VERSION

version 0.08

SYNOPSIS

use Parse::Keyword { try => \&try_parser };
use Exporter 'import';
our @EXPORT = 'try';
sub try {
my ($try, $catch) = @_;
&Try::Tiny::try($try, ($catch ? (&Try::Tiny::catch($catch)) : ()));
}
sub try_parser {
lex_read_space;
die "syntax error" unless lex_peek eq '{';
my $try = parse_block;
lex_read_space;
my $catch;
if (lex_peek(6) =~ /^catch\b/) {
lex_read(5);
lex_read_space;
die "syntax error" unless lex_peek eq '{';
$catch = parse_block;
}
return (sub { ($try, $catch) }, 1);
}

DESCRIPTION

DO NOT USE!

This module has fundamental errors in the way it handles closures, which are not fixable. Runtime keywords will never be able to work properly with the current design of this module. There are certain cases where this module is still safe to use (keywords that only have effect at compile time, or keywords that never call any of the "parse_*" functions), but that is limiting enough to make this module mostly worthless, and I likely won't be continuing to maintain it. Be warned!

NOTE: The API of this module is still in flux. I may make backwards-incompatible changes as I figure out how it should look.

This module allows you to write keyword-based syntax extensions without requiring you to write any C code yourself. It is similar to Devel::Declare, except that it uses the Perl parser API introduced in Perl 5.14 in order to allow you to parse parts of things using perl's own parser, rather than having to fake it with balanced brace matching or other fragile things.

To use this module, you should pass a hashref to the "use" statement. The keys of this hashref are subroutines in the current package which should have special parsing behavior attached to them, and the values are coderefs which should be used to implement the custom parsing behavior.

The parsing coderefs will be called when perl encounters a call to the keyword that you attached custom parsing to. The current parser state will be directly after parsing the keyword. The parser function will receive the name of the keyword as a parameter, and should return a coderef which, when called at runtime, will produce the arguments to the function. In addition, if your keyword should be parsed as a statement (for instance, if you don't want to require a trailing semicolon), you can return a second, true value.

In order to actually handle the parsing itself, this module also exports various parsing functions, which you can call. See below for details.

FUNCTIONS

lex_peek($n)

Returns a string consisting of the next $n characters in the input (or next one character, if $n isn't given). This string may be shorter than $n characters if there are fewer than $n characters remaining to read. The current position in the buffer to be parsed is not moved. See ``PL_parser->linestr'' in perlapi and ``lex_next_chunk'' in perlapi for more information.

NOTE: This function currently only returns text that is on the current line, unless the current line has been fully read (via "lex_read"). This is due to a bug in perl itself, and this restriction will hopefully be lifted in a future version of this module, so don't depend on it. See the ``BUGS'' section for more information.

lex_read($n)

Moves the current position in the parsing buffer forward by $n characters (or one character, if $n isn't given). See ``lex_read_to'' in perlapi for more details.

lex_read_space

Moves the current position in the parsing buffer forward past any whitespace or comments. See ``lex_read_space'' in perlapi for more details.

lex_stuff($str)

Inserts $str into the current parsing buffer at the current location, so that future calls to "lex_peek" and such will see it. Note that this does not move the current position in the parsing buffer, so multiple calls to "lex_stuff" at the same location will end up inserted into the buffer in reverse order. See ``lex_stuff_sv'' in perlapi for more information.

parse_block, parse_stmtseq, parse_fullstmt, parse_barestmt, parse_fullexpr, parse_listexpr, parse_termexpr, parse_arithexpr

These functions parse the specified amount of Perl code, and return a coderef which will evaluate that code when executed. They each take an optional boolean parameter that should be true if you are creating a subroutine which will be going in the symbol table, or in other more obscure situations involving closures (the CVf_ANON flag will be set on the created coderef if this is not passed - see "t/unavailable.t" in this distribution). See ``parse_block'' in perlapi, ``parse_stmtseq'' in perlapi, ``parse_fullstmt'' in perlapi, ``parse_barestmt'' in perlapi, ``parse_fullexpr'' in perlapi, parse_listexpr, parse_termexpr, and ``parse_arithexpr'' in perlapi for more details.

compiling_package

Returns the name of the package that the keyword which is currently being parsed was called in. This should be used instead of "caller" if you want to do something like install a subroutine in the calling package.

SUPPORT

You can find this documentation for this module with the perldoc command.

    perldoc Parse::Keyword

You can also look for information at:

• MetaCPAN

  <https://metacpan.org/release/Parse-Keyword>

• RT: CPAN's request tracker

  <http://rt.cpan.org/NoAuth/Bugs.html?Dist=Parse-Keyword>

• Github

  <https://github.com/doy/parse-keyword>

• CPAN Ratings

  <http://cpanratings.perl.org/d/Parse-Keyword>

AUTHOR

Jesse Luehrs <[email protected]>

COPYRIGHT AND LICENSE

This software is Copyright (c) 2013 by Jesse Luehrs.

This is free software, licensed under:

  The MIT (X11) License