# # Copyright (c) 2003,2004,2005 Alexander Taler (dissent@0--0.org) # # All rights reserved. This program is free software; you can redistribute it # and/or modify it under the same terms as Perl itself. # package VCS::LibCVS::Datum; use strict; use Carp; use VCS::LibCVS::Datum::DirectoryName; use VCS::LibCVS::Datum::Entry; use VCS::LibCVS::Datum::FileContents; use VCS::LibCVS::Datum::FileMode; use VCS::LibCVS::Datum::FileName; use VCS::LibCVS::Datum::LogMessage; use VCS::LibCVS::Datum::PathName; use VCS::LibCVS::Datum::RevisionNumber; use VCS::LibCVS::Datum::Root; use VCS::LibCVS::Datum::String; use VCS::LibCVS::Datum::TagSpec; use VCS::LibCVS::Datum::Time; =head1 NAME VCS::LibCVS::Datum - a piece of CVS data =head1 SYNOPSIS $mode = VCS::LibCVS::Datum::Mode->new("u=rw,g=rw,o=r"); $rq = VCS::LibCVS::Client::Request::Modified->new(["file",$mode,$file]); =head1 DESCRIPTION Datum represents a single piece of CVS data, such as an entries line, a file mode or a tag spec. It has subclasses for each type of CVS data. Datum should not to be instantiated, instead its subclasses should be instantiated. To access the contained data see the accessors for each subclass. =cut # Datum provides some generic facilities to help the implementation of its # simpler subclasses; those which consist of a predetermined number of lines # which can be named. To use these facilities a subclass of Datum overrides # the _data_names() routine, which returns a list of names used as keys into # the private hash. All the routines defined in Datum use this facility. ############################################################################### # Class constants ############################################################################### use constant REVISION => '$Header: /cvsroot/libcvs-perl/libcvs-perl/VCS/LibCVS/Datum.pm,v 1.14 2005/10/10 12:52:11 dissent Exp $ '; ############################################################################### # Private variables ############################################################################### # These are specified by _data_names(). # Each of them is a single line of text, without a terminating newline. ############################################################################### # Class routines ############################################################################### =head1 CLASS ROUTINES =head2 B $data = Datum::Subclass->new($data) Only call this on subclasses of Datum. Some subclasses override this constructor. =over 4 =item return type: Datum::Subclass =item argument 1 type: . . . =over 2 =item E<32>E<32>option 1: IO::Handle An IO::Handle object from which the Datum will be read. Most Datum are line oriented, so they will read one or more lines from the IO::handle. =item E<32>E<32>option 2: scalar If the Datum consists of a single line, it can be passed as a scalar. =item E<32>E<32>option 3: \@array ref If the Datum consists of one or more lines, they can be passed as an array ref. =item E<32>E<32>option 4: Datum::Subclass If the argument is an object of the type which is being constructed, the argument itself will be returned. It will I be copied. This is used by the Client::Request constructor which blindly passes its args to the Datum constructor. That way the user has the freedom to construct their own Datum and pass it to the Client::Request constructor. =back =back Construct a new Datum. The IO::Handle option is used for reading a Datum from the server, the others when constructing it locally. =cut sub new { my ($class, $arg_data) = @_; # if the argument is already an object of the right type, just return it return $arg_data if (UNIVERSAL::isa($arg_data, $class)); my $that = bless {}, $class; # if the argument is an iohandle, read in the right number of lines, and # overwrite $arg_data with that. if (UNIVERSAL::isa($arg_data, "IO::Handle")) { my @arg_data = map { $arg_data->getline(); } $that->_data_names(); $arg_data = \@arg_data; } # a scalar is good if one line is needed, otherwise expect an array ref if (!ref($arg_data)) { confess "Datum needs multiple lines" if (@{[$that->_data_names]} != 1); $that->_add_line(($that->_data_names)[0], $arg_data); } elsif (ref($arg_data) eq "ARRAY") { confess "Wrong number of lines" if (@{[$that->_data_names]} != @$arg_data); foreach my $n ($that->_data_names) {$that->_add_line($n, shift @$arg_data)} } else { confess "Wrong type for data of Arg:" . ref($arg_data); } return $that; } ############################################################################### # Instance routines ############################################################################### =head1 INSTANCE ROUTINES =head2 B $datum_string = $datum->as_string() =over 4 =item return type: string scalar =back Returns the Datum as a string. =cut sub as_string { my $self = shift; my $as_string = $self->as_protocol_string(); chomp $as_string; return $as_string; } =head2 B $datum->protocol_print($file_handle) =over 4 =item return type: undef =item argument 1 type: IO::Handle =back Prints the Datum to the IO::Handle. The output will be formatted for sending to the cvs server, including the placement of newlines. =cut sub protocol_print { my ($self, $ioh) = @_; $ioh->print($self->as_protocol_string()); } =head2 B $datum_string = $datum->as_protocol_string() =over 4 =item return type: string scalar =back Returns the Datum as a string suitable for being sent to the server, including the placement of newlines. =cut sub as_protocol_string { my $self = shift; my $p_string = ""; foreach my $name ($self->_data_names) { $p_string .= $self->{$name} . "\n"; } return $p_string; } =head2 B if ($datum1->equals($datum2)) { =over 4 =item return type: boolean =item argument 1 type: VCS::LibCVS::Datum =back Returns true if the data contain the same information (and are of the same type) =cut sub equals { my $self = shift; my $other = shift; return 0 if ref($self) ne ref($other); foreach my $key ($self->_data_names()) { return 0 if $self->{$key} ne $other->{$key}; } return 1; } ############################################################################### # Private routines ############################################################################### # Returns a list of strings, which are names of the keys in the $self hash # which hold the lines of the Datum. If the Datum is something other than a # list of lines, the empty list will be returned. sub _data_names { return (); } # Add a single named line to the internal hash, ensuring that it contains no # newlines sub _add_line { my ($self, $name, $line) = @_; chomp($line); confess "Embedded newline in Datum line: $name: $line" if $line =~ /\n/; return $self->{$name} = $line; } =head1 SEE ALSO VCS::LibCVS::Datum::DirectoryName VCS::LibCVS::Datum::Entry VCS::LibCVS::Datum::FileContents VCS::LibCVS::Datum::FileMode VCS::LibCVS::Datum::FileName VCS::LibCVS::Datum::PathName VCS::LibCVS::Datum::Root VCS::LibCVS::Datum::String VCS::LibCVS::Datum::TagSpec VCS::LibCVS::Datum::Time =cut 1;