user/dev discussion of public-inbox itself
 help / color / Atom feed
From: Eric Wong <e@80x24.org>
To: meta@public-inbox.org
Subject: [PATCH 4/6] lazy load Xapian and make it optional for v2
Date: Wed, 15 May 2019 06:33:52 +0000
Message-ID: <20190515063354.52259-5-e@80x24.org> (raw)
In-Reply-To: <20190515063354.52259-1-e@80x24.org>

More tests work without Search::Xapian, now.
Usability issues still need to be fixed
---
 INSTALL                      |  2 +-
 TODO                         |  2 -
 ci/deps.perl                 |  7 +--
 ci/profiles.sh               |  3 +-
 lib/PublicInbox/Inbox.pm     | 19 ++++----
 lib/PublicInbox/Search.pm    | 84 +++++++++++++++++++++---------------
 lib/PublicInbox/SearchIdx.pm | 50 +++++++++++++--------
 lib/PublicInbox/WWW.pm       |  7 ++-
 script/public-inbox-index    | 25 ++++++++---
 script/public-inbox-init     |  1 +
 script/public-inbox-purge    |  5 +++
 t/admin.t                    |  2 +-
 t/cgi.t                      | 16 +++++--
 t/indexlevels-mirror.t       | 17 +++++---
 t/nntp.t                     |  2 +-
 t/nntpd.t                    | 15 +++++--
 t/over.t                     |  3 +-
 t/psgi_bad_mids.t            |  3 +-
 t/psgi_scan_all.t            |  4 +-
 t/psgi_search.t              |  4 +-
 t/purge.t                    |  2 +-
 t/search-thr-index.t         |  5 ++-
 t/search.t                   |  7 +--
 t/v1reindex.t                |  3 +-
 t/v2reindex.t                |  5 ++-
 t/watch_maildir_v2.t         |  4 +-
 26 files changed, 184 insertions(+), 113 deletions(-)

diff --git a/INSTALL b/INSTALL
index 313a295..72e0763 100644
--- a/INSTALL
+++ b/INSTALL
@@ -69,7 +69,7 @@ Numerous optional modules are likely to be useful as well:
 - Search::Xapian                   deb: libsearch-xapian-perl
                                    pkg: p5-Search-Xapian
                                    rpm: perl-Search-Xapian
-                                   (for v2, HTTP search)
+                                   (HTTP search)
 
 - Net::Server                      deb: libnet-server-perl
                                    pkg: pkg-Net-Server
diff --git a/TODO b/TODO
index d947b0f..4953439 100644
--- a/TODO
+++ b/TODO
@@ -69,8 +69,6 @@ all need to be considered for everything we introduce)
 
 * large mbox/Maildir/MH/NNTP spool import (see PublicInbox::Import)
 
-* Allow NNTP and more of PSGI code to work without Xapian
-
 * Read-only WebDAV interface to the git repo so it can be mounted
   via davfs2 or fusedav to avoid full clones.
   davfs2 needs Range: request support for this to be feasible:
diff --git a/ci/deps.perl b/ci/deps.perl
index e0fda01..32b0226 100755
--- a/ci/deps.perl
+++ b/ci/deps.perl
@@ -70,12 +70,7 @@ if (@precious) {
 
 
 # bare minimum for v2
-$profiles->{v2essential} = [ @{$profiles->{essential}}, qw(
-	DBD::SQLite
-	DBI
-	Search::Xapian
-	xapian-compact
-) ];
+$profiles->{v2essential} = [ @{$profiles->{essential}}, qw(DBD::SQLite DBI) ];
 
 # package names which can't be mapped automatically:
 my $non_auto = {
diff --git a/ci/profiles.sh b/ci/profiles.sh
index 31140f3..5d7c2cf 100755
--- a/ci/profiles.sh
+++ b/ci/profiles.sh
@@ -62,8 +62,9 @@ essential devtest-
 EOF
 	;;
 debian-sid|debian-9|debian-10) sed "s/^/$PKG_FMT /" <<EOF
-all devtest-
 all devtest
+all devtest Search::Xapian-
+all devtest-
 v2essential
 essential
 essential devtest-
diff --git a/lib/PublicInbox/Inbox.pm b/lib/PublicInbox/Inbox.pm
index dc186b7..813ed99 100644
--- a/lib/PublicInbox/Inbox.pm
+++ b/lib/PublicInbox/Inbox.pm
@@ -25,7 +25,7 @@ sub cleanup_task () {
 	my $next = {};
 	for my $ibx (values %$CLEANUP) {
 		my $again;
-		foreach my $f (qw(mm search)) {
+		foreach my $f (qw(mm search over)) {
 			delete $ibx->{$f} if SvREFCNT($ibx->{$f}) == 1;
 		}
 		my $expire = time - 60;
@@ -37,7 +37,7 @@ sub cleanup_task () {
 				$again = 1 if $git->cleanup($expire);
 			}
 		}
-		$again ||= !!($ibx->{mm} || $ibx->{search});
+		$again ||= !!($ibx->{over} || $ibx->{mm} || $ibx->{search});
 		$next->{"$ibx"} = $ibx if $again;
 	}
 	$CLEANUP = $next;
@@ -175,14 +175,17 @@ sub search ($;$) {
 		require PublicInbox::Search;
 		PublicInbox::Search->new($self, $self->{altid});
 	};
-	# TODO: lazily load Xapian
-	# return $srch if $over_only || eval { $srch->xdb };
-	# undef;
+	($over_only || eval { $srch->xdb }) ? $srch : undef;
 }
 
 sub over ($) {
-	my $srch = search($_[0], 1) or return;
-	$srch->{over_ro};
+	my ($self) = @_;
+	my $srch = search($self, 1) or return;
+	$self->{over} ||= eval {
+		my $over = $srch->{over_ro};
+		$over->dbh_new; # may fail
+		$over;
+	}
 }
 
 sub try_cat {
@@ -290,7 +293,7 @@ sub nntp_url {
 sub nntp_usable {
 	my ($self) = @_;
 	my $ret = mm($self) && over($self);
-	$self->{mm} = $self->{search} = undef;
+	$self->{mm} = $self->{over} = $self->{search} = undef;
 	$ret;
 }
 
diff --git a/lib/PublicInbox/Search.pm b/lib/PublicInbox/Search.pm
index e79ec0f..b1e62f4 100644
--- a/lib/PublicInbox/Search.pm
+++ b/lib/PublicInbox/Search.pm
@@ -12,11 +12,21 @@ use constant TS => 0;  # Received: header in Unix time
 use constant YYYYMMDD => 1; # Date: header for searching in the WWW UI
 use constant DT => 2; # Date: YYYYMMDDHHMMSS
 
-use Search::Xapian qw/:standard/;
 use PublicInbox::SearchMsg;
 use PublicInbox::MIME;
 use PublicInbox::MID qw/id_compress/;
 use PublicInbox::Over;
+my $QP_FLAGS;
+sub load_xapian () {
+	$QP_FLAGS ||= eval {
+		require Search::Xapian;
+		Search::Xapian->import(qw(:standard));
+
+		# n.b. FLAG_PURE_NOT is expensive not suitable for a public
+		# website as it could become a denial-of-service vector
+		FLAG_PHRASE()|FLAG_BOOLEAN()|FLAG_LOVEHATE()|FLAG_WILDCARD();
+	};
+};
 
 # This is English-only, everything else is non-standard and may be confused as
 # a prefix common in patch emails
@@ -41,10 +51,6 @@ use constant {
 	#      (commit 83425ef12e4b65cdcecd11ddcb38175d4a91d5a0)
 	# 14 - fix ghost root vivification
 	SCHEMA_VERSION => 15,
-
-	# n.b. FLAG_PURE_NOT is expensive not suitable for a public website
-	# as it could become a denial-of-service vector
-	QP_FLAGS => FLAG_PHRASE|FLAG_BOOLEAN|FLAG_LOVEHATE|FLAG_WILDCARD,
 };
 
 my %bool_pfx_external = (
@@ -113,18 +119,43 @@ EOF
 );
 chomp @HELP;
 
-sub xdir {
-	my ($self) = @_;
+sub xdir ($;$) {
+	my ($self, $rdonly) = @_;
 	if ($self->{version} == 1) {
 		"$self->{mainrepo}/public-inbox/xapian" . SCHEMA_VERSION;
 	} else {
 		my $dir = "$self->{mainrepo}/xap" . SCHEMA_VERSION;
+		return $dir if $rdonly;
+
 		my $part = $self->{partition};
 		defined $part or die "partition not given";
 		$dir .= "/$part";
 	}
 }
 
+sub xdb ($) {
+	my ($self) = @_;
+	$self->{xdb} ||= do {
+		load_xapian();
+		my $dir = xdir($self, 1);
+		if ($self->{version} >= 2) {
+			my $xdb;
+			foreach my $part (<$dir/*>) {
+				-d $part && $part =~ m!/\d+\z! or next;
+				my $sub = Search::Xapian::Database->new($part);
+				if ($xdb) {
+					$xdb->add_database($sub);
+				} else {
+					$xdb = $sub;
+				}
+			}
+			$xdb;
+		} else {
+			Search::Xapian::Database->new($dir);
+		}
+	};
+}
+
 sub new {
 	my ($class, $mainrepo, $altid) = @_;
 	my $version = 1;
@@ -138,33 +169,16 @@ sub new {
 		altid => $altid,
 		version => $version,
 	}, $class;
-	my $dir;
-	if ($version >= 2) {
-		$dir = "$self->{mainrepo}/xap" . SCHEMA_VERSION;
-		my $xdb;
-		my $parts = 0;
-		foreach my $part (<$dir/*>) {
-			-d $part && $part =~ m!/\d+\z! or next;
-			$parts++;
-			my $sub = Search::Xapian::Database->new($part);
-			if ($xdb) {
-				$xdb->add_database($sub);
-			} else {
-				$xdb = $sub;
-			}
-		}
-		$self->{xdb} = $xdb;
-	} else {
-		$dir = $self->xdir;
-		$self->{xdb} = Search::Xapian::Database->new($dir);
-	}
+	my $dir = xdir($self, 1);
 	$self->{over_ro} = PublicInbox::Over->new("$dir/over.sqlite3");
 	$self;
 }
 
 sub reopen {
 	my ($self) = @_;
-	$self->{xdb}->reopen;
+	if (my $xdb = $self->{xdb}) {
+		$xdb->reopen;
+	}
 	$self; # make chaining easier
 }
 
@@ -175,7 +189,8 @@ sub query {
 	if ($query_string eq '' && !$opts->{mset}) {
 		$self->{over_ro}->recent($opts);
 	} else {
-		my $query = $self->qp->parse_query($query_string, QP_FLAGS);
+		my $qp = qp($self);
+		my $query = $qp->parse_query($query_string, $QP_FLAGS);
 		$opts->{relevance} = 1 unless exists $opts->{relevance};
 		_do_enquire($self, $query, $opts);
 	}
@@ -213,7 +228,8 @@ sub _do_enquire {
 
 sub _enquire_once {
 	my ($self, $query, $opts) = @_;
-	my $enquire = Search::Xapian::Enquire->new($self->{xdb});
+	my $xdb = xdb($self);
+	my $enquire = Search::Xapian::Enquire->new($xdb);
 	$enquire->set_query($query);
 	$opts ||= {};
         my $desc = !$opts->{asc};
@@ -246,13 +262,13 @@ sub qp {
 
 	my $qp = $self->{query_parser};
 	return $qp if $qp;
-
+	my $xdb = xdb($self);
 	# new parser
 	$qp = Search::Xapian::QueryParser->new;
-	$qp->set_default_op(OP_AND);
-	$qp->set_database($self->{xdb});
+	$qp->set_default_op(OP_AND());
+	$qp->set_database($xdb);
 	$qp->set_stemmer($self->stemmer);
-	$qp->set_stemming_strategy(STEM_SOME);
+	$qp->set_stemming_strategy(STEM_SOME());
 	$qp->set_max_wildcard_expansion(100);
 	$qp->add_valuerangeprocessor(
 		Search::Xapian::NumberValueRangeProcessor->new(YYYYMMDD, 'd:'));
diff --git a/lib/PublicInbox/SearchIdx.pm b/lib/PublicInbox/SearchIdx.pm
index 1b86f72..135b5eb 100644
--- a/lib/PublicInbox/SearchIdx.pm
+++ b/lib/PublicInbox/SearchIdx.pm
@@ -54,11 +54,10 @@ sub new {
 				die("Invalid indexlevel $ibx->{indexlevel}\n");
 			}
 		}
-	} else { # v1
+	} else { # FIXME: old tests: old tests
 		$ibx = { mainrepo => $git_dir, version => 1 };
 	}
 	$ibx = PublicInbox::InboxWritable->new($ibx);
-	require Search::Xapian::WritableDatabase;
 	my $self = bless {
 		mainrepo => $mainrepo,
 		-inbox => $ibx,
@@ -84,25 +83,36 @@ sub new {
 	$self;
 }
 
+sub need_xapian ($) { $_[0]->{indexlevel} =~ $xapianlevels }
+
 sub _xdb_release {
 	my ($self) = @_;
-	my $xdb = delete $self->{xdb} or croak 'not acquired';
-	$xdb->close;
+	if (need_xapian($self)) {
+		my $xdb = delete $self->{xdb} or croak 'not acquired';
+		$xdb->close;
+	}
 	$self->lock_release if $self->{creat};
 	undef;
 }
 
 sub _xdb_acquire {
 	my ($self) = @_;
-	croak 'already acquired' if $self->{xdb};
+	my $flag;
 	my $dir = $self->xdir;
-	my $flag = Search::Xapian::DB_OPEN;
+	if (need_xapian($self)) {
+		croak 'already acquired' if $self->{xdb};
+		PublicInbox::Search::load_xapian();
+		require Search::Xapian::WritableDatabase;
+		$flag = $self->{creat} ?
+			Search::Xapian::DB_CREATE_OR_OPEN() :
+			Search::Xapian::DB_OPEN();
+	}
 	if ($self->{creat}) {
 		require File::Path;
 		$self->lock_acquire;
 		File::Path::mkpath($dir);
-		$flag = Search::Xapian::DB_CREATE_OR_OPEN;
 	}
+	return unless defined $flag;
 	$self->{xdb} = Search::Xapian::WritableDatabase->new($dir, $flag);
 }
 
@@ -342,7 +352,7 @@ sub add_message {
 		$num = index_mm($self, $mime);
 	}
 	eval {
-		if ($self->{indexlevel} =~ $xapianlevels) {
+		if (need_xapian($self)) {
 			$self->add_xapian($mime, $num, $oid, $mids, $mid0)
 		}
 		if (my $over = $self->{over}) {
@@ -383,7 +393,6 @@ sub batch_do {
 # v1 only, where $mid is unique
 sub remove_message {
 	my ($self, $mid) = @_;
-	my $db = $self->{xdb};
 	$mid = mid_clean($mid);
 
 	if (my $over = $self->{over}) {
@@ -394,7 +403,8 @@ sub remove_message {
 			warn "<$mid> missing for removal from overview\n";
 		}
 	}
-	return if $self->{indexlevel} !~ $xapianlevels;
+	return unless need_xapian($self);
+	my $db = $self->{xdb};
 	my $nr = 0;
 	eval {
 		batch_do($self, 'Q' . $mid, sub {
@@ -413,10 +423,12 @@ sub remove_message {
 # MID is a hint in V2
 sub remove_by_oid {
 	my ($self, $oid, $mid) = @_;
-	my $db = $self->{xdb};
 
 	$self->{over}->remove_oid($oid, $mid) if $self->{over};
 
+	return unless need_xapian($self);
+	my $db = $self->{xdb};
+
 	# XXX careful, we cannot use batch_do here since we conditionally
 	# delete documents based on other factors, so we cannot call
 	# find_doc_ids twice.
@@ -664,7 +676,7 @@ sub _last_x_commit {
 	my ($self, $mm) = @_;
 	my $lm = $mm->last_commit || '';
 	my $lx = '';
-	if ($self->{indexlevel} =~ $xapianlevels) {
+	if (need_xapian($self)) {
 		$lx = $self->{xdb}->get_metadata('last_commit') || '';
 	} else {
 		$lx = $lm;
@@ -695,7 +707,7 @@ sub _index_sync {
 		$self->{over}->disconnect;
 		$git->cleanup;
 		delete $self->{txn};
-		$xdb->cancel_transaction;
+		$xdb->cancel_transaction if $xdb;
 		$xdb = _xdb_release($self);
 
 		# ensure we leak no FDs to "git log" with Xapian <= 1.2
@@ -717,7 +729,7 @@ sub _index_sync {
 			}
 			$dbh->commit;
 		}
-		if ($newest && $self->{indexlevel} =~ $xapianlevels) {
+		if ($newest && need_xapian($self)) {
 			my $cur = $xdb->get_metadata('last_commit');
 			if (need_update($self, $cur, $newest)) {
 				$xdb->set_metadata('last_commit', $newest);
@@ -785,7 +797,7 @@ sub begin_txn_lazy {
 	$self->{-inbox}->with_umask(sub {
 		my $xdb = $self->{xdb} || $self->_xdb_acquire;
 		$self->{over}->begin_lazy if $self->{over};
-		$xdb->begin_transaction;
+		$xdb->begin_transaction if $xdb;
 		$self->{txn} = 1;
 		$xdb;
 	});
@@ -795,14 +807,18 @@ sub commit_txn_lazy {
 	my ($self) = @_;
 	delete $self->{txn} or return;
 	$self->{-inbox}->with_umask(sub {
-		$self->{xdb}->commit_transaction;
+		if (my $xdb = $self->{xdb}) {
+			$xdb->commit_transaction;
+		}
 		$self->{over}->commit_lazy if $self->{over};
 	});
 }
 
 sub worker_done {
 	my ($self) = @_;
-	die "$$ $0 xdb not released\n" if $self->{xdb};
+	if (need_xapian($self)) {
+		die "$$ $0 xdb not released\n" if $self->{xdb};
+	}
 	die "$$ $0 still in transaction\n" if $self->{txn};
 }
 
diff --git a/lib/PublicInbox/WWW.pm b/lib/PublicInbox/WWW.pm
index 0f963dc..8e1b1af 100644
--- a/lib/PublicInbox/WWW.pm
+++ b/lib/PublicInbox/WWW.pm
@@ -149,8 +149,11 @@ sub preload {
 	require PublicInbox::MIME;
 	require Digest::SHA;
 	require POSIX;
-
-	foreach (qw(PublicInbox::Search PublicInbox::SearchView
+	eval {
+		require PublicInbox::Search;
+		PublicInbox::Search::load_xapian();
+	};
+	foreach (qw(PublicInbox::SearchView
 			PublicInbox::Mbox IO::Compress::Gzip
 			PublicInbox::NewsWWW)) {
 		eval "require $_;";
diff --git a/script/public-inbox-index b/script/public-inbox-index
index b353093..53def9a 100755
--- a/script/public-inbox-index
+++ b/script/public-inbox-index
@@ -26,10 +26,12 @@ if ($@) {
 my $reindex;
 my $prune;
 my $jobs = undef;
+my $indexlevel;
 my %opts = (
 	'--reindex' => \$reindex,
 	'--jobs|j=i' => \$jobs,
 	'--prune' => \$prune,
+        'L|indexlevel=s' => \$indexlevel,
 );
 GetOptions(%opts) or die "bad command-line args\n$usage";
 die "--jobs must be positive\n" if defined $jobs && $jobs < 0;
@@ -55,18 +57,27 @@ defined($config) and $config->each_inbox(sub {
 });
 
 foreach my $dir (@dirs) {
-	if (!ref($dir) && -f "$dir/inbox.lock") { # v2
-		my $ibx = { mainrepo => $dir, name => 'unnamed' };
+	if (!ref($dir)) {
+		unless (-d $dir) {
+			die "$dir does not appear to be an inbox repository\n";
+		}
+		my $ibx = {
+			mainrepo => $dir,
+			name => 'unnamed',
+			indexlevel => $indexlevel,
+			version => -f "$dir/inbox.lock" ? 2 : 1,
+		};
 		$dir = PublicInbox::Inbox->new($ibx);
+	} elsif (defined $indexlevel && !defined($dir->{indexlevel})) {
+		# XXX: users can shoot themselves in the foot, with this...
+		$dir->{indexlevel} = $indexlevel;
 	}
-	index_dir($dir);
+
+	index_inbox($dir);
 }
 
-sub index_dir {
+sub index_inbox {
 	my ($repo) = @_;
-	if (!ref $repo && ! -d $repo) {
-		die "$repo does not appear to be an inbox repository\n";
-	}
 	if (ref($repo) && ($repo->{version} || 1) == 2) {
 		eval { require PublicInbox::V2Writable };
 		die "v2 requirements not met: $@\n" if $@;
diff --git a/script/public-inbox-init b/script/public-inbox-init
index 8bb7845..9f0bd1b 100755
--- a/script/public-inbox-init
+++ b/script/public-inbox-init
@@ -109,6 +109,7 @@ my $ibx = PublicInbox::Inbox->new({
 	name => $name,
 	version => $version,
 	-primary_address => $address[0],
+	indexlevel => $indexlevel,
 });
 
 if ($version >= 2) {
diff --git a/script/public-inbox-purge b/script/public-inbox-purge
index 688dd95..264bcde 100755
--- a/script/public-inbox-purge
+++ b/script/public-inbox-purge
@@ -63,6 +63,11 @@ if ($all) {
 				mainrepo => $dir,
 			});
 		};
+
+		# somebody could "rm -r" all the Xapian directories;
+		# let them purge the overview, at least
+		$ibx->{indexlevel} ||= 'basic' unless $ibx->search;
+
 		push @inboxes, $ibx;
 	}
 
diff --git a/t/admin.t b/t/admin.t
index b5a4383..3790c9e 100644
--- a/t/admin.t
+++ b/t/admin.t
@@ -50,7 +50,7 @@ SKIP: {
 
 # v2
 SKIP: {
-	for my $m (qw(DBD::SQLite Search::Xapian)) {
+	for my $m (qw(DBD::SQLite)) {
 		skip "$m missing", 5 unless eval "require $m";
 	}
 	use_ok 'PublicInbox::V2Writable';
diff --git a/t/cgi.t b/t/cgi.t
index b24bbc4..d3172bf 100644
--- a/t/cgi.t
+++ b/t/cgi.t
@@ -30,6 +30,7 @@ my $cfgpfx = "publicinbox.test";
 	my %cfg = (
 		"$cfgpfx.address" => $addr,
 		"$cfgpfx.mainrepo" => $maindir,
+		"$cfgpfx.indexlevel" => 'basic',
 	);
 	while (my ($k,$v) = each %cfg) {
 		is(0, system(qw(git config --file), $pi_config, $k, $v),
@@ -39,9 +40,12 @@ my $cfgpfx = "publicinbox.test";
 
 use_ok 'PublicInbox::Git';
 use_ok 'PublicInbox::Import';
-use_ok 'Email::MIME';
-my $git = PublicInbox::Git->new($maindir);
-my $im = PublicInbox::Import->new($git, 'test', $addr);
+use_ok 'PublicInbox::Inbox';
+use_ok 'PublicInbox::V1Writable';
+use_ok 'PublicInbox::Config';
+my $cfg = PublicInbox::Config->new($pi_config);
+my $ibx = $cfg->lookup_name('test');
+my $im = PublicInbox::V1Writable->new($ibx);
 
 {
 	local $ENV{HOME} = $home;
@@ -103,8 +107,9 @@ EOF
 	like($res->{head}, qr/^Status: 501 /, "search not-yet-enabled");
 	my $indexed;
 	eval {
+		require DBD::SQLite;
 		require PublicInbox::SearchIdx;
-		my $s = PublicInbox::SearchIdx->new($maindir, 1);
+		my $s = PublicInbox::SearchIdx->new($ibx, 1);
 		$s->index_sync;
 		$indexed = 1;
 	};
@@ -120,6 +125,7 @@ EOF
 		};
 	} else {
 		like($res->{head}, qr/^Status: 501 /, "search not available");
+		SKIP: { skip 'DBD::SQLite not available', 2 };
 	}
 
 	my $have_xml_feed = eval { require XML::Feed; 1 } if $indexed;
@@ -132,6 +138,8 @@ EOF
 		my $p = XML::Feed->parse(\($res->{body}));
 		is($p->format, "Atom", "parsed atom feed");
 		is(scalar $p->entries, 3, "parsed three entries");
+	} else {
+		SKIP: { skip 'DBD::SQLite or XML::Feed missing', 2 };
 	}
 }
 
diff --git a/t/indexlevels-mirror.t b/t/indexlevels-mirror.t
index e25b827..cac7050 100644
--- a/t/indexlevels-mirror.t
+++ b/t/indexlevels-mirror.t
@@ -10,8 +10,7 @@ require './t/common.perl';
 require_git(2.6);
 my $this = (split('/', __FILE__))[-1];
 
-# TODO: remove Search::Xapian as a requirement for basic
-foreach my $mod (qw(DBD::SQLite Search::Xapian)) {
+foreach my $mod (qw(DBD::SQLite)) {
 	eval "require $mod";
 	plan skip_all => "$mod missing for $this" if $@;
 }
@@ -47,8 +46,11 @@ sub import_index_incremental {
 	$im->done;
 
 	# index master (required for v1)
-	is(system($index, $ibx->{mainrepo}), 0, 'index master OK');
-	my $ro_master = PublicInbox::Inbox->new({mainrepo => $ibx->{mainrepo}});
+	is(system($index, $ibx->{mainrepo}, "-L$level"), 0, 'index master OK');
+	my $ro_master = PublicInbox::Inbox->new({
+		mainrepo => $ibx->{mainrepo},
+		indexlevel => $level
+	});
 	my ($nr, $msgs) = $ro_master->recent;
 	is($nr, 1, 'only one message in master, so far');
 	is($msgs->[0]->{mid}, 'm@1', 'first message in master indexed');
@@ -75,7 +77,10 @@ sub import_index_incremental {
 	is(system($index, $mirror), 0, "v$v index mirror OK");
 
 	# read-only access
-	my $ro_mirror = PublicInbox::Inbox->new({mainrepo => $mirror});
+	my $ro_mirror = PublicInbox::Inbox->new({
+		mainrepo => $mirror,
+		indexlevel => 'basic'
+	});
 	($nr, $msgs) = $ro_mirror->recent;
 	is($nr, 1, 'only one message, so far');
 	is($msgs->[0]->{mid}, 'm@1', 'read first message');
@@ -94,7 +99,7 @@ sub import_index_incremental {
 		['m@1','m@2'], 'got both messages in mirror');
 
 	# incremental index master (required for v1)
-	is(system($index, $ibx->{mainrepo}), 0, 'index master OK');
+	is(system($index, $ibx->{mainrepo}, "-L$level"), 0, 'index master OK');
 	($nr, $msgs) = $ro_master->recent;
 	is($nr, 2, '2nd message seen in master');
 	is_deeply([sort { $a cmp $b } map { $_->{mid} } @$msgs],
diff --git a/t/nntp.t b/t/nntp.t
index c4b9737..7ab801c 100644
--- a/t/nntp.t
+++ b/t/nntp.t
@@ -4,7 +4,7 @@ use strict;
 use warnings;
 use Test::More;
 
-foreach my $mod (qw(DBD::SQLite Search::Xapian Data::Dumper)) {
+foreach my $mod (qw(DBD::SQLite Data::Dumper)) {
 	eval "require $mod";
 	plan skip_all => "$mod missing for nntp.t" if $@;
 }
diff --git a/t/nntpd.t b/t/nntpd.t
index b8a84a0..c7ea319 100644
--- a/t/nntpd.t
+++ b/t/nntpd.t
@@ -3,7 +3,7 @@
 use strict;
 use warnings;
 use Test::More;
-foreach my $mod (qw(DBD::SQLite Search::Xapian)) {
+foreach my $mod (qw(DBD::SQLite)) {
 	eval "require $mod";
 	plan skip_all => "$mod missing for nntpd.t" if $@;
 }
@@ -55,12 +55,13 @@ my $ibx = {
 	name => $group,
 	version => $version,
 	-primary_address => $addr,
+	indexlevel => 'basic',
 };
 $ibx = PublicInbox::Inbox->new($ibx);
 {
 	local $ENV{HOME} = $home;
 	my @cmd = ($init, $group, $mainrepo, 'http://example.com/', $addr);
-	push @cmd, "-V$version";
+	push @cmd, "-V$version", '-Lbasic';
 	is(system(@cmd), 0, 'init OK');
 	is(system(qw(git config), "--file=$home/.public-inbox/config",
 			"publicinbox.$group.newsgroup", $group),
@@ -71,8 +72,8 @@ $ibx = PublicInbox::Inbox->new($ibx);
 	if ($version == 2) {
 		$im = PublicInbox::V2Writable->new($ibx);
 	} elsif ($version == 1) {
-		my $git = PublicInbox::Git->new($mainrepo);
-		$im = PublicInbox::Import->new($git, 'test', $addr);
+		use_ok 'PublicInbox::V1Writable';
+		$im = PublicInbox::V1Writable->new($ibx);
 	} else {
 		die "unsupported version: $version";
 	}
@@ -261,6 +262,12 @@ EOF
 		is($rdr, waitpid($rdr, 0), 'reader done');
 		is($? >> 8, 0, 'no errors');
 	}
+	SKIP: {
+		my @of = `lsof -p $pid 2>/dev/null`;
+		skip('lsof broken', 1) if (!scalar(@of) || $?);
+		my @xap = grep m!Search/Xapian!, @of;
+		is_deeply(\@xap, [], 'Xapian not loaded in nntpd');
+	}
 	{
 		setsockopt($s, IPPROTO_TCP, TCP_NODELAY, 1);
 		syswrite($s, 'HDR List-id 1-');
diff --git a/t/over.t b/t/over.t
index 7a3c972..c0d9d5e 100644
--- a/t/over.t
+++ b/t/over.t
@@ -5,8 +5,7 @@ use warnings;
 use Test::More;
 use File::Temp qw/tempdir/;
 use Compress::Zlib qw(compress);
-# FIXME: allow using Over w/o Xapian
-foreach my $mod (qw(DBD::SQLite Search::Xapian)) {
+foreach my $mod (qw(DBD::SQLite)) {
 	eval "require $mod";
 	plan skip_all => "$mod missing for over.t" if $@;
 }
diff --git a/t/psgi_bad_mids.t b/t/psgi_bad_mids.t
index 5008f5b..71eefe5 100644
--- a/t/psgi_bad_mids.t
+++ b/t/psgi_bad_mids.t
@@ -7,7 +7,7 @@ use File::Temp qw/tempdir/;
 use PublicInbox::MIME;
 use PublicInbox::Config;
 use PublicInbox::WWW;
-my @mods = qw(DBD::SQLite Search::Xapian HTTP::Request::Common Plack::Test
+my @mods = qw(DBD::SQLite HTTP::Request::Common Plack::Test
 		URI::Escape Plack::Builder);
 foreach my $mod (@mods) {
 	eval "require $mod";
@@ -22,6 +22,7 @@ my $ibx = {
 	name => 'bad-mids',
 	version => 2,
 	-primary_address => 'test@example.com',
+	indexlevel => 'basic',
 };
 $ibx = PublicInbox::Inbox->new($ibx);
 my $im = PublicInbox::V2Writable->new($ibx, 1);
diff --git a/t/psgi_scan_all.t b/t/psgi_scan_all.t
index e9c439e..2f54c82 100644
--- a/t/psgi_scan_all.t
+++ b/t/psgi_scan_all.t
@@ -6,8 +6,7 @@ use Test::More;
 use Email::MIME;
 use File::Temp qw/tempdir/;
 use PublicInbox::Config;
-my @mods = qw(HTTP::Request::Common Plack::Test URI::Escape Search::Xapian
-	DBD::SQLite);
+my @mods = qw(HTTP::Request::Common Plack::Test URI::Escape DBD::SQLite);
 foreach my $mod (@mods) {
 	eval "require $mod";
 	plan skip_all => "$mod missing for psgi_scan_all.t" if $@;
@@ -26,6 +25,7 @@ foreach my $i (1..2) {
 		mainrepo => $mainrepo,
 		name => "test-$i",
 		version => 2,
+		indexlevel => 'basic',
 		-primary_address => $addr,
 	};
 	my $ibx = PublicInbox::Inbox->new($opt);
diff --git a/t/psgi_search.t b/t/psgi_search.t
index da6cc68..1adc1d6 100644
--- a/t/psgi_search.t
+++ b/t/psgi_search.t
@@ -8,13 +8,13 @@ use Email::MIME;
 use PublicInbox::Config;
 use PublicInbox::WWW;
 use bytes (); # only for bytes::length
-my @mods = qw(PublicInbox::SearchIdx HTTP::Request::Common Plack::Test
+my @mods = qw(Search::Xapian HTTP::Request::Common Plack::Test
 		URI::Escape Plack::Builder);
 foreach my $mod (@mods) {
 	eval "require $mod";
 	plan skip_all => "$mod missing for psgi_search.t" if $@;
 }
-use_ok $_ foreach @mods;
+use_ok $_ foreach (@mods, qw(PublicInbox::SearchIdx));
 my $tmpdir = tempdir('pi-psgi-search.XXXXXX', TMPDIR => 1, CLEANUP => 1);
 my $git_dir = "$tmpdir/a.git";
 
diff --git a/t/purge.t b/t/purge.t
index 574935e..c1e0e9a 100644
--- a/t/purge.t
+++ b/t/purge.t
@@ -6,7 +6,7 @@ use Test::More;
 use File::Temp qw/tempdir/;
 require './t/common.perl';
 require_git(2.6);
-my @mods = qw(IPC::Run DBI DBD::SQLite Search::Xapian);
+my @mods = qw(IPC::Run DBI DBD::SQLite);
 foreach my $mod (@mods) {
 	eval "require $mod";
 	plan skip_all => "missing $mod for t/purge.t" if $@;
diff --git a/t/search-thr-index.t b/t/search-thr-index.t
index ed3e4e7..848dc56 100644
--- a/t/search-thr-index.t
+++ b/t/search-thr-index.t
@@ -7,8 +7,9 @@ use Test::More;
 use File::Temp qw/tempdir/;
 use PublicInbox::MID qw(mids);
 use Email::MIME;
-eval { require PublicInbox::SearchIdx; };
-plan skip_all => "Xapian missing for search" if $@;
+eval { require Search::Xapian };
+plan skip_all => "Search::Xapian missing for search" if $@;
+require PublicInbox::SearchIdx;
 my $tmpdir = tempdir('pi-search-thr-index.XXXXXX', TMPDIR => 1, CLEANUP => 1);
 my $git_dir = "$tmpdir/a.git";
 
diff --git a/t/search.t b/t/search.t
index 493e00d..a9d0c92 100644
--- a/t/search.t
+++ b/t/search.t
@@ -3,8 +3,9 @@
 use strict;
 use warnings;
 use Test::More;
-eval { require PublicInbox::SearchIdx; };
-plan skip_all => "Xapian missing for search" if $@;
+eval { require Search::Xapian };
+plan skip_all => "Search::Xapian missing for search" if $@;
+require PublicInbox::SearchIdx;
 use File::Temp qw/tempdir/;
 use Email::MIME;
 my $tmpdir = tempdir('pi-search-XXXXXX', TMPDIR => 1, CLEANUP => 1);
@@ -12,7 +13,7 @@ my $git_dir = "$tmpdir/a.git";
 my ($root_id, $last_id);
 
 is(0, system(qw(git init --shared -q --bare), $git_dir), "git init (main)");
-eval { PublicInbox::Search->new($git_dir) };
+eval { PublicInbox::Search->new($git_dir)->xdb };
 ok($@, "exception raised on non-existent DB");
 
 my $rw = PublicInbox::SearchIdx->new($git_dir, 1);
diff --git a/t/v1reindex.t b/t/v1reindex.t
index 33a36fa..402ecd7 100644
--- a/t/v1reindex.t
+++ b/t/v1reindex.t
@@ -209,8 +209,7 @@ ok(!-d $xap, 'Xapian directories removed again');
 	delete $ibx->{mm};
 	is_deeply([ $ibx->mm->minmax ], $minmax, 'minmax unchanged');
 	is($ibx->mm->num_highwater, 10, 'num_highwater as expected');
-	my $mset = $ibx->search->reopen->query('hello world', {mset=>1});
-	is($mset->size, 0, "no Xapian search results");
+	isnt($ibx->search, 'no search for basic');
 
 	my ($min, $max) = $ibx->mm->minmax;
 	is_deeply($ibx->mm->msg_range(\$min, $max), $msgmap, 'msgmap unchanged');
diff --git a/t/v2reindex.t b/t/v2reindex.t
index c416629..2a6fc55 100644
--- a/t/v2reindex.t
+++ b/t/v2reindex.t
@@ -225,8 +225,9 @@ ok(!-d $xap, 'Xapian directories removed again');
 	delete $ibx->{mm};
 	is_deeply([ $ibx->mm->minmax ], $minmax, 'minmax unchanged');
 	is($ibx->mm->num_highwater, 10, 'num_highwater as expected');
-	my $mset = $ibx->search->query('freedom', {mset=>1});
-	is($mset->size, 0, "search fails on indexlevel='basic'");
+
+	isnt($ibx->search, 'no search for basic');
+
 	for (<"$xap/*/*">) { $sizes{$ibx->{indexlevel}} += -s _ if -f $_ }
 	ok($sizes{medium} > $sizes{basic}, 'basic is smaller than medium');
 
diff --git a/t/watch_maildir_v2.t b/t/watch_maildir_v2.t
index 5f96891..b2cff4b 100644
--- a/t/watch_maildir_v2.t
+++ b/t/watch_maildir_v2.t
@@ -7,12 +7,12 @@ use Cwd;
 use PublicInbox::Config;
 require './t/common.perl';
 require_git(2.6);
-my @mods = qw(Filesys::Notify::Simple PublicInbox::V2Writable);
+my @mods = qw(Search::Xapian DBD::SQLite Filesys::Notify::Simple);
 foreach my $mod (@mods) {
 	eval "require $mod";
 	plan skip_all => "$mod missing for watch_maildir_v2.t" if $@;
 }
-
+require PublicInbox::V2Writable;
 my $tmpdir = tempdir('watch_maildir-v2-XXXXXX', TMPDIR => 1, CLEANUP => 1);
 my $mainrepo = "$tmpdir/v2";
 my $maildir = "$tmpdir/md";
-- 
EW


  parent reply index

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-05-15  6:33 [PATCH 0/6] make Search::Xapian optional for v2, NNTP, etc Eric Wong
2019-05-15  6:33 ` [PATCH 1/6] inbox: add ->over method to ease access Eric Wong
2019-05-15  6:33 ` [PATCH 2/6] nntp: use Inbox->over directly Eric Wong
2019-05-15  6:33 ` [PATCH 3/6] www: use Inbox->over where appropriate Eric Wong
2019-05-15  6:33 ` Eric Wong [this message]
2019-05-15  6:33 ` [PATCH 5/6] searchidx: do not create empty Xapian partitions for basic Eric Wong
2019-05-15  6:33 ` [PATCH 6/6] admin: improve warnings and errors for missing modules Eric Wong
2019-05-22  0:35 ` [PATCH 0/6] make Search::Xapian optional for v2, NNTP, etc Eric Wong

Reply instructions:

You may reply publically 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=20190515063354.52259-5-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

user/dev discussion of public-inbox itself

Archives are clonable:
	git clone --mirror https://public-inbox.org/meta
	git clone --mirror http://czquwvybam4bgbro.onion/meta
	git clone --mirror http://hjrcffqmbrq6wope.onion/meta
	git clone --mirror http://ou63pmih66umazou.onion/meta

Newsgroups are available over NNTP:
	nntp://news.public-inbox.org/inbox.comp.mail.public-inbox.meta
	nntp://ou63pmih66umazou.onion/inbox.comp.mail.public-inbox.meta
	nntp://czquwvybam4bgbro.onion/inbox.comp.mail.public-inbox.meta
	nntp://hjrcffqmbrq6wope.onion/inbox.comp.mail.public-inbox.meta
	nntp://news.gmane.org/gmane.mail.public-inbox.general

 note: .onion URLs require Tor: https://www.torproject.org/

AGPL code for this site: git clone https://public-inbox.org/ public-inbox