NAME
   Devel::Declare - Adding keywords to perl, in perl

SYNOPSIS
     use Method::Signatures;
     # or ...
     use MooseX::Declare;
     # etc.

     # Use some new and exciting syntax like:
     method hello (Str :$who, Int :$age where { $_ > 0 }) {
       $self->say("Hello ${who}, I am ${age} years old!");
     }

DESCRIPTION
   Devel::Declare can install subroutines called declarators which locally
   take over Perl's parser, allowing the creation of new syntax.

   This document describes how to create a simple declarator.

USAGE
   We'll demonstrate the usage of "Devel::Declare" with a motivating
   example: a new "method" keyword, which acts like the builtin "sub", but
   automatically unpacks $self and the other arguments.

     package My::Methods;
     use Devel::Declare;

 Creating a declarator with "setup_for"
   You will typically create

     sub import {
       my $class = shift;
       my $caller = caller;

       Devel::Declare->setup_for(
           $caller,
           { method => { const => \&parser } }
       );
       no strict 'refs';
       *{$caller.'::method'} = sub (&) {};
     }

   Starting from the end of this import routine, you'll see that we're
   creating a subroutine called "method" in the caller's namespace. Yes,
   that's just a normal subroutine, and it does nothing at all (yet!) Note
   the prototype "(&)" which means that the caller would call it like so:

       method {
           my ($self, $arg1, $arg2) = @_;
           ...
       }

   However we want to be able to call it like this

       method foo ($arg1, $arg2) {
           ...
       }

   That's why we call "setup_for" above, to register the declarator
   'method' with a custom parser, as per the next section. It acts on an
   optype, usually 'const' as above. (Other valid values are 'check' and
   'rv2cv').

   For a simpler way to install new methods, see also
   Devel::Declare::MethodInstaller::Simple

 Writing a parser subroutine
   This subroutine is called at *compilation* time, and allows you to read
   the custom syntaxes that we want (in a syntax that may or may not be
   valid core Perl 5) and munge it so that the result will be parsed by the
   "perl" compiler.

   For this example, we're defining some globals for convenience:

       our ($Declarator, $Offset);

   Then we define a parser subroutine to handle our declarator. We'll look
   at this in a few chunks.

       sub parser {
         local ($Declarator, $Offset) = @_;

   "Devel::Declare" provides some very low level utility methods to parse
   character strings. We'll define some useful higher level routines below
   for convenience, and we can use these to parse the various elements in
   our new syntax.

   Notice how our parser subroutine is invoked at compile time, when the
   "perl" parser is pointed just *before* the declarator name.

         skip_declarator;          # step past 'method'
         my $name = strip_name;    # strip out the name 'foo', if present
         my $proto = strip_proto;  # strip out the prototype '($arg1, $arg2)', if present

   Now we can prepare some code to 'inject' into the new subroutine. For
   example we might want the method as above to have "my ($self, $arg1,
   $arg2) = @_" injected at the beginning of it. We also do some clever
   stuff with scopes that we'll look at shortly.

         my $inject = make_proto_unwrap($proto);
         if (defined $name) {
           $inject = scope_injector_call().$inject;
         }
         inject_if_block($inject);

   We've now managed to change "method ($arg1, $arg2) { ... }" into "method
   { injected_code; ... }". This will compile... but we've lost the name of
   the method!

   In a cute (or horrifying, depending on your perspective) trick, we
   temporarily change the definition of the subroutine "method" itself, to
   specialise it with the $name we stripped, so that it assigns the code
   block to that name.

   Even though the *next* time "method" is compiled, it will be redefined
   again, "perl" caches these definitions in its parse tree, so we'll
   always get the right one!

   Note that we also handle the case where there was no name, allowing an
   anonymous method analogous to an anonymous subroutine.

         if (defined $name) {
           $name = join('::', Devel::Declare::get_curstash_name(), $name)
             unless ($name =~ /::/);
           shadow(sub (&) { no strict 'refs'; *{$name} = shift; });
         } else {
           shadow(sub (&) { shift });
         }
       }

 Parser utilities in detail
   For simplicity, we're using global variables like $Offset in these
   examples. You may prefer to look at Devel::Declare::Context::Simple,
   which encapsulates the context much more cleanly.

  "skip_declarator"
   This simple parser just moves across a 'token'. The common case is to
   skip the declarator, i.e. to move to the end of the string 'method' and
   before the prototype and code block.

       sub skip_declarator {
         $Offset += Devel::Declare::toke_move_past_token($Offset);
       }

  "toke_move_past_token"
   This builtin parser simply moves past a 'token' (matching
   "/[a-zA-Z_]\w*/") It takes an offset into the source document, and skips
   past the token. It returns the number of characters skipped.

  "strip_name"
   This parser skips any whitespace, then scans the next word (again
   matching a 'token'). We can then analyse the current line, and
   manipulate it (using pure Perl). In this case we take the name of the
   method out, and return it.

       sub strip_name {
         skipspace;
         if (my $len = Devel::Declare::toke_scan_word($Offset, 1)) {
           my $linestr = Devel::Declare::get_linestr();
           my $name = substr($linestr, $Offset, $len);
           substr($linestr, $Offset, $len) = '';
           Devel::Declare::set_linestr($linestr);
           return $name;
         }
         return;
       }

  "toke_scan_word"
   This builtin parser, given an offset into the source document, matches a
   'token' as above but does not skip. It returns the length of the token
   matched, if any.

  "get_linestr"
   This builtin returns the full text of the current line of the source
   document.

  "set_linestr"
   This builtin sets the full text of the current line of the source
   document. Beware that injecting a newline into the middle of the line is
   likely to fail in surprising ways. Generally, Perl's parser can rely on
   the `current line' actually being only a single line. Use other kinds of
   whitespace instead, in the code that you inject.

  "skipspace"
   This parser skips whitsepace.

       sub skipspace {
         $Offset += Devel::Declare::toke_skipspace($Offset);
       }

  "toke_skipspace"
   This builtin parser, given an offset into the source document, skips
   over any whitespace, and returns the number of characters skipped.

  "strip_proto"
   This is a more complex parser that checks if it's found something that
   starts with '(' and returns everything till the matching ')'.

       sub strip_proto {
         skipspace;

         my $linestr = Devel::Declare::get_linestr();
         if (substr($linestr, $Offset, 1) eq '(') {
           my $length = Devel::Declare::toke_scan_str($Offset);
           my $proto = Devel::Declare::get_lex_stuff();
           Devel::Declare::clear_lex_stuff();
           $linestr = Devel::Declare::get_linestr();
           substr($linestr, $Offset, $length) = '';
           Devel::Declare::set_linestr($linestr);
           return $proto;
         }
         return;
       }

  "toke_scan_str"
   This builtin parser uses Perl's own parsing routines to match a
   "stringlike" expression. Handily, this includes bracketed expressions
   (just think about things like "q(this is a quote)").

   Also it Does The Right Thing with nested delimiters (like "q(this (is
   (a) quote))").

   It returns the effective length of the expression matched. Really, what
   it returns is the difference in position between where the string
   started, within the buffer, and where it finished. If the string
   extended across multiple lines then the contents of the buffer may have
   been completely replaced by the new lines, so this position difference
   is not the same thing as the actual length of the expression matched.
   However, because moving backward in the buffer causes problems, the
   function arranges for the effective length to always be positive,
   padding the start of the buffer if necessary.

   Use "get_lex_stuff" to get the actual matched text, the content of the
   string. Because of the behaviour around multiline strings, you can't
   reliably get this from the buffer. In fact, after the function returns,
   you can't rely on any content of the buffer preceding the end of the
   string.

   If the string being scanned is not well formed (has no closing
   delimiter), "toke_scan_str" returns "undef". In this case you cannot
   rely on the contents of the buffer.

  "get_lex_stuff"
   This builtin returns what was matched by "toke_scan_str". To avoid
   segfaults, you should call "clear_lex_stuff" immediately afterwards.

 Munging the subroutine
   Let's look at what we need to do in detail.

  "make_proto_unwrap"
   We may have defined our method in different ways, which will result in a
   different value for our prototype, as parsed above. For example:

       method foo         {  # undefined
       method foo ()      {  # ''
       method foo ($arg1) {  # '$arg1'

   We deal with them as follows, and return the appropriate "my ($self,
   ...) = @_;" string.

       sub make_proto_unwrap {
         my ($proto) = @_;
         my $inject = 'my ($self';
         if (defined $proto) {
           $inject .= ", $proto" if length($proto);
           $inject .= ') = @_; ';
         } else {
           $inject .= ') = shift;';
         }
         return $inject;
       }

  "inject_if_block"
   Now we need to inject it after the opening '{' of the method body. We
   can do this with the building blocks we defined above like "skipspace"
   and "get_linestr".

       sub inject_if_block {
         my $inject = shift;
         skipspace;
         my $linestr = Devel::Declare::get_linestr;
         if (substr($linestr, $Offset, 1) eq '{') {
           substr($linestr, $Offset+1, 0) = $inject;
           Devel::Declare::set_linestr($linestr);
         }
       }

  "scope_injector_call"
   We want to be able to handle both named and anonymous methods. i.e.

       method foo () { ... }
       my $meth = method () { ... };

   These will then get rewritten as

       method { ... }
       my $meth = method { ... };

   where 'method' is a subroutine that takes a code block. Spot the
   problem? The first one doesn't have a semicolon at the end of it! Unlike
   'sub' which is a builtin, this is just a normal statement, so we need to
   terminate it. Luckily, using "B::Hooks::EndOfScope", we can do this!

     use B::Hooks::EndOfScope;

   We'll add this to what gets 'injected' at the beginning of the method
   source.

     sub scope_injector_call {
       return ' BEGIN { MethodHandlers::inject_scope }; ';
     }

   So at the beginning of every method, we are passing a callback that will
   get invoked at the *end* of the method's compilation... i.e. exactly
   then the closing '}' is compiled.

     sub inject_scope {
       on_scope_end {
         my $linestr = Devel::Declare::get_linestr;
         my $offset = Devel::Declare::get_linestr_offset;
         substr($linestr, $offset, 0) = ';';
         Devel::Declare::set_linestr($linestr);
       };
     }

 Shadowing each method.
  "shadow"
   We override the current definition of 'method' using "shadow".

       sub shadow {
         my $pack = Devel::Declare::get_curstash_name;
         Devel::Declare::shadow_sub("${pack}::${Declarator}", $_[0]);
       }

   For a named method we invoked like this:

       shadow(sub (&) { no strict 'refs'; *{$name} = shift; });

   So in the case of a "method foo { ... }", this call would redefine
   "method" to be a subroutine that exports 'sub foo' as the (munged)
   contents of "{...}".

   The case of an anonymous method is also cute:

       shadow(sub (&) { shift });

   This means that

       my $meth = method () { ... };

   is rewritten with "method" taking the codeblock, and returning it as is
   to become the value of $meth.

  "get_curstash_name"
   This returns the package name *currently being compiled*.

  "shadow_sub"
   Handles the details of redefining the subroutine.

SEE ALSO
   One of the best ways to learn "Devel::Declare" is still to look at
   modules that use it:

   <http://cpants.perl.org/dist/used_by/Devel-Declare>.

AUTHORS
   Matt S Trout - <[email protected]> - original author

   Company: http://www.shadowcat.co.uk/ Blog: http://chainsawblues.vox.com/

   Florian Ragwitz <[email protected]> - maintainer

   osfameron <[email protected]> - first draft of documentation

COPYRIGHT AND LICENSE
   This library is free software under the same terms as perl itself

   Copyright (c) 2007, 2008, 2009 Matt S Trout

   Copyright (c) 2008, 2009 Florian Ragwitz

   stolen_chunk_of_toke.c based on toke.c from the perl core, which is

   Copyright (C) 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999,
   2000, 2001, 2002, 2003, 2004, 2005, 2006, by Larry Wall and others

You are viewing proxied material from ftp.icm.edu.pl. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.