3 # ABSTRACT: minimal try/catch with proper preservation of $@
10 use Exporter 5.57 'import';
11 our @EXPORT = our @EXPORT_OK = qw(try catch finally);
14 $Carp::Internal{+__PACKAGE__}++;
17 my $su = $INC{'Sub/Util.pm'} && defined &Sub::Util::set_subname;
18 my $sn = $INC{'Sub/Name.pm'} && eval { Sub::Name->VERSION(0.08) };
20 $su = eval { require Sub::Util; } && defined &Sub::Util::set_subname;
22 $sn = eval { require Sub::Name; Sub::Name->VERSION(0.08) };
26 *_subname = $su ? \&Sub::Util::set_subname
27 : $sn ? \&Sub::Name::subname
29 *_HAS_SUBNAME = ($su || $sn) ? sub(){1} : sub(){0};
32 # Need to prototype as @ not $$ because of the way Perl evaluates the prototype.
33 # Keeping it at $$ means you only ever get 1 sub because we need to eval in a list
34 # context & not a scalar one
37 my ( $try, @code_refs ) = @_;
39 # we need to save this here, the eval block will be in scalar context due
41 my $wantarray = wantarray;
43 # work around perl bug by explicitly initializing these, due to the likelyhood
44 # this will be used in global destruction (perl rt#119311)
45 my ( $catch, @finally ) = ();
47 # find labeled blocks in the argument list.
48 # catch and finally tag the blocks by blessing a scalar reference to them.
49 foreach my $code_ref (@code_refs) {
51 if ( ref($code_ref) eq 'Try::Tiny::Catch' ) {
52 croak 'A try() may not be followed by multiple catch() blocks'
54 $catch = ${$code_ref};
55 } elsif ( ref($code_ref) eq 'Try::Tiny::Finally' ) {
56 push @finally, ${$code_ref};
59 'try() encountered an unexpected argument ('
60 . ( defined $code_ref ? $code_ref : 'undef' )
61 . ') - perhaps a missing semi-colon before or'
66 # FIXME consider using local $SIG{__DIE__} to accumulate all errors. It's
67 # not perfect, but we could provide a list of additional errors for
70 # name the blocks if we have Sub::Name installed
72 _subname("${caller}::try {...} " => $try)
75 # save the value of $@ so we can set $@ back to it in the beginning of the eval
76 # and restore $@ after the eval finishes
81 # failed will be true if the eval dies, because 1 will not be returned
83 my $failed = not eval {
86 # evaluate the try block in the correct context
89 } elsif ( defined $wantarray ) {
95 return 1; # properly set $failed to false
98 # preserve the current error and reset the original value of $@
102 # set up a scope guard to invoke the finally block at the end
104 map { Try::Tiny::ScopeGuard->_new($_, $failed ? $error : ()) }
107 # at this point $failed contains a true value if the eval died, even if some
108 # destructor overwrote $@ as the eval was unwinding.
110 # if we got an error, invoke the catch block.
112 # This works like given($error), but is backwards compatible and
113 # sets $_ in the dynamic scope for the body of C<$catch>
115 return $catch->($error);
118 # in case when() was used without an explicit return, the C<for>
119 # loop will be aborted and there's no useful return value
124 # no failure, $@ is back to what it was, everything is fine
125 return $wantarray ? @ret : $ret[0];
130 my ( $block, @rest ) = @_;
132 croak 'Useless bare catch()' unless wantarray;
135 _subname("${caller}::catch {...} " => $block)
138 bless(\$block, 'Try::Tiny::Catch'),
144 my ( $block, @rest ) = @_;
146 croak 'Useless bare finally()' unless wantarray;
149 _subname("${caller}::finally {...} " => $block)
152 bless(\$block, 'Try::Tiny::Finally'),
158 package # hide from PAUSE
159 Try::Tiny::ScopeGuard;
161 use constant UNSTABLE_DOLLARAT => ($] < '5.013002') ? 1 : 0;
169 my ($code, @args) = @{ $_[0] };
171 local $@ if UNSTABLE_DOLLARAT;
177 "Execution of finally() block $code resulted in an exception, which "
178 . '*CAN NOT BE PROPAGATED* due to fundamental limitations of Perl. '
179 . 'Your program will continue as if this event never took place. '
180 . "Original exception text follows:\n\n"
181 . (defined $@ ? $@ : '$@ left undefined...')
196 You can use Try::Tiny's C<try> and C<catch> to expect and handle exceptional
197 conditions, avoiding quirks in Perl and common mistakes:
199 # handle errors with a catch handler
203 warn "caught error: $_"; # not $@
206 You can also use it like a standalone C<eval> to catch and ignore any error
207 conditions. Obviously, this is an extreme measure not to be undertaken
210 # just silence errors
217 This module provides bare bones C<try>/C<catch>/C<finally> statements that are designed to
218 minimize common mistakes with eval blocks, and NOTHING else.
220 This is unlike L<TryCatch> which provides a nice syntax and avoids adding
221 another call stack layer, and supports calling C<return> from the C<try> block to
222 return from the parent subroutine. These extra features come at a cost of a few
223 dependencies, namely L<Devel::Declare> and L<Scope::Upper> which are
224 occasionally problematic, and the additional catch filtering uses L<Moose>
225 type constraints which may not be desirable either.
227 The main focus of this module is to provide simple and reliable error handling
228 for those having a hard time installing L<TryCatch>, but who still want to
229 write correct C<eval> blocks without 5 lines of boilerplate each time.
231 It's designed to work as correctly as possible in light of the various
232 pathological edge cases (see L</BACKGROUND>) and to be compatible with any style
233 of error values (simple strings, references, objects, overloaded objects, etc).
235 If the C<try> block dies, it returns the value of the last statement executed in
236 the C<catch> block, if there is one. Otherwise, it returns C<undef> in scalar
237 context or the empty list in list context. The following examples all
238 assign C<"bar"> to C<$x>:
240 my $x = try { die "foo" } catch { "bar" };
241 my $x = try { die "foo" } || "bar";
242 my $x = (try { die "foo" }) // "bar";
244 my $x = eval { die "foo" } || "bar";
246 You can add C<finally> blocks, yielding the following:
249 try { die 'foo' } finally { $x = 'bar' };
250 try { die 'foo' } catch { warn "Got a die: $_" } finally { $x = 'bar' };
252 C<finally> blocks are always executed making them suitable for cleanup code
253 which cannot be handled using local. You can add as many C<finally> blocks to a
254 given C<try> block as you like.
256 Note that adding a C<finally> block without a preceding C<catch> block
257 suppresses any errors. This behaviour is consistent with using a standalone
258 C<eval>, but it is not consistent with C<try>/C<finally> patterns found in
259 other programming languages, such as Java, Python, Javascript or C#. If you
260 learnt the C<try>/C<finally> pattern from one of these languages, watch out for
265 All functions are exported by default using L<Exporter>.
267 If you need to rename the C<try>, C<catch> or C<finally> keyword consider using
268 L<Sub::Import> to get L<Sub::Exporter>'s flexibility.
274 Takes one mandatory C<try> subroutine, an optional C<catch> subroutine and C<finally>
277 The mandatory subroutine is evaluated in the context of an C<eval> block.
279 If no error occurred the value from the first block is returned, preserving
282 If there was an error and the second subroutine was given it will be invoked
283 with the error in C<$_> (localized) and as that block's first and only
286 C<$@> does B<not> contain the error. Inside the C<catch> block it has the same
287 value it had before the C<try> block was executed.
289 Note that the error may be false, but if that happens the C<catch> block will
292 Once all execution is finished then the C<finally> block, if given, will execute.
296 Intended to be used in the second argument position of C<try>.
298 Returns a reference to the subroutine it was given but blessed as
299 C<Try::Tiny::Catch> which allows try to decode correctly what to do
300 with this code reference.
304 Inside the C<catch> block the caught error is stored in C<$_>, while previous
305 value of C<$@> is still available for use. This value may or may not be
306 meaningful depending on what happened before the C<try>, but it might be a good
307 idea to preserve it in an error stack.
309 For code that captures C<$@> when throwing new errors (i.e.
310 L<Class::Throwable>), you'll need to do:
331 Intended to be the second or third element of C<try>. C<finally> blocks are always
332 executed in the event of a successful C<try> or if C<catch> is run. This allows
333 you to locate cleanup code which cannot be done via C<local()> e.g. closing a file
336 When invoked, the C<finally> block is passed the error that was caught. If no
337 error was caught, it is passed nothing. (Note that the C<finally> block does not
338 localize C<$_> with the error, since unlike in a C<catch> block, there is no way
339 to know if C<$_ == undef> implies that there were no errors.) In other words,
340 the following code does just what you would expect:
345 # ...code run in case of error
348 print "The try block died with: @_\n";
350 print "The try block ran without error.\n";
354 B<You must always do your own error handling in the C<finally> block>. C<Try::Tiny> will
355 not do anything about handling possible errors coming from code located in these
358 Furthermore B<exceptions in C<finally> blocks are not trappable and are unable
359 to influence the execution of your program>. This is due to limitation of
360 C<DESTROY>-based scope guards, which C<finally> is implemented on top of. This
361 may change in a future version of Try::Tiny.
363 In the same way C<catch()> blesses the code reference this subroutine does the same
364 except it bless them as C<Try::Tiny::Finally>.
370 There are a number of issues with C<eval>.
374 When you run an C<eval> block and it succeeds, C<$@> will be cleared, potentially
375 clobbering an error that is currently being caught.
377 This causes action at a distance, clearing previous errors your caller may have
380 C<$@> must be properly localized before invoking C<eval> in order to avoid this
383 More specifically, C<$@> is clobbered at the beginning of the C<eval>, which
384 also makes it impossible to capture the previous error before you die (for
385 instance when making exception objects with error stacks).
387 For this reason C<try> will actually set C<$@> to its previous value (the one
388 available before entering the C<try> block) in the beginning of the C<eval>
391 =head2 Localizing $@ silently masks errors
393 Inside an C<eval> block, C<die> behaves sort of like:
397 return_undef_from_eval();
400 This means that if you were polite and localized C<$@> you can't die in that
401 scope, or your error will be discarded (printing "Something's wrong" instead).
403 The workaround is very ugly:
414 =head2 $@ might not be a true value
422 because due to the previous caveats it may have been unset.
424 C<$@> could also be an overloaded error object that evaluates to false, but
425 that's asking for trouble anyway.
427 The classic failure mode is:
429 sub Object::DESTROY {
434 my $obj = Object->new;
443 In this case since C<Object::DESTROY> is not localizing C<$@> but still uses
444 C<eval>, it will set C<$@> to C<"">.
446 The destructor is called when the stack is unwound, after C<die> sets C<$@> to
447 C<"foo at Foo.pm line 42\n">, so by the time C<if ( $@ )> is evaluated it has
448 been cleared by C<eval> in the destructor.
450 The workaround for this is even uglier than the previous ones. Even though we
451 can't save the value of C<$@> from code that doesn't localize, we can at least
452 be sure the C<eval> was aborted due to an error:
454 my $failed = not eval {
460 This is because an C<eval> that caught a C<die> will always return a false
465 Using Perl 5.10 you can use L<perlsyn/"Switch statements">.
467 =for stopwords topicalizer
469 The C<catch> block is invoked in a topicalizer context (like a C<given> block),
470 but note that you can't return a useful value from C<catch> using the C<when>
471 blocks without an explicit C<return>.
473 This is somewhat similar to Perl 6's C<CATCH> blocks. You can use it to
474 concisely match errors:
479 when (/^Can't locate .*?\.pm in \@INC/) { } # ignore
489 C<@_> is not available within the C<try> block, so you need to copy your
490 argument list. In case you want to work with argument values directly via C<@_>
491 aliasing (i.e. allow C<$_[1] = "foo">), you need to pass C<@_> by reference:
494 my ( $self, @args ) = @_;
495 try { $self->bar(@args) }
503 try { $_ = $self->bar($_) for @$args }
508 C<return> returns from the C<try> block, not from the parent sub (note that
509 this is also how C<eval> works, but not how L<TryCatch> works):
519 say "this text WILL be displayed, even though an exception is thrown";
522 Instead, you should capture the return value:
529 return unless $success;
531 say "This text WILL NEVER appear!";
534 sub parent_sub_with_catch {
540 # do something with $_
541 return undef; #see note
543 return unless $success;
545 say "This text WILL NEVER appear!";
548 Note that if you have a C<catch> block, it must return C<undef> for this to work,
549 since if a C<catch> block exists, its return value is returned in place of C<undef>
550 when an exception is thrown.
554 C<try> introduces another caller stack frame. L<Sub::Uplevel> is not used. L<Carp>
555 will not report this when using full stack traces, though, because
556 C<%Carp::Internal> is used. This lack of magic is considered a feature.
558 =for stopwords unhygienically
562 The value of C<$_> in the C<catch> block is not guaranteed to be the value of
563 the exception thrown (C<$@>) in the C<try> block. There is no safe way to
564 ensure this, since C<eval> may be used unhygienically in destructors. The only
565 guarantee is that the C<catch> will be called if an exception is thrown.
569 The return value of the C<catch> block is not ignored, so if testing the result
570 of the expression for truth on success, be sure to return a false value from
578 return; # avoid returning a true value;
585 C<$SIG{__DIE__}> is still in effect.
587 Though it can be argued that C<$SIG{__DIE__}> should be disabled inside of
588 C<eval> blocks, since it isn't people have grown to rely on it. Therefore in
589 the interests of compatibility, C<try> does not disable C<$SIG{__DIE__}> for
590 the scope of the error throwing code.
594 Lexical C<$_> may override the one set by C<catch>.
596 For example Perl 5.10's C<given> form uses a lexical C<$_>, creating some
604 warn $_; # will print $foo, not the error
605 warn $_[0]; # instead, get the error like this
610 Note that this behavior was changed once again in L<Perl5 version 18
611 |https://metacpan.org/module/perldelta#given-now-aliases-the-global-_>.
612 However, since the entirety of lexical C<$_> is now L<considered experimental
613 |https://metacpan.org/module/perldelta#Lexical-_-is-now-experimental>, it
614 is unclear whether the new version 18 behavior is final.
624 Much more feature complete, more convenient semantics, but at the cost of
625 implementation complexity.
629 Automatic error throwing for builtin functions and more. Also designed to
630 work well with C<given>/C<when>.
634 A lightweight role for rolling your own exception classes.
638 Exception object implementation with a C<try> statement. Does not localize
641 =item L<Exception::Class::TryCatch>
643 Provides a C<catch> statement, but properly calling C<eval> is your
646 The C<try> keyword pushes C<$@> onto an error stack, avoiding some of the
647 issues with C<$@>, but you still need to localize to prevent clobbering.
651 =head1 LIGHTNING TALK
653 I gave a lightning talk about this module, you can see the slides (Firefox
656 L<http://web.archive.org/web/20100628040134/http://nothingmuch.woobling.org/talks/takahashi.xul>
660 L<http://web.archive.org/web/20100305133605/http://nothingmuch.woobling.org/talks/yapc_asia_2009/try_tiny.yml>
662 =head1 VERSION CONTROL
664 L<http://github.com/doy/try-tiny/>