# This tool is copyright (c) 2005, Martin Langhoff.
# It is released under the Gnu Public License, version 2.
#
-# The basic idea is to walk the output of tla abrowse,
-# fetch the changesets and apply them.
+# The basic idea is to walk the output of tla abrowse,
+# fetch the changesets and apply them.
#
=head1 Invocation
- git-archimport [ -h ] [ -v ] [ -o ] [ -a ] [ -f ] [ -T ]
- [ -D depth] [ -t tempdir ] <archive>/<branch> [ <archive>/<branch> ]
+ git archimport [ -h ] [ -v ] [ -o ] [ -a ] [ -f ] [ -T ]
+ [ -D depth] [ -t tempdir ] <archive>/<branch> [ <archive>/<branch> ]
Imports a project from one or more Arch repositories. It will follow branches
and repositories within the namespaces defined by the <archive/branch>
parameters supplied. If it cannot find the remote branch a merge comes from
-it will just import it as a regular commit. If it can find it, it will mark it
+it will just import it as a regular commit. If it can find it, it will mark it
as a merge whenever possible.
See man (1) git-archimport for more details.
- create tag objects instead of ref tags
- audit shell-escaping of filenames
- hide our private tags somewhere smarter
- - find a way to make "cat *patches | patch" safe even when patchfiles are missing newlines
+ - find a way to make "cat *patches | patch" safe even when patchfiles are missing newlines
- sort and apply patches by graphing ancestry relations instead of just
relying in dates supplied in the changeset itself.
tla ancestry-graph -m could be helpful here...
=head1 Devel tricks
-Add print in front of the shell commands invoked via backticks.
+Add print in front of the shell commands invoked via backticks.
=head1 Devel Notes
sub usage() {
print STDERR <<END;
-Usage: ${\basename $0} # fetch/update GIT from Arch
+Usage: git archimport # fetch/update GIT from Arch
[ -h ] [ -v ] [ -o ] [ -a ] [ -f ] [ -T ] [ -D depth ] [ -t tempdir ]
repository/arch-branch [ repository/arch-branch] ...
END
# values associated with keys:
# =1 - Arch version / git 'branch' detected via abrowse on a limit
# >1 - Arch version / git 'branch' of an auxiliary branch we've merged
-my %arch_branches = map { $_ => 1 } @ARGV;
+my %arch_branches = map { my $branch = $_; $branch =~ s/:[^:]*$//; $branch => 1 } @ARGV;
+
+# $branch_name_map:
+# maps arch branches to git branch names
+my %branch_name_map = map { m/^(.*):([^:]*)$/; $1 => $2 } grep { m/:/ } @ARGV;
$ENV{'TMPDIR'} = $opt_t if $opt_t; # $ENV{TMPDIR} will affect tempdir() calls:
my $tmp = tempdir('git-archimport-XXXXXX', TMPDIR => 1, CLEANUP => 1);
closedir DIR
}
+my $default_archive; # default Arch archive
my %reachable = (); # Arch repositories we can access
my %unreachable = (); # Arch repositories we can't access :<
my @psets = (); # the collection
my $stage = shift;
while (my ($limit, $level) = each %arch_branches) {
next unless $level == $stage;
-
- open ABROWSE, "$TLA abrowse -fkD --merges $limit |"
+
+ open ABROWSE, "$TLA abrowse -fkD --merges $limit |"
or die "Problems with tla abrowse: $!";
-
+
my %ps = (); # the current one
my $lastseen = '';
-
+
while (<ABROWSE>) {
chomp;
-
+
# first record padded w 8 spaces
if (s/^\s{8}\b//) {
my ($id, $type) = split(m/\s+/, $_, 2);
push (@psets, \%last_ps);
$psets{ $last_ps{id} } = \%last_ps;
}
-
+
my $branch = extract_versionname($id);
%ps = ( id => $id, branch => $branch );
if (%last_ps && ($last_ps{branch} eq $branch)) {
$ps{parent_id} = $last_ps{id};
}
-
+
$arch_branches{$branch} = 1;
$lastseen = 'id';
$ps{type} = 't';
# read which revision we've tagged when we parse the log
$ps{tag} = $1;
- } else {
+ } else {
warn "Unknown type $type";
}
$arch_branches{$branch} = 1;
$lastseen = 'id';
- } elsif (s/^\s{10}//) {
- # 10 leading spaces or more
+ } elsif (s/^\s{10}//) {
+ # 10 leading spaces or more
# indicate commit metadata
-
+
# date
if ($lastseen eq 'id' && m/^(\d{4}-\d\d-\d\d \d\d:\d\d:\d\d)/){
$ps{date} = $1;
} elsif ($lastseen eq 'merges' && s/^\s{2}//) {
my $id = $_;
push (@{$ps{merges}}, $id);
-
+
# aggressive branch finding:
if ($opt_D) {
my $branch = extract_versionname($id);
my $repo = extract_reponame($branch);
-
+
if (archive_reachable($repo) &&
!defined $arch_branches{$branch}) {
$arch_branches{$branch} = $stage + 1;
if (@psets && $psets[$#psets]{branch} eq $ps{branch}) {
$temp{parent_id} = $psets[$#psets]{id};
}
- push (@psets, \%temp);
+ push (@psets, \%temp);
$psets{ $temp{id} } = \%temp;
- }
-
+ }
+
close ABROWSE or die "$TLA abrowse failed on $limit\n";
}
} # end foreach $root
while (my $file = readdir(DIR)) {
# skip non-interesting-files
next unless -f "$ptag_dir/$file";
-
+
# convert first '--' to '/' from old git-archimport to use
# as an archivename/c--b--v private tag
if ($file !~ m!,!) {
my $fq_cvbr = shift; # archivename/[[[[category]branch]version]revision]
return (split(/\//, $fq_cvbr))[0];
}
-
+
sub extract_versionname {
my $name = shift;
$name =~ s/--(?:patch|version(?:fix)?|base)-\d+$//;
}
# convert a fully-qualified revision or version to a unique dirname:
-# normalperson@yhbt.net-05/mpd--uclinux--1--patch-2
+# normalperson@yhbt.net-05/mpd--uclinux--1--patch-2
# becomes: normalperson@yhbt.net-05,mpd--uclinux--1
#
# the git notion of a branch is closer to
return $ret;
}
-*git_branchname = $opt_o ? *old_style_branchname : *tree_dirname;
+*git_default_branchname = $opt_o ? *old_style_branchname : *tree_dirname;
+
+# retrieve default archive, since $branch_name_map keys might not include it
+sub get_default_archive {
+ if (!defined $default_archive) {
+ $default_archive = safe_pipe_capture($TLA,'my-default-archive');
+ chomp $default_archive;
+ }
+ return $default_archive;
+}
+
+sub git_branchname {
+ my $revision = shift;
+ my $name = extract_versionname($revision);
+
+ if (exists $branch_name_map{$name}) {
+ return $branch_name_map{$name};
+
+ } elsif ($name =~ m#^([^/]*)/(.*)$#
+ && $1 eq get_default_archive()
+ && exists $branch_name_map{$2}) {
+ # the names given in the command-line lacked the archive.
+ return $branch_name_map{$2};
+
+ } else {
+ return git_default_branchname($revision);
+ }
+}
sub process_patchset_accurate {
my $ps = shift;
-
+
# switch to that branch if we're not already in that branch:
if (-e "$git_dir/refs/heads/$ps->{branch}") {
system('git-checkout','-f',$ps->{branch}) == 0 or die "$! $?\n";
my $rm = safe_pipe_capture('git-ls-files','--others','-z');
rmtree(split(/\0/,$rm)) if $rm;
}
-
+
# Apply the import/changeset/merge into the working tree
my $dir = sync_to_ps($ps);
# read the new log entry:
parselog($ps, \@commitlog);
if ($ps->{id} =~ /--base-0$/ && $ps->{id} ne $psets[0]{id}) {
- # this should work when importing continuations
+ # this should work when importing continuations
if ($ps->{tag} && (my $branchpoint = eval { ptag($ps->{tag}) })) {
-
+
# find where we are supposed to branch from
- system('git-checkout','-f','-b',$ps->{branch},
- $branchpoint) == 0 or die "$! $?\n";
-
+ if (! -e "$git_dir/refs/heads/$ps->{branch}") {
+ system('git-branch',$ps->{branch},$branchpoint) == 0 or die "$! $?\n";
+
+ # We trust Arch with the fact that this is just a tag,
+ # and it does not affect the state of the tree, so
+ # we just tag and move on. If the user really wants us
+ # to consolidate more branches into one, don't tag because
+ # the tag name would be already taken.
+ tag($ps->{id}, $branchpoint);
+ ptag($ps->{id}, $branchpoint);
+ print " * Tagged $ps->{id} at $branchpoint\n";
+ }
+ system('git-checkout','-f',$ps->{branch}) == 0 or die "$! $?\n";
+
# remove any old stuff that got leftover:
my $rm = safe_pipe_capture('git-ls-files','--others','-z');
rmtree(split(/\0/,$rm)) if $rm;
-
- # If we trust Arch with the fact that this is just
- # a tag, and it does not affect the state of the tree
- # then we just tag and move on
- tag($ps->{id}, $branchpoint);
- ptag($ps->{id}, $branchpoint);
- print " * Tagged $ps->{id} at $branchpoint\n";
return 0;
} else {
warn "Tagging from unknown id unsupported\n" if $ps->{tag};
}
# allow multiple bases/imports here since Arch supports cherry-picks
# from unrelated trees
- }
-
+ }
+
# update the index with all the changes we got
system('git-diff-files --name-only -z | '.
'git-update-index --remove -z --stdin') == 0 or die "$! $?\n";
# does not handle permissions or any renames involving directories
sub process_patchset_fast {
my $ps = shift;
- #
+ #
# create the branch if needed
#
if ($ps->{type} eq 'i' && !$import) {
# new branch! we need to verify a few things
die "Branch on a non-tag!" unless $ps->{type} eq 't';
my $branchpoint = ptag($ps->{tag});
- die "Tagging from unknown id unsupported: $ps->{tag}"
+ die "Tagging from unknown id unsupported: $ps->{tag}"
unless $branchpoint;
-
+
# find where we are supposed to branch from
- system('git-checkout','-b',$ps->{branch},$branchpoint);
-
- # If we trust Arch with the fact that this is just
- # a tag, and it does not affect the state of the tree
- # then we just tag and move on
- tag($ps->{id}, $branchpoint);
- ptag($ps->{id}, $branchpoint);
- print " * Tagged $ps->{id} at $branchpoint\n";
+ if (! -e "$git_dir/refs/heads/$ps->{branch}") {
+ system('git-branch',$ps->{branch},$branchpoint) == 0 or die "$! $?\n";
+
+ # We trust Arch with the fact that this is just a tag,
+ # and it does not affect the state of the tree, so
+ # we just tag and move on. If the user really wants us
+ # to consolidate more branches into one, don't tag because
+ # the tag name would be already taken.
+ tag($ps->{id}, $branchpoint);
+ ptag($ps->{id}, $branchpoint);
+ print " * Tagged $ps->{id} at $branchpoint\n";
+ }
+ system('git-checkout',$ps->{branch}) == 0 or die "$! $?\n";
return 0;
- }
+ }
die $! if $?;
- }
+ }
#
# Apply the import/changeset/merge into the working tree
- #
+ #
if ($ps->{type} eq 'i' || $ps->{type} eq 't') {
apply_import($ps) or die $!;
$stats{import_or_tag}++;
# prepare update git's index, based on what arch knows
# about the pset, resolve parents, etc
#
-
- my @commitlog = safe_pipe_capture($TLA,'cat-archive-log',$ps->{id});
+
+ my @commitlog = safe_pipe_capture($TLA,'cat-archive-log',$ps->{id});
die "Error in cat-archive-log: $!" if $?;
-
+
parselog($ps,\@commitlog);
# imports don't give us good info
if (@$ren % 2) {
die "Odd number of entries in rename!?";
}
-
+
while (@$ren) {
my $from = shift @$ren;
- my $to = shift @$ren;
+ my $to = shift @$ren;
unless (-d dirname($to)) {
mkpath(dirname($to)); # will die on err
"Things may be a bit slow\n";
*process_patchset = *process_patchset_accurate;
}
-
+
foreach my $ps (@psets) {
# process patchsets
$ps->{branch} = git_branchname($ps->{id});
#
- # ensure we have a clean state
- #
+ # ensure we have a clean state
+ #
if (my $dirty = `git-diff-files`) {
die "Unclean tree when about to process $ps->{id} " .
" - did we fail to commit cleanly before?\n$dirty";
}
die $! if $?;
-
+
#
# skip commits already in repo
#
my $tree = `git-write-tree`;
die "cannot write tree $!" if $?;
chomp $tree;
-
+
#
# Who's your daddy?
#
close HEAD;
chomp $p;
push @par, '-p', $p;
- } else {
+ } else {
if ($ps->{type} eq 's') {
warn "Could not find the right head for the branch $ps->{branch}";
}
}
}
-
+
if ($ps->{merges}) {
push @par, find_parents($ps);
}
- #
+ #
# Commit, tag and clean state
#
$ENV{TZ} = 'GMT';
$ENV{GIT_COMMITTER_EMAIL} = $ps->{email};
$ENV{GIT_COMMITTER_DATE} = $ps->{date};
- my $pid = open2(*READER, *WRITER,'git-commit-tree',$tree,@par)
+ my $pid = open2(*READER, *WRITER,'git-commit-tree',$tree,@par)
or die $!;
- print WRITER $ps->{summary},"\n";
- print WRITER $ps->{message},"\n";
-
+ print WRITER $ps->{summary},"\n\n";
+
+ # only print message if it's not empty, to avoid a spurious blank line;
+ # also append an extra newline, so there's a blank line before the
+ # following "git-archimport-id:" line.
+ print WRITER $ps->{message},"\n\n" if ($ps->{message} ne "");
+
# make it easy to backtrack and figure out which Arch revision this was:
print WRITER 'git-archimport-id: ',$ps->{id},"\n";
-
+
close WRITER;
my $commitid = <READER>; # read
chomp $commitid;
}
#
# Update the branch
- #
+ #
open HEAD, ">","$git_dir/refs/heads/$ps->{branch}";
print HEAD $commitid;
close HEAD;
sub sync_to_ps {
my $ps = shift;
my $tree_dir = $tmp.'/'.tree_dirname($ps->{id});
-
+
$opt_v && print "sync_to_ps($ps->{id}) method: ";
if (-d $tree_dir) {
safe_pipe_capture($TLA,'get','--no-pristine',$ps->{id},$tree_dir);
$stats{get_new}++;
}
-
+
# added -I flag to rsync since we're going to fast! AIEEEEE!!!!
system('rsync','-aI','--delete','--exclude',$git_dir,
# '--exclude','.arch-inventory',
mkpath($tmp);
safe_pipe_capture($TLA,'get','-s','--no-pristine',$ps->{id},"$tmp/import");
- die "Cannot get import: $!" if $?;
+ die "Cannot get import: $!" if $?;
system('rsync','-aI','--delete', '--exclude',$git_dir,
'--exclude','.arch-ids','--exclude','{arch}',
"$tmp/import/", './');
die "Cannot rsync import:$!" if $?;
-
+
rmtree("$tmp/import");
die "Cannot remove tempdir: $!" if $?;
-
+
return 1;
}
# get the changeset
safe_pipe_capture($TLA,'get-changeset',$ps->{id},"$tmp/changeset");
die "Cannot get changeset: $!" if $?;
-
+
# apply patches
if (`find $tmp/changeset/patches -type f -name '*.patch'`) {
# this can be sped up considerably by doing
# (find | xargs cat) | patch
# but that can get mucked up by patches
- # with missing trailing newlines or the standard
+ # with missing trailing newlines or the standard
# 'missing newline' flag in the patch - possibly
# produced with an old/buggy diff.
# slow and safe, we invoke patch once per patchfile
# bring in new files
system('rsync','-aI','--exclude',$git_dir,
- '--exclude','.arch-ids',
+ '--exclude','.arch-ids',
'--exclude', '{arch}',
"$tmp/changeset/new-files-archive/",'./');
removed_files => 1,
removed_directories => 1,
);
-
+
chomp (@$log);
while ($_ = shift @$log) {
if (/^Continuation-of:\s*(.*)/) {
$ps->{tag} = $1;
$key = undef;
} elsif (/^Summary:\s*(.*)$/ ) {
- # summary can be multiline as long as it has a leading space
+ # summary can be multiline as long as it has a leading space.
+ # we squeeze it onto a single line, though.
$ps->{summary} = [ $1 ];
$key = 'summary';
} elsif (/^Creator: (.*)\s*<([^\>]+)>/) {
}
}
}
-
- # post-processing:
- $ps->{summary} = join("\n",@{$ps->{summary}})."\n";
+
+ # drop leading empty lines from the log message
+ while (@$log && $log->[0] eq '') {
+ shift @$log;
+ }
+ if (exists $ps->{summary} && @{$ps->{summary}}) {
+ $ps->{summary} = join(' ', @{$ps->{summary}});
+ }
+ elsif (@$log == 0) {
+ $ps->{summary} = 'empty commit message';
+ } else {
+ $ps->{summary} = $log->[0] . '...';
+ }
$ps->{message} = join("\n",@$log);
-
+
# skip Arch control files, unescape pika-escaped files
foreach my $k (keys %want_headers) {
next unless (defined $ps->{$k});
# write/read a tag
sub tag {
my ($tag, $commit) = @_;
-
+
if ($opt_o) {
$tag =~ s|/|--|g;
} else {
- # don't use subdirs for tags yet, it could screw up other porcelains
- $tag =~ s|/|,|g;
+ my $patchname = $tag;
+ $patchname =~ s/.*--//;
+ $tag = git_branchname ($tag) . '--' . $patchname;
}
-
+
if ($commit) {
open(C,">","$git_dir/refs/tags/$tag")
or die "Cannot create tag $tag: $!\n";
my ($tag, $commit) = @_;
# don't use subdirs for tags yet, it could screw up other porcelains
- $tag =~ s|/|,|g;
-
+ $tag =~ s|/|,|g;
+
my $tag_file = "$ptag_dir/$tag";
my $tag_branch_dir = dirname($tag_file);
mkpath($tag_branch_dir) unless (-d $tag_branch_dir);
or die "Cannot write tag $tag: $!\n";
close(C)
or die "Cannot write tag $tag: $!\n";
- $rptags{$commit} = $tag
+ $rptags{$commit} = $tag
unless $tag =~ m/--base-0$/;
} else { # read
# if the tag isn't there, return 0
# Identify what branches are merging into me
# and whether we are fully merged
# git-merge-base <headsha> <headsha> should tell
- # me what the base of the merge should be
+ # me what the base of the merge should be
#
my $ps = shift;
}
#
- # foreach branch find a merge base and walk it to the
+ # foreach branch find a merge base and walk it to the
# head where we are, collecting the merged patchsets that
# Arch has recorded. Keep that in @have
# Compare that with the commits on the other branch
# between merge-base and the tip of the branch (@need)
# and see if we have a series of consecutive patches
# starting from the merge base. The tip of the series
- # of consecutive patches merged is our new parent for
+ # of consecutive patches merged is our new parent for
# that branch.
#
foreach my $branch (keys %branches) {
next unless -e "$git_dir/refs/heads/$branch";
my $mergebase = `git-merge-base $branch $ps->{branch}`;
- if ($?) {
- # Don't die here, Arch supports one-way cherry-picking
- # between branches with no common base (or any relationship
- # at all beforehand)
- warn "Cannot find merge base for $branch and $ps->{branch}";
- next;
- }
+ if ($?) {
+ # Don't die here, Arch supports one-way cherry-picking
+ # between branches with no common base (or any relationship
+ # at all beforehand)
+ warn "Cannot find merge base for $branch and $ps->{branch}";
+ next;
+ }
chomp $mergebase;
# now walk up to the mergepoint collecting what patches we have
# merge what we have with what ancestors have
%have = (%have, %ancestorshave);
- # see what the remote branch has - these are the merges we
+ # see what the remote branch has - these are the merges we
# will want to have in a consecutive series from the mergebase
my $otherbranchtip = git_rev_parse($branch);
my @needraw = `git-rev-list --topo-order $otherbranchtip ^$mergebase`;
foreach my $needps (@needraw) { # get the psets
$needps = commitid2pset($needps);
# git-rev-list will also
- # list commits merged in via earlier
+ # list commits merged in via earlier
# merges. we are only interested in commits
# from the branch we're looking at
if ($branch eq $needps->{branch}) {
next unless ref $psets{$p}{merges};
my @merges = @{$psets{$p}{merges}};
foreach my $merge (@merges) {
- if ($parents{$merge}) {
+ if ($parents{$merge}) {
delete $parents{$merge};
}
}
sub commitid2pset {
my $commitid = shift;
chomp $commitid;
- my $name = $rptags{$commitid}
+ my $name = $rptags{$commitid}
|| die "Cannot find reverse tag mapping for $commitid";
$name =~ s|,|/|;
- my $ps = $psets{$name}
+ my $ps = $psets{$name}
|| (print Dumper(sort keys %psets)) && die "Cannot find patchset for $name";
return $ps;
}
my $archive = shift;
return 1 if $reachable{$archive};
return 0 if $unreachable{$archive};
-
+
if (system "$TLA whereis-archive $archive >/dev/null") {
if ($opt_a && (system($TLA,'register-archive',
"http://mirrors.sourcecontrol.net/$archive") == 0)) {
return 1;
}
}
-