXCEPTIONS section for an explanation of how exceptions |
108 | See the EXCEPTIONS section for an explanation of how exceptions |
109 | (i.e. C<die>) are handled inside guard blocks. |
109 | (i.e. C<die>) are handled inside guard blocks. |
110 | |
110 | |
111 | Example: acquire a Coro::Semaphore for a second by registering a |
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. |
112 | timer. The timer callback references the guard used to unlock it |
|
|
113 | again. (Please ignore the fact that C<Coro::Semaphore> has a C<guard> |
|
|
114 | method that does this already): |
113 | |
115 | |
114 | use Guard; |
116 | use Guard; |
115 | use AnyEvent; |
117 | use AnyEvent; |
116 | use Coro::Semaphore; |
118 | use Coro::Semaphore; |
117 | |
119 | |
… | |
… | |
131 | |
133 | |
132 | The advantage of doing this with a guard instead of simply calling C<< |
134 | The advantage of doing this with a guard instead of simply calling C<< |
133 | $sem->down >> in the callback is that you can opt not to create the timer, |
135 | $sem->down >> in the callback is that you can opt not to create the timer, |
134 | or your code can throw an exception before it can create the timer, or you |
136 | or your code can throw an exception before it can create the timer, or you |
135 | can create multiple timers or other event watchers and only when the last |
137 | can create multiple timers or other event watchers and only when the last |
136 | one gets executed will the lock be unlocked. |
138 | one gets executed will the lock be unlocked. Using the C<guard>, you do |
|
|
139 | not have to worry about catching all the places where you have to unlock |
|
|
140 | the semaphore. |
137 | |
141 | |
138 | =item Guard::cancel $guard |
142 | =item Guard::cancel $guard |
139 | |
143 | |
140 | Calling this function will "disable" the guard object returned by the |
144 | Calling this function will "disable" the guard object returned by the |
141 | C<guard> function, i.e. it will free the BLOCK originally passed to |
145 | C<guard> function, i.e. it will free the BLOCK originally passed to |
… | |
… | |
165 | Therefore, whenever a guard block throws an exception, it will be caught, |
169 | Therefore, whenever a guard block throws an exception, it will be caught, |
166 | and this module will call the code reference stored in C<$Guard::DIED> |
170 | and this module will call the code reference stored in C<$Guard::DIED> |
167 | (with C<$@> set to the actual exception), which is similar to how most |
171 | (with C<$@> set to the actual exception), which is similar to how most |
168 | event loops handle this case. |
172 | event loops handle this case. |
169 | |
173 | |
|
|
174 | The default for C<$Guard::DIED> is to call C<warn "$@">. |
|
|
175 | |
|
|
176 | The C<$@> variable will be restored to its value before the guard call in |
|
|
177 | all cases, so guards will not disturb C<$@> in any way. |
|
|
178 | |
170 | The code reference stored in C<$Guard::DIED> should not die (behaviour is |
179 | The code reference stored in C<$Guard::DIED> should not die (behaviour is |
171 | not guaranteed, but right now, the exception will simply be ignored). |
180 | not guaranteed, but right now, the exception will simply be ignored). |
172 | |
|
|
173 | The default for C<$Guard::DIED> is to call C<warn "$@">. |
|
|
174 | |
181 | |
175 | =head1 AUTHOR |
182 | =head1 AUTHOR |
176 | |
183 | |
177 | Marc Lehmann <schmorp@schmorp.de> |
184 | Marc Lehmann <schmorp@schmorp.de> |
178 | http://home.schmorp.de/ |
185 | http://home.schmorp.de/ |