package Bigtop; use strict; use warnings; use warnings::register; use Carp; use File::Spec; our $VERSION = '0.16'; sub write_file { my $file_name = shift; my $content = shift; my $no_overwrite = shift || 0; if ( $no_overwrite and -e $file_name ) { if ( warnings::enabled() ) { warnings::warn( "$file_name already exists, skipping it.\n" ); } return; } my $FILE; unless ( open $FILE, '>', $file_name ) { croak "Couldn't write to $file_name: $!\n"; } print $FILE $content; unless( close $FILE ) { croak "Trouble closing $file_name: $!\n"; } } sub make_module_path { my $module_dir = shift; my $module_name = shift; my @sub_dirs = split /::/, $module_name; # loop through subdirs, so we can make all the intervening paths foreach my $subdir ( 'lib', @sub_dirs ) { $module_dir = File::Spec->catdir( $module_dir, $subdir ); mkdir $module_dir; # mkdir can fail for two reasons, either the dir already exists, # in which case we don't care. Or, the directory could not be # made, then caller will notice when trying to write files to it } if ( wantarray ) { return $module_dir, @sub_dirs; } else { $module_dir; } } =begin NICE_TRY sub setup_template { my $class = shift; my $template_text = shift || $class->get_default_template(); return if ( $class->template_is_setup() ); warn "about to bind for $class"; eval { package $class; Inline->bind( TT => $template_text, POST_CHOMP => 1, TRIM_LEADING_SPACE => 0, TRIM_TRAILING_SPACE => 0, ); }; die $@ if $@; $class->template_is_setup( 1 ); } =end NICE_TRY =cut 1; __END__ =head1 NAME Bigtop - A web application data language processor =head1 SYNOPSIS See Bigtop::Docs::TentTut or Bigtop::Docs::Tutorial for how to create a Bigtop file. The real synopsis: vi your_app.bigtop (or use tentmaker see Bigtop::Docs::TentTut) bigtop --create your_app.bigtop all Modify your bigtop file and try again: bigtop docs/your_app.bigtop all =head1 DESCRIPTION Bigtop is a language for describing the data of a web application. Usually this data will be stored in a relational database. Once you have a description of your data, you can generate a web application from it. This includes all the pieces you need like: the sql statements ready for feeding to your database command line tool, the httpd.conf you need to Include in the httpd.conf on your system, the modules that will handle the web requests, the models that make the database tables look like classes, etc. If you need to alter the data model in the future, you can change your original description to match the new reality, then regenerate the application without fear of losing hand written code (though you may have to modify some of it to reflect the new reality). =head1 FUNCTIONS This module is really a place holder, but it does provide some developer routines (which are not exported): =over 4 =item write_file use Bigtop; Bigtop::write_file( $file_name, $file_content, $no_overwrite ) This attempts to write $file_content to $file_name and dies on failures of open or close. Further, if you pass a true no_overwrite flag, it will check to see if the file exists and refuse to overwrite it. In that case, the user gets a warning that the file has been skipped because it already exists. Since write_file dies on failures, you should wrap it in an eval if you don't want its errors to be fatal. You may also say no warnings in the eval block to prevent skipped messages when you ask for no overwrite. =item make_module_path (Note that make_module_path uses File::Spec, so even though Unix directory syntax is shown below, the function should work in other places.) use Bigtop; Bigtop::make_module_path( $build_dir, $module_name ); This attempts to make all the directories from $build_dir to the home of the module. It assumes that lib comes immediately after $build_dir. For example, a call like: Bigtop::make_module_path( '/home/username/App-Name', 'App::Name::Subname' ); Attempts to make these directories: /home/username/App-Name/lib /home/username/App-Name/lib/App /home/username/App-Name/lib/App/Name /home/username/App-Name/lib/App/Name/Subname It doesn't report failures. Making directories can fail because the directories already exist (in which case you probably don't care) or because they could not be written (in which case you'll notice soon enough, when you try to write to them). =back =head2 EXPORT None. =head1 SEE ALSO The work of Bigtop is handled by its pieces: Bigtop::Parser Bigtop::Init Bigtop::Init::Std Bigtop::SQL Bigtop::SQL::Postgres Bigtop::CGI Bigtop::CGI::Gantry Bigtop::Control Bigtop::Control::Gantry Bigtop::HttpdConf Bigtop::HttpdConf::Gantry Bigtop::Model Bigtop::Model::GantryCDBI Bigtop::Model::Gantry Bigtop::SiteLook::GantryDefault The backends come in types. Ideally, these types all share a set of keywords which are defined in the type's module. So C<Bigtop::SQL> is meant to define the KEYWORDS that all Bigtop::SQL::* modules use. They may define others, but only if they are specific to the generated target. For example, there might be some Postgres specific keyword which doesn't apply to other databases. It should be defined in Bigtop::SQL::Postgres. (But currently only on Bigtop::Control::Gantry, of the backend implementations, actually defines target specifc keywords). =head1 JOIN US Bigtop is discussed on the Gantry mailing list. Please visit http://www.usegantry.org, and click on the Mailing List tab under the banner, for instructions. Bigtop source is available for svn checkout: svn checkout http://svn.usegantry.org/repo/bigtop/trunk/ =head1 AUTHOR Phil Crow, E<lt>philcrow2000@yahoo.comE<gt> =head1 COPYRIGHT AND LICENSE Copyright (C) 2005-6, Phil Crow This library is free software; you can redistribute it and/or modify it under the same terms as Perl itself, either Perl version 5.8.6 or, at your option, any later version of Perl 5 you may have available. =cut