=head1 NAME

Mail::Header - manipulate MIME headers

=head1 SYNOPSIS

 use Mail::Header;
    
 my $head = Mail::Header->new;
 my $head = Mail::Header->new( \*STDIN );
 my $head = Mail::Header->new( [<>], Modify => 0);

=head1 DESCRIPTION

Read, write, create, and manipulate MIME headers, the leading part
of each modern e-mail message, but also used in other protocols
like HTTP.  The fields are kept in L<Mail::Field|Mail::Field> objects.

Be aware that the header fields each have a name part, which shall
be treated case-insensitive, and a content part, which may be folded
over multiple lines.  

Mail::Header does not always follow the RFCs strict enough, does not
help you with character encodings.  It does not use weak references
where it could (because those did not exist when the module was written)
which costs some performance and make the implementation a little more
complicated.  The Mail::Message::Head implementation is much newer
and therefore better.

=head1 METHODS

=head2 Constructors

=over 4

=item $obj-E<gt>B<dup>()

Create a duplicate of the current object.

=item $obj-E<gt>B<new>([ARG], [OPTIONS])

=item Mail::Header-E<gt>B<new>([ARG], [OPTIONS])

ARG may be either a file descriptor (reference to a GLOB)
or a reference to an array. If given the new object will be
initialized with headers either from the array of read from 
the file descriptor.

OPTIONS is a list of options given in the form of key-value
pairs, just like a hash table. Valid options are

 -Option    --Default
  FoldLength  79
  MailFrom    'KEEP'
  Modify      false

=over 2

=item FoldLength => INTEGER

The default length of line to be used when folding header lines.
See L<fold_length()|Mail::Header/"Accessors">.

=item MailFrom => 'IGNORE'|'COERCE'|'KEEP'|'ERROR'

See method L<mail_from()|Mail::Header/"Accessors">.

=item Modify => BOOLEAN

If this value is I<true> then the headers will be re-formatted,
otherwise the format of the header lines will remain unchanged.

=back

=back

=head2 "Fake" constructors

Be warned that the next constructors all require an already created
header object, of which the original content will be destroyed.

=over 4

=item $obj-E<gt>B<empty>()

Empty an existing C<Mail::Header> object of all lines.

=item $obj-E<gt>B<extract>(ARRAY)

Extract a header from the given array into an existing Mail::Header
object. C<extract> B<will modify> this array.
Returns the object that the method was called on.

=item $obj-E<gt>B<header>([ARRAY])

C<header> does multiple operations. First it will extract a header from
the ARRAY, if given. It will then reformat the header (if reformatting
is permitted), and finally return a reference to an array which
contains the header in a printable form.

=item $obj-E<gt>B<header_hashref>([HASH])

As L<header()|Mail::Header/""Fake" constructors">, but it will eventually set headers from a hash
reference, and it will return the headers as a hash reference.

example: 

 $fields->{From} = 'Tobias Brox <tobix@cpan.org>';
 $fields->{To}   = ['you@somewhere', 'me@localhost'];
 $head->header_hashref($fields);

=item $obj-E<gt>B<read>(FILEHANDLE)

Read a header from the given file descriptor into an existing Mail::Header
object.

=back

=head2 Accessors

=over 4

=item $obj-E<gt>B<fold_length>([TAG], [LENGTH])

Set the default fold length for all tags or just one. With no arguments
the default fold length is returned. With two arguments it sets the fold
length for the given tag and returns the previous value. If only C<LENGTH>
is given it sets the default fold length for the current object.

In the two argument form C<fold_length> may be called as a static method,
setting default fold lengths for tags that will be used by B<all>
C<Mail::Header> objects. See the C<fold> method for
a description on how C<Mail::Header> uses these values.

=item $obj-E<gt>B<mail_from>('IGNORE'|'COERCE'|'KEEP'|'ERROR')

This specifies what to do when a C<`From '> line is encountered.
Valid values are C<IGNORE> - ignore and discard the header,
C<ERROR> - invoke an error (call die), C<COERCE> - rename them as Mail-From
and C<KEEP> - keep them.

=item $obj-E<gt>B<modify>([VALUE])

If C<VALUE> is I<false> then C<Mail::Header> will not do any automatic
reformatting of the headers, other than to ensure that the line
starts with the tags given.

=back

=head2 Processing

=over 4

=item $obj-E<gt>B<add>(TAG, LINE [, INDEX])

Add a new line to the header. If TAG is C<undef> the the tag will be
extracted from the beginning of the given line. If INDEX is given,
the new line will be inserted into the header at the given point, otherwise
the new line will be appended to the end of the header.

=item $obj-E<gt>B<as_string>()

Returns the header as a single string.

=item $obj-E<gt>B<cleanup>()

Remove any header line that, other than the tag, only contains whitespace

=item $obj-E<gt>B<combine>(TAG [, WITH])

Combine all instances of TAG into one. The lines will be
joined together WITH, or a single space if not given. The new
item will be positioned in the header where the first instance was, all
other instances of TAG will be removed.

=item $obj-E<gt>B<count>(TAG)

Returns the number of times the given atg appears in the header

=item $obj-E<gt>B<delete>(TAG [, INDEX ])

Delete a tag from the header. If an INDEX id is given, then the Nth instance
of the tag will be removed. If no INDEX is given, then all instances
of tag will be removed.

=item $obj-E<gt>B<fold>([LENGTH])

Fold the header. If LENGTH is not given, then C<Mail::Header> uses the
following rules to determine what length to fold a line.

=item $obj-E<gt>B<get>(TAG [, INDEX])

Get the text from a line. If an INDEX is given, then the text of the Nth
instance will be returned. If it is not given the return value depends on the
context in which C<get> was called. In an array context a list of all the
text from all the instances of the TAG will be returned. In a scalar context
the text for the first instance will be returned.

The lines are unfolded, but still terminated with a new-line (see C<chomp>)

=item $obj-E<gt>B<print>([FILEHANDLE])

Print the header to the given file descriptor, or C<STDOUT> if no
file descriptor is given.

=item $obj-E<gt>B<replace>(TAG, LINE [, INDEX ])

Replace a line in the header.  If TAG is C<undef> the the tag will be
extracted from the beginning of the given line. If INDEX is given
the new line will replace the Nth instance of that tag, otherwise the
first instance of the tag is replaced. If the tag does not appear in the
header then a new line will be appended to the header.

=item $obj-E<gt>B<tags>()

Returns an array of all the tags that exist in the header. Each tag will
only appear in the list once. The order of the tags is not specified.

=item $obj-E<gt>B<unfold>([TAG])

Unfold all instances of the given tag so that they do not spread across
multiple lines. If C<TAG> is not given then all lines are unfolded.

The unfolding process is wrong but (for compatibility reasons) will
not be repaired: only one blank at the start of the line should be
removed, not all of them.

=back

=head1 SEE ALSO

This module is part of the MailTools distribution,
F<http://perl.overmeer.net/mailtools/>.

=head1 AUTHORS

The MailTools bundle was developed by Graham Barr.  Later, Mark
Overmeer took over maintenance without commitment to further development.

Mail::Cap by Gisle Aas E<lt>aas@oslonett.noE<gt>.
Mail::Field::AddrList by Peter Orbaek E<lt>poe@cit.dkE<gt>.
Mail::Mailer and Mail::Send by Tim Bunce E<lt>Tim.Bunce@ig.co.ukE<gt>.
For other contributors see ChangeLog.

=head1 LICENSE

Copyrights 1995-2000 Graham Barr E<lt>gbarr@pobox.comE<gt> and
2001-2007 Mark Overmeer E<lt>perl@overmeer.netE<gt>.

This program is free software; you can redistribute it and/or modify it
under the same terms as Perl itself.
See F<http://www.perl.com/perl/misc/Artistic.html>