path: root/init-repository

#!/usr/bin/env perl
#############################################################################
##
## Copyright (C) 2015 The Qt Company Ltd.
## Contact: http://www.qt.io/licensing/
##
## This file is part of the utilities of the Qt Toolkit.
##
## $QT_BEGIN_LICENSE:LGPL21$
## Commercial License Usage
## Licensees holding valid commercial Qt licenses may use this file in
## accordance with the commercial license agreement provided with the
## Software or, alternatively, in accordance with the terms contained in
## a written agreement between you and The Qt Company. For licensing terms
## and conditions see http://www.qt.io/terms-conditions. For further
## information use the contact form at http://www.qt.io/contact-us.
##
## GNU Lesser General Public License Usage
## Alternatively, this file may be used under the terms of the GNU Lesser
## General Public License version 2.1 or version 3 as published by the Free
## Software Foundation and appearing in the file LICENSE.LGPLv21 and
## LICENSE.LGPLv3 included in the packaging of this file. Please review the
## following information to ensure the GNU Lesser General Public License
## requirements will be met: https://www.gnu.org/licenses/lgpl.html and
## http://www.gnu.org/licenses/old-licenses/lgpl-2.1.html.
##
## As a special exception, The Qt Company gives you certain additional
## rights. These rights are described in The Qt Company LGPL Exception
## version 1.1, included in the file LGPL_EXCEPTION.txt in this package.
##
## $QT_END_LICENSE$
##
#############################################################################

use v5.8;
use strict;
use warnings;

package Qt::InitRepository;


=head1 NAME

init-repository - initialize the Qt5 repository and all submodules

=head1 SYNOPSIS

  ./init-repository [options]

This script may be run after an initial `git clone' of Qt5 in order to check
out all submodules.


=head1 OPTIONS

B<Global options:>

=over

=item --force, -f

Force initialization (even if the submodules are already checked out).


=item --force-hooks

Force initialization of hooks (even if there are already hooks in checked out
submodules).


=item --quiet, -q

Be quiet.  Will exit cleanly if the repository is already initialized.

=back


B<Module options:>

=over

=item --no-webkit

Skip webkit and webkit examples submodules.
It may be desirable to skip these modules due to the large size of the webkit
git repository.

=item --module-subset=<module1>,<module2>...

Only initialize the specified subset of modules given as the argument. Specified
modules must already exist in .gitmodules.
The string "all" results in cloning all known modules. The default is the set of
maintained modules.
Module names may be prefixed with a dash to exclude them from a bigger set.

=item --no-update

Skip the `git submodule update' command.

=item --branch

Instead of checking out specific SHA1s, check out the submodule branches that
correspond with the current supermodule commit.
By default, this option will cause local commits in the submodules to be rebased.
With --no-update, the branches will be checked out, but their heads will not move.

=item --ignore-submodules

Set git config to ignore submodules by default when doing operations on the
qt5 repo, such as `pull', `fetch', `diff' etc.

After using this option, pass `--ignore-submodules=none' to git to override
it as needed.

=back


B<Repository options:>

=over

=item --berlin

Switch to internal URLs and make use of the Berlin git mirrors.
(Implies `--mirror').

=item --oslo

Switch to internal URLs and make use of the Oslo git mirrors.
(Implies `--mirror').


=item --codereview-username <Gerrit/JIRA username>

Specify the user name for the (potentially) writable `gerrit' remote
for each module, for use with the Gerrit code review tool.

If this option is omitted, the gerrit remote is created without a username
and port number, and thus relies on a correct SSH configuration.


=item --alternates <path to other Qt5 repo>

Adds alternates for each submodule to another full qt5 checkout. This makes
this qt5 checkout very small, as it will use the object store of the
alternates before unique objects are stored in its own object store.

This option has no effect when using `--no-update'.

B<NOTE:> This will make this repo dependent on the alternate, which is
potentially dangerous!  The dependency can be broken by also using
the `--copy-objects' option, or by running C<git repack -a> in each
submodule, where required.  Please read the note about the `--shared' option
in the documentation of `git clone' for more information.


=item --copy-objects

When `--alternates' is used, automatically do a C<git repack -a> in each
submodule after cloning, to ensure that the repositories are independent
from the source used as a reference for cloning.

Note that this negates the disk usage benefits gained from the use of
`--alternates'.


=item --mirror <url-base>

Uses <url-base> as the base URL for submodule git mirrors.

For example:

  --mirror user@machine:/foo/bar/

...will use the following as a mirror for qtbase:

  user@machine:/foo/bar/qt/qtbase.git

The mirror is permitted to contain a subset of the submodules; any
missing modules will fall back to the canonical URLs.

=back

=cut

use Carp         qw( confess             );
use English      qw( -no_match_vars      );
use Getopt::Long qw( GetOptionsFromArray );
use Pod::Usage   qw( pod2usage           );
use Cwd          qw( getcwd              );

my $GERRIT_SSH_BASE
    = 'ssh://@USER@codereview.qt-project.org@PORT@/';

my $BER_MIRROR_URL_BASE
    = 'git://hegel/';

my $OSLO_MIRROR_URL_BASE
    = 'git://qilin/';

sub new
{
    my ($class, @arguments) = @_;

    my $self = {};
    bless $self, $class;
    $self->parse_arguments(@arguments);

    return $self;
}

# Like `system', but possibly log the command, and die on non-zero exit code
sub exe
{
    my ($self, @cmd) = @_;

    if (!$self->{quiet}) {
        print "+ @cmd\n";
    }

    if (system(@cmd) != 0) {
        confess "@cmd exited with status $CHILD_ERROR";
    }

    return;
}

sub parse_arguments
{
    my ($self, @args) = @_;

    %{$self} = (%{$self},
        'alternates'          => "",
        'branch'              => 0,
        'codereview-username' => "",
        'detach-alternates'   => 0 ,
        'force'               => 0 ,
        'force-hooks'         => 0 ,
        'ignore-submodules'   => 0 ,
        'mirror-url'          => "",
        'update'              => 1 ,
        'webkit'              => 1 ,
        'module-subset'       => "default",
    );

    GetOptionsFromArray(\@args,
        'alternates=s'      =>  \$self->{qw{ alternates        }},
        'branch'            =>  \$self->{qw{ branch            }},
        'codereview-username=s' => \$self->{qw{ codereview-username }},
        'copy-objects'      =>  \$self->{qw{ detach-alternates }},
        'force|f'           =>  \$self->{qw{ force             }},
        'force-hooks'       =>  \$self->{qw{ force-hooks       }},
        'ignore-submodules' =>  \$self->{qw{ ignore-submodules }},
        'mirror=s'          =>  \$self->{qw{ mirror-url        }},
        'quiet'             =>  \$self->{qw{ quiet             }},
        'update!'           =>  \$self->{qw{ update            }},
        'webkit!'           =>  \$self->{qw{ webkit            }},
        'module-subset=s'   =>  \$self->{qw{ module-subset     }},

        'help|?'            =>  sub { pod2usage(1);               },

        'berlin' => sub {
            $self->{'mirror-url'}        = $BER_MIRROR_URL_BASE;
        },
        'oslo' => sub {
            $self->{'mirror-url'}        = $OSLO_MIRROR_URL_BASE;
        },
    ) || pod2usage(2);

    # Replace any double trailing slashes from end of mirror
    $self->{'mirror-url'} =~ s{//+$}{/};

    $self->{'module-subset'} = [ split(/,/, $self->{'module-subset'}) ];
    if (!$self->{webkit}) {
        push @{$self->{'module-subset'}}, "-qtwebkit", "-qtwebkit-examples";
    }

    return;
}

sub check_if_already_initialized
{
    my ($self) = @_;

    # We consider the repo as `initialized' if submodule.qtbase.url is set
    if (qx(git config --get submodule.qtbase.url)) {
        if (!$self->{force}) {
            exit 0 if ($self->{quiet});
            print "Will not reinitialize already initialized repository (use -f to force)!\n";
            exit 1;
        }
    }

    return;
}

sub git_submodule_init
{
    my ($self, @init_args) = @_;

    if ($self->{quiet}) {
        unshift @init_args, '--quiet';
    }
    $self->exe('git', 'submodule', 'init', @init_args);

    my $template = getcwd()."/.commit-template";
    if (-e $template) {
        $self->exe('git', 'config', 'commit.template', $template);
    }

    return;
}

sub git_clone_all_submodules
{
    my ($self, $my_repo_base, $co_branch, @subset) = @_;

    my %subdirs = ();
    my %subbranches = ();
    my %subbases = ();
    my %subinits = ();
    my @submodconfig = qx(git config -l -f .gitmodules);
    foreach my $line (@submodconfig) {
        # Example line: submodule.qtqa.url=../qtqa.git
        next if ($line !~ /^submodule\.([^.=]+)\.([^.=]+)=(.*)$/);
        if ($2 eq "path") {
            $subdirs{$1} = $3;
        } elsif ($2 eq "branch") {
            $subbranches{$1} = $3;
        } elsif ($2 eq "url") {
            my ($mod, $base) = ($1, $3);
            next if ($base !~ /^\.\.\//);
            $base = $my_repo_base.'/'.$base;
            while ($base =~ s,/(?!\.\./)[^/]+/\.\./,/,g) {}
            $subbases{$mod} = $base;
        } elsif ($2 eq "update") {
            push @subset, '-'.$1 if ($3 eq 'ignore');
        } elsif ($2 eq "initrepo") {
            $subinits{$1} = ($3 eq "yes" or $3 eq "true");
        }
    }

    my %include = ();
    foreach my $mod (@subset) {
        if ($mod eq "all") {
            map { $include{$_} = 1; } keys %subbases;
        } elsif ($mod eq "default") {
            map { $include{$_} = 1; } grep { $subinits{$_} } keys %subbases;
        } elsif ($mod =~ s/^-//) {
            delete $include{$mod};
        } else {
            $include{$mod} = 1;
        }
    }

    my @modules = sort keys %include;

    $self->git_submodule_init(map { $subdirs{$_} } @modules);

    # manually clone each repo here, so we can easily use reference repos, mirrors etc
    my @configresult = qx(git config -l);
    foreach my $line (@configresult) {
        # Example line: submodule.qtqa.url=git://gitorious.org/qt/qtqa.git
        next if ($line !~ /submodule\.([^.=]+)\.url=/);
        my $module = $1;

        if (!defined($include{$module})) {
            $self->exe('git', 'config', '--remove-section', "submodule.$module");
            next;
        }

        if ($self->{'ignore-submodules'}) {
            $self->exe('git', 'config', "submodule.$module.ignore", 'all');
        }
    }

    foreach my $module (@modules) {
        $self->git_clone_one_submodule($subdirs{$module}, $subbases{$module}, $subbranches{$module});
    }

    if ($co_branch) {
        foreach my $module (@modules) {
            my $branch = $subbranches{$module};
            die("No branch defined for submodule $module.\n") if (!defined($branch));
            my $orig_cwd = getcwd();
            chdir($module) or confess "chdir $module: $OS_ERROR";
            my $br = qx(git rev-parse -q --verify $branch);
            if (!$br) {
                $self->exe('git', 'checkout', '-b', $branch, "origin/$branch");
            } else {
                $self->exe('git', 'checkout', $branch);
            }
            chdir("$orig_cwd") or confess "chdir $orig_cwd: $OS_ERROR";
        }
    }
    if ($self->{update}) {
        my @cmd = ('git', 'submodule', 'update', '--no-fetch');
        push @cmd, '--remote', '--rebase' if ($co_branch);
        $self->exe(@cmd);

        foreach my $module (@modules) {
            if (-f $module.'/.gitmodules') {
                my $orig_cwd = getcwd();
                chdir($module) or confess "chdir $module: $OS_ERROR";
                $self->git_clone_all_submodules($subbases{$module}, 0, "all");
                chdir("$orig_cwd") or confess "chdir $orig_cwd: $OS_ERROR";
            }
        }
    }

    return;
}

sub git_add_remotes
{
    my ($self, $gerrit_repo_basename) = @_;

    my $gerrit_repo_url = $GERRIT_SSH_BASE;
    # If given a username, make a "verbose" remote.
    # Otherwise, rely on proper SSH configuration.
    if ($self->{'codereview-username'}) {
        $gerrit_repo_url =~ s,\@USER\@,$self->{'codereview-username'}\@,;
        $gerrit_repo_url =~ s,\@PORT\@,:29418,;
    } else {
        $gerrit_repo_url =~ s,\@[^\@]+\@,,g;
    }

    $gerrit_repo_url .= $gerrit_repo_basename;
    $self->exe('git', 'config', 'remote.gerrit.url', $gerrit_repo_url);
    $self->exe('git', 'config', 'remote.gerrit.fetch', '+refs/heads/*:refs/remotes/gerrit/*', '/heads/');
}

sub git_clone_one_submodule
{
    my ($self, $submodule, $repo_basename, $branch) = @_;

    my $alternates            = $self->{ 'alternates'        };
    my $mirror_url            = $self->{ 'mirror-url'        };
    my $protocol              = $self->{ 'protocol'          };

    # `--reference FOO' args for the clone, if any.
    my @reference_args;

    if ($alternates) {
        # alternates is a qt5 repo, so the submodule will be under that.
        if (-e "$alternates/$submodule/.git") {
            @reference_args = ('--reference', "$alternates/$submodule");
        }
        else {
            print " *** $alternates/$submodule not found, ignoring alternate for this submodule\n";
        }
    }

    my $url = $self->{'base-url'}.$repo_basename;
    my $mirror;
    if ($mirror_url) {
        $mirror = $mirror_url.$repo_basename;
    }

    if ($mirror) {
        # Only use the mirror if it can be reached.
        eval { $self->exe('git', 'ls-remote', $mirror, 'test/if/mirror/exists') };
        if ($@) {
            warn "mirror [$mirror] is not accessible; $url will be used\n";
            undef $mirror;
        }
    }

    my $do_clone = (! -e "$submodule/.git");
    if ($do_clone) {
        push @reference_args, '--branch', $branch if ($branch);
        $self->exe('git', 'clone', @reference_args,
                   ($mirror ? $mirror : $url), $submodule);
    }

    my $orig_cwd = getcwd();
    chdir($submodule) or confess "chdir $submodule: $OS_ERROR";

    if ($mirror) {
        # This is only for the user's convenience - we make no use of it.
        $self->exe('git', 'config', 'remote.mirror.url', $mirror);
        $self->exe('git', 'config', 'remote.mirror.fetch', '+refs/heads/*:refs/remotes/mirror/*');
    }

    if (!$do_clone && $self->{update}) {
        # If we didn't clone, fetch from the right location. We always update
        # the origin remote, so that submodule update --remote works.
        $self->exe('git', 'config', 'remote.origin.url', ($mirror ? $mirror : $url));
        $self->exe('git', 'fetch', 'origin');
    }

    if (!($do_clone || $self->{update}) || $mirror) {
        # Leave the origin configured to the canonical URL. It's already correct
        # if we cloned/fetched without a mirror; otherwise it may be anything.
        $self->exe('git', 'config', 'remote.origin.url', $url);
    }

    my $template = getcwd()."/../.commit-template";
    if (-e $template) {
        $self->exe('git', 'config', 'commit.template', $template);
    }

    $self->git_add_remotes($repo_basename);

    if ($self->{'detach-alternates'}) {
        $self->exe('git', 'repack', '-a');

        my $alternates_path = '.git/objects/info/alternates';
        if (-e $alternates_path) {
            unlink($alternates_path) || confess "unlink $alternates_path: $OS_ERROR";
        }
    }

    chdir($orig_cwd) or confess "cd $orig_cwd: $OS_ERROR";

    return;
}

sub ensure_link
{
    my ($self, $src, $tgt) = @_;
    return if (!$self->{'force-hooks'} and -f $tgt);
    unlink($tgt); # In case we have a dead symlink or pre-existing hook
    print "Aliasing $src\n      as $tgt ...\n" if (!$self->{quiet});
    return if eval { symlink($src, $tgt) };
    # Windows doesn't do (proper) symlinks. As the post_commit script needs
    # them to locate itself, we write a forwarding script instead.
    open SCRIPT, ">".$tgt or die "Cannot create forwarding script $tgt: $!\n";
    print SCRIPT "#!/bin/sh\nexec `dirname \$0`/$src \"\$\@\"\n";
    close SCRIPT;
}

sub git_install_hooks
{
    my ($self) = @_;

    return if (!-d 'qtrepotools/git-hooks');

    # Force C locale as git submodule returns the localized string "Entering"
    local $ENV{LC_ALL} = 'C';
    chomp(my @modules = `git submodule foreach :`);
    push @modules, "";
    for my $module (@modules) {
        $module =~ s,^Entering \'([^\']+)\'$,$1/,;
        my $rel = $module;
        $rel =~ s,[^/]+,..,g;
        $rel .= "../../qtrepotools/git-hooks/";
        $self->ensure_link($rel.'gerrit_commit_msg_hook', $module.'.git/hooks/commit-msg');
        $self->ensure_link($rel.'git_post_commit_hook', $module.'.git/hooks/post-commit');
    }
}

sub run
{
    my ($self) = @_;

    $self->check_if_already_initialized;

    chomp(my $url = `git config remote.origin.url`);
    die("Have no origin remote.\n") if (!$url);
    $url =~ s,\.git$,,;
    $url =~ s,qt/qt5$,,;
    $self->{'base-url'} = $url;

    $self->git_clone_all_submodules('qt/qt5', $self->{branch}, @{$self->{'module-subset'}});

    $self->git_add_remotes('qt/qt5');

    $self->git_install_hooks;

    return;
}

#==============================================================================

Qt::InitRepository->new(@ARGV)->run if (!caller);
1;