From: Eric Wong <e@80x24.org>
To: meta@public-inbox.org
Subject: [PATCH 7/7] fetch: use manifest.js.gz for v1
Date: Sun, 12 Sep 2021 07:47:16 +0000 [thread overview]
Message-ID: <20210912074716.4713-8-e@80x24.org> (raw)
In-Reply-To: <20210912074716.4713-1-e@80x24.org>
This is gentler to the remote HTTP server in the no-op case and
will allow client migrations to some v2-ish format without
forcing the client to redownload everything.
---
lib/PublicInbox/Fetch.pm | 147 +++++++++++++++++++++--------------
lib/PublicInbox/LeiMirror.pm | 6 +-
2 files changed, 90 insertions(+), 63 deletions(-)
diff --git a/lib/PublicInbox/Fetch.pm b/lib/PublicInbox/Fetch.pm
index d795731c..9613a582 100644
--- a/lib/PublicInbox/Fetch.pm
+++ b/lib/PublicInbox/Fetch.pm
@@ -35,49 +35,13 @@ sub remote_url ($$) {
my $fh = popen_rd($cmd, undef, { -C => $dir, 2 => $lei->{2} });
my $url = <$fh>;
close $fh or return;
- chomp $url;
+ $url =~ s!/*\n!!s;
$url;
}
-sub do_fetch {
- my ($cls, $lei, $cd) = @_;
- my $ibx_ver;
- my $curl = PublicInbox::LeiCurl->new($lei) or return;
- my $dir = PublicInbox::Admin::resolve_inboxdir($cd, \$ibx_ver);
- if ($ibx_ver == 1) {
- my $url = remote_url($lei, $dir) //
- die "E: $dir missing remote.origin.url\n";
- my $uri = URI->new($url);
- my $torsocks = $curl->torsocks($lei, $uri);
- my $opt = { -C => $dir };
- my $cmd = [ @$torsocks, fetch_cmd($lei, $opt) ];
- my $cerr = PublicInbox::LeiMirror::run_reap($lei, $cmd, $opt);
- $lei->child_error($cerr, "@$cmd failed") if $cerr;
- return;
- }
- # v2:
- opendir my $dh, "$dir/git" or die "opendir $dir/git: $!";
- my @epochs = sort { $b <=> $a } map { substr($_, 0, -4) + 0 }
- grep(/\A[0-9]+\.git\z/, readdir($dh));
- my ($git_url, $epoch);
- for my $nr (@epochs) { # try newest epoch, first
- my $edir = "$dir/git/$nr.git";
- if (defined(my $url = remote_url($lei, $edir))) {
- $git_url = $url;
- $epoch = $nr;
- last;
- } else {
- warn "W: $edir missing remote.origin.url\n";
- }
- }
- $git_url or die "Unable to determine git URL\n";
- my $inbox_url = $git_url;
- $inbox_url =~ s!/git/$epoch(?:\.git)?/?\z!! or
- $inbox_url =~ s!/$epoch(?:\.git)?/?\z!! or die <<EOM;
-Unable to infer inbox URL from <$git_url>
-EOM
- $lei->qerr("# inbox URL: $inbox_url/");
- my $muri = URI->new("$inbox_url/manifest.js.gz");
+sub do_manifest ($$$) {
+ my ($lei, $dir, $ibx_uri) = @_;
+ my $muri = URI->new("$ibx_uri/manifest.js.gz");
my $ft = File::Temp->new(TEMPLATE => 'manifest-XXXX',
UNLINK => 1, DIR => $dir);
my $fn = $ft->filename;
@@ -91,13 +55,16 @@ EOM
$lei->err($@) if $@;
push @opt, '-z', $mf if defined($m0);
}
- my $curl_cmd = $curl->for_uri($lei, $muri, @opt);
+ my $curl_cmd = $lei->{curl}->for_uri($lei, $muri, @opt);
my $opt = {};
$opt->{$_} = $lei->{$_} for (0..2);
my $cerr = PublicInbox::LeiMirror::run_reap($lei, $curl_cmd, $opt);
- return $lei->child_error($cerr, "@$curl_cmd failed") if $cerr;
- return if !-s $ft; # 304 Not Modified via curl -z
-
+ if ($cerr) {
+ return [ 404 ] if ($cerr >> 8) == 22; # 404 Missing
+ $lei->child_error($cerr, "@$curl_cmd failed");
+ return;
+ }
+ return [ 304 ] if !-s $ft; # 304 Not Modified via curl -z
my $m1 = PublicInbox::LeiMirror::decode_manifest($ft, $fn, $muri);
my $mdiff = { %$m1 };
@@ -110,36 +77,96 @@ EOM
my $t1 = $cur->{modified} // next;
delete($mdiff->{$k}) if $f0 eq $f1 && $t0 == $t1;
}
- my $ibx_uri = URI->new("$inbox_url/");
my ($path_pfx, $v1_bare, @v2_epochs) =
PublicInbox::LeiMirror::deduce_epochs($mdiff, $ibx_uri->path);
- defined($v1_bare) and die <<EOM;
+ [ 200, $path_pfx, $v1_bare, \@v2_epochs, $muri, $ft, $mf ];
+}
+
+sub do_fetch {
+ my ($cls, $lei, $cd) = @_;
+ my $ibx_ver;
+ $lei->{curl} //= PublicInbox::LeiCurl->new($lei) or return;
+ my $dir = PublicInbox::Admin::resolve_inboxdir($cd, \$ibx_ver);
+ my ($ibx_uri, @git_dir, @epochs);
+ if ($ibx_ver == 1) {
+ my $url = remote_url($lei, $dir) //
+ die "E: $dir missing remote.origin.url\n";
+ $ibx_uri = URI->new($url);
+ } else { # v2:
+ opendir my $dh, "$dir/git" or die "opendir $dir/git: $!";
+ @epochs = sort { $b <=> $a } map { substr($_, 0, -4) + 0 }
+ grep(/\A[0-9]+\.git\z/, readdir($dh));
+ my ($git_url, $epoch);
+ for my $nr (@epochs) { # try newest epoch, first
+ my $edir = "$dir/git/$nr.git";
+ if (defined(my $url = remote_url($lei, $edir))) {
+ $git_url = $url;
+ $epoch = $nr;
+ last;
+ } else {
+ warn "W: $edir missing remote.origin.url\n";
+ }
+ }
+ $git_url or die "Unable to determine git URL\n";
+ my $inbox_url = $git_url;
+ $inbox_url =~ s!/git/$epoch(?:\.git)?/?\z!! or
+ $inbox_url =~ s!/$epoch(?:\.git)?/?\z!! or die <<EOM;
+Unable to infer inbox URL from <$git_url>
+EOM
+ $ibx_uri = URI->new($inbox_url);
+ }
+ $lei->qerr("# inbox URL: $ibx_uri/");
+ my $res = do_manifest($lei, $dir, $ibx_uri) or return;
+ my ($code, $path_pfx, $v1_bare, $v2_epochs, $muri, $ft, $mf) = @$res;
+ return if $code == 304;
+ if ($code == 404) {
+ # any pre-manifest.js.gz instances running? Just fetch all
+ # existing ones and unconditionally try cloning the next
+ $v2_epochs = [ map {;
+ "$dir/git/$_.git";
+ } @epochs ];
+ push @$v2_epochs, "$dir/git/".($epochs[-1] + 1) if @epochs;
+ } else {
+ $code == 200 or die "BUG unexpected code $code\n";
+ }
+ if ($ibx_ver == 2) {
+ defined($v1_bare) and warn <<EOM;
E: got v1 `$v1_bare' when expecting v2 epoch(s) in <$muri>, WTF?
EOM
- my @epoch_nr = sort { $a <=> $b }
- map { my ($nr) = (m!/([0-9]+)\.git\z!g) } @v2_epochs;
-
+ @git_dir = map { "$dir/git/$_.git" } sort { $a <=> $b }
+ map { my ($nr) = (m!/([0-9]+)\.git\z!g) } @$v2_epochs;
+ } else {
+ $v1_bare eq $dir or warn "$v1_bare != $dir";
+ $git_dir[0] = $v1_bare // $dir;
+ }
# n.b. this expects all epochs are from the same host
- my $torsocks = $curl->torsocks($lei, $muri);
- for my $nr (@epoch_nr) {
- my $dir = "$dir/git/$nr.git";
+ my $torsocks = $lei->{curl}->torsocks($lei, $muri);
+ for my $d (@git_dir) {
my $cmd;
- my $opt = {};
- if (-d $dir) {
- $opt->{-C} = $dir;
+ my $opt = {}; # for spawn
+ if (-d $d) {
+ $opt->{-C} = $d;
$cmd = [ @$torsocks, fetch_cmd($lei, $opt) ];
} else {
my $e_uri = $ibx_uri->clone;
- $e_uri->path($ibx_uri->path."git/$nr.git");
+ my ($epath) = ($d =~ m!/(git/[0-9]+\.git)\z!);
+ $e_uri->path($ibx_uri->path.$epath);
$cmd = [ @$torsocks,
PublicInbox::LeiMirror::clone_cmd($lei, $opt),
- $$e_uri, $dir ];
+ $$e_uri, $d];
}
my $cerr = PublicInbox::LeiMirror::run_reap($lei, $cmd, $opt);
- return $lei->child_error($cerr, "@$cmd failed") if $cerr;
+ # do not bail on clone failure if we didn't have a manifest
+ if ($cerr && ($code == 200 || -d $d)) {
+ $lei->child_error($cerr, "@$cmd failed");
+ return;
+ }
+ }
+ if ($ft) {
+ my $fn = $ft->filename;
+ rename($fn, $mf) or die "E: rename($fn, $mf): $!\n";
+ $ft->unlink_on_destroy(0);
}
- rename($fn, $mf) or die "E: rename($fn, $mf): $!\n";
- $ft->unlink_on_destroy(0);
}
1;
diff --git a/lib/PublicInbox/LeiMirror.pm b/lib/PublicInbox/LeiMirror.pm
index fe1cefe2..23813dcf 100644
--- a/lib/PublicInbox/LeiMirror.pm
+++ b/lib/PublicInbox/LeiMirror.pm
@@ -295,15 +295,15 @@ EOM
$uri->clone
} @v2_epochs;
clone_v2($self, \@v2_epochs);
- my $fin = "$self->{dst}/manifest.js.gz";
- rename($fn, $fin) or die "E: rename($fn, $fin): $!";
- $ft->unlink_on_destroy(0);
} elsif (defined $v1_bare) {
clone_v1($self);
} else {
die "E: confused by <$uri>, possible matches:\n\t",
join(', ', sort keys %$m), "\n";
}
+ my $fin = "$self->{dst}/manifest.js.gz";
+ rename($fn, $fin) or die "E: rename($fn, $fin): $!";
+ $ft->unlink_on_destroy(0);
}
sub start_clone_url {
prev parent reply other threads:[~2021-09-12 7:47 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-09-12 7:47 [PATCH 0/7] new public-inbox-{clone,fetch} commands Eric Wong
2021-09-12 7:47 ` [PATCH 1/7] lei_mirror: simplify error reporting Eric Wong
2021-09-12 7:47 ` [PATCH 2/7] lei_mirror: fix error message Eric Wong
2021-09-12 7:47 ` [PATCH 3/7] new public-inbox-{clone,fetch} commands Eric Wong
2021-09-12 7:47 ` [PATCH 4/7] clone|lei_mirror: write description in mirrors Eric Wong
2021-09-12 7:47 ` [PATCH 5/7] import: do not write a "description" file Eric Wong
2021-09-12 7:47 ` [PATCH 6/7] init: set a useful description Eric Wong
2021-09-12 7:47 ` Eric Wong [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
List information: https://public-inbox.org/README
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20210912074716.4713-8-e@80x24.org \
--to=e@80x24.org \
--cc=meta@public-inbox.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).