1 # PAUSE doesn't seem to case about this in t/role-basic/lib, but just in case ...
2 package # Hide from PAUSE
8 use vars qw(@EXPORT @EXPORT_OK $VERSION @ISA);
17 $VERSION = eval $VERSION;
19 @EXPORT = @EXPORT_OK = qw(try catch finally);
21 $Carp::Internal{+__PACKAGE__}++;
23 # Need to prototype as @ not $$ because of the way Perl evaluates the prototype.
24 # Keeping it at $$ means you only ever get 1 sub because we need to eval in a list
25 # context & not a scalar one
28 my ( $try, @code_refs ) = @_;
30 # we need to save this here, the eval block will be in scalar context due
32 my $wantarray = wantarray;
34 my ( $catch, @finally );
36 # find labeled blocks in the argument list.
37 # catch and finally tag the blocks by blessing a scalar reference to them.
38 foreach my $code_ref (@code_refs) {
39 next unless $code_ref;
41 my $ref = ref($code_ref);
43 if ( $ref eq 'Try::Tiny::Catch' ) {
44 $catch = ${$code_ref};
45 } elsif ( $ref eq 'Try::Tiny::Finally' ) {
46 push @finally, ${$code_ref};
49 confess("Unknown code ref type given '${ref}'. Check your usage & try again");
53 # save the value of $@ so we can set $@ back to it in the beginning of the eval
56 my ( @ret, $error, $failed );
58 # FIXME consider using local $SIG{__DIE__} to accumulate all errors. It's
59 # not perfect, but we could provide a list of additional errors for
63 # localize $@ to prevent clobbering of previous value by a successful
67 # failed will be true if the eval dies, because 1 will not be returned
72 # evaluate the try block in the correct context
75 } elsif ( defined $wantarray ) {
81 return 1; # properly set $fail to false
84 # copy $@ to $error; when we leave this scope, local $@ will revert $@
85 # back to its previous value
89 # set up a scope guard to invoke the finally block at the end
91 map { Try::Tiny::ScopeGuard->_new($_, $failed ? $error : ()) }
94 # at this point $failed contains a true value if the eval died, even if some
95 # destructor overwrote $@ as the eval was unwinding.
97 # if we got an error, invoke the catch block.
99 # This works like given($error), but is backwards compatible and
100 # sets $_ in the dynamic scope for the body of C<$catch>
102 return $catch->($error);
105 # in case when() was used without an explicit return, the C<for>
106 # loop will be aborted and there's no useful return value
111 # no failure, $@ is back to what it was, everything is fine
112 return $wantarray ? @ret : $ret[0];
117 my ( $block, @rest ) = @_;
120 bless(\$block, 'Try::Tiny::Catch'),
126 my ( $block, @rest ) = @_;
129 bless(\$block, 'Try::Tiny::Finally'),
135 package # hide from PAUSE
136 Try::Tiny::ScopeGuard;
144 my @guts = @{ shift() };
145 my $code = shift @guts;
158 Try::Tiny - minimal try/catch with proper localization of $@
162 # handle errors with a catch handler
166 warn "caught error: $_"; # not $@
169 # just silence errors
176 This module provides bare bones C<try>/C<catch>/C<finally> statements that are designed to
177 minimize common mistakes with eval blocks, and NOTHING else.
179 This is unlike L<TryCatch> which provides a nice syntax and avoids adding
180 another call stack layer, and supports calling C<return> from the try block to
181 return from the parent subroutine. These extra features come at a cost of a few
182 dependencies, namely L<Devel::Declare> and L<Scope::Upper> which are
183 occasionally problematic, and the additional catch filtering uses L<Moose>
184 type constraints which may not be desirable either.
186 The main focus of this module is to provide simple and reliable error handling
187 for those having a hard time installing L<TryCatch>, but who still want to
188 write correct C<eval> blocks without 5 lines of boilerplate each time.
190 It's designed to work as correctly as possible in light of the various
191 pathological edge cases (see L<BACKGROUND>) and to be compatible with any style
192 of error values (simple strings, references, objects, overloaded objects, etc).
194 If the try block dies, it returns the value of the last statement executed in
195 the catch block, if there is one. Otherwise, it returns C<undef> in scalar
196 context or the empty list in list context. The following two examples both
197 assign C<"bar"> to C<$x>.
199 my $x = try { die "foo" } catch { "bar" };
201 my $x = eval { die "foo" } || "bar";
203 You can add finally blocks making the following true.
206 try { die 'foo' } finally { $x = 'bar' };
207 try { die 'foo' } catch { warn "Got a die: $_" } finally { $x = 'bar' };
209 Finally blocks are always executed making them suitable for cleanup code
210 which cannot be handled using local. You can add as many finally blocks to a
211 given try block as you like.
215 All functions are exported by default using L<Exporter>.
217 If you need to rename the C<try>, C<catch> or C<finally> keyword consider using
218 L<Sub::Import> to get L<Sub::Exporter>'s flexibility.
224 Takes one mandatory try subroutine, an optional catch subroutine & finally
227 The mandatory subroutine is evaluated in the context of an C<eval> block.
229 If no error occurred the value from the first block is returned, preserving
232 If there was an error and the second subroutine was given it will be invoked
233 with the error in C<$_> (localized) and as that block's first and only
236 C<$@> does B<not> contain the error. Inside the C<catch> block it has the same
237 value it had before the C<try> block was executed.
239 Note that the error may be false, but if that happens the C<catch> block will
242 Once all execution is finished then the finally block if given will execute.
246 Intended to be used in the second argument position of C<try>.
248 Returns a reference to the subroutine it was given but blessed as
249 C<Try::Tiny::Catch> which allows try to decode correctly what to do
250 with this code reference.
254 Inside the catch block the caught error is stored in C<$_>, while previous
255 value of C<$@> is still available for use. This value may or may not be
256 meaningful depending on what happened before the C<try>, but it might be a good
257 idea to preserve it in an error stack.
259 For code that captures C<$@> when throwing new errors (i.e.
260 L<Class::Throwable>), you'll need to do:
281 Intended to be the second or third element of C<try>. Finally blocks are always
282 executed in the event of a successful C<try> or if C<catch> is run. This allows
283 you to locate cleanup code which cannot be done via C<local()> e.g. closing a file
286 When invoked, the finally block is passed the error that was caught. If no
287 error was caught, it is passed nothing. In other words, the following code
288 does just what you would expect:
293 # ...code run in case of error
296 print "The try block died with: @_\n";
298 print "The try block ran without error.\n";
302 B<You must always do your own error handling in the finally block>. C<Try::Tiny> will
303 not do anything about handling possible errors coming from code located in these
306 In the same way C<catch()> blesses the code reference this subroutine does the same
307 except it bless them as C<Try::Tiny::Finally>.
313 There are a number of issues with C<eval>.
317 When you run an eval block and it succeeds, C<$@> will be cleared, potentially
318 clobbering an error that is currently being caught.
320 This causes action at a distance, clearing previous errors your caller may have
323 C<$@> must be properly localized before invoking C<eval> in order to avoid this
326 More specifically, C<$@> is clobbered at the beginning of the C<eval>, which
327 also makes it impossible to capture the previous error before you die (for
328 instance when making exception objects with error stacks).
330 For this reason C<try> will actually set C<$@> to its previous value (before
331 the localization) in the beginning of the C<eval> block.
333 =head2 Localizing $@ silently masks errors
335 Inside an eval block C<die> behaves sort of like:
339 return_undef_from_eval();
342 This means that if you were polite and localized C<$@> you can't die in that
343 scope, or your error will be discarded (printing "Something's wrong" instead).
345 The workaround is very ugly:
356 =head2 $@ might not be a true value
364 because due to the previous caveats it may have been unset.
366 C<$@> could also be an overloaded error object that evaluates to false, but
367 that's asking for trouble anyway.
369 The classic failure mode is:
371 sub Object::DESTROY {
376 my $obj = Object->new;
385 In this case since C<Object::DESTROY> is not localizing C<$@> but still uses
386 C<eval>, it will set C<$@> to C<"">.
388 The destructor is called when the stack is unwound, after C<die> sets C<$@> to
389 C<"foo at Foo.pm line 42\n">, so by the time C<if ( $@ )> is evaluated it has
390 been cleared by C<eval> in the destructor.
392 The workaround for this is even uglier than the previous ones. Even though we
393 can't save the value of C<$@> from code that doesn't localize, we can at least
394 be sure the eval was aborted due to an error:
396 my $failed = not eval {
402 This is because an C<eval> that caught a C<die> will always return a false
407 Using Perl 5.10 you can use L<perlsyn/"Switch statements">.
409 The C<catch> block is invoked in a topicalizer context (like a C<given> block),
410 but note that you can't return a useful value from C<catch> using the C<when>
411 blocks without an explicit C<return>.
413 This is somewhat similar to Perl 6's C<CATCH> blocks. You can use it to
414 concisely match errors:
419 when (/^Can't locate .*?\.pm in \@INC/) { } # ignore
429 C<@_> is not available within the C<try> block, so you need to copy your
430 arglist. In case you want to work with argument values directly via C<@_>
431 aliasing (i.e. allow C<$_[1] = "foo">), you need to pass C<@_> by reference:
434 my ( $self, @args ) = @_;
435 try { $self->bar(@args) }
443 try { $_ = $self->bar($_) for @$args }
448 C<return> returns from the C<try> block, not from the parent sub (note that
449 this is also how C<eval> works, but not how L<TryCatch> works):
452 try { return "foo" };
460 C<try> introduces another caller stack frame. L<Sub::Uplevel> is not used. L<Carp>
461 will not report this when using full stack traces, though, because
462 C<%Carp::Internal> is used. This lack of magic is considered a feature.
466 The value of C<$_> in the C<catch> block is not guaranteed to be the value of
467 the exception thrown (C<$@>) in the C<try> block. There is no safe way to
468 ensure this, since C<eval> may be used unhygenically in destructors. The only
469 guarantee is that the C<catch> will be called if an exception is thrown.
473 The return value of the C<catch> block is not ignored, so if testing the result
474 of the expression for truth on success, be sure to return a false value from
482 return; # avoid returning a true value;
489 C<$SIG{__DIE__}> is still in effect.
491 Though it can be argued that C<$SIG{__DIE__}> should be disabled inside of
492 C<eval> blocks, since it isn't people have grown to rely on it. Therefore in
493 the interests of compatibility, C<try> does not disable C<$SIG{__DIE__}> for
494 the scope of the error throwing code.
498 Lexical C<$_> may override the one set by C<catch>.
500 For example Perl 5.10's C<given> form uses a lexical C<$_>, creating some
508 warn $_; # will print $foo, not the error
509 warn $_[0]; # instead, get the error like this
522 Much more feature complete, more convenient semantics, but at the cost of
523 implementation complexity.
527 Automatic error throwing for builtin functions and more. Also designed to
528 work well with C<given>/C<when>.
532 A lightweight role for rolling your own exception classes.
536 Exception object implementation with a C<try> statement. Does not localize
539 =item L<Exception::Class::TryCatch>
541 Provides a C<catch> statement, but properly calling C<eval> is your
544 The C<try> keyword pushes C<$@> onto an error stack, avoiding some of the
545 issues with C<$@>, but you still need to localize to prevent clobbering.
549 =head1 LIGHTNING TALK
551 I gave a lightning talk about this module, you can see the slides (Firefox
554 L<http://nothingmuch.woobling.org/talks/takahashi.xul?data=yapc_asia_2009/try_tiny.txt>
558 L<http://nothingmuch.woobling.org/talks/yapc_asia_2009/try_tiny.yml>
560 =head1 VERSION CONTROL
562 L<http://github.com/nothingmuch/try-tiny/>
566 Yuval Kogman E<lt>nothingmuch@woobling.orgE<gt>
570 Copyright (c) 2009 Yuval Kogman. All rights reserved.
571 This program is free software; you can redistribute
572 it and/or modify it under the terms of the MIT license.