X-Git-Url: http://www.git.stargrave.org/?p=public-inbox.git;a=blobdiff_plain;f=t%2Fpsgi_search.t;h=3da93eda718cd7335c2b1f747f2a7f16e2bf880e;hp=1df38691c1c07a1b3eb825683d691325c3be1316;hb=refs%2Fheads%2Fmaster;hpb=3d41aa23f35501ca92aab8aa42980fa73f7fa74f diff --git a/t/psgi_search.t b/t/psgi_search.t index 1df38691..27946391 100644 --- a/t/psgi_search.t +++ b/t/psgi_search.t @@ -1,71 +1,162 @@ -# Copyright (C) 2017-2018 all contributors +#!perl -w +# Copyright (C) all contributors # License: AGPL-3.0+ use strict; -use warnings; -use Test::More; -use File::Temp qw/tempdir/; -use Email::MIME; +use v5.10.1; +use PublicInbox::TestCommon; +use IO::Uncompress::Gunzip qw(gunzip); +use PublicInbox::Eml; use PublicInbox::Config; -use PublicInbox::WWW; -my @mods = qw(PublicInbox::SearchIdx HTTP::Request::Common Plack::Test +use PublicInbox::Inbox; +my @mods = qw(DBD::SQLite 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; -my $tmpdir = tempdir('pi-psgi-search.XXXXXX', TMPDIR => 1, CLEANUP => 1); -my $git_dir = "$tmpdir/a.git"; - -is(0, system(qw(git init -q --bare), $git_dir), "git init (main)"); -my $rw = PublicInbox::SearchIdx->new($git_dir, 1); -ok($rw, "search indexer created"); -my $data = <<'EOF'; -Subject: test -Message-Id: -From: Ævar Arnfjörð Bjarmason -To: git@vger.kernel.org +require_mods(@mods); +use_ok($_) for (qw(HTTP::Request::Common Plack::Test)); +use_ok 'PublicInbox::WWW'; +use_ok 'PublicInbox::SearchIdx'; +my ($tmpdir, $for_destroy) = tmpdir(); +local $ENV{TZ} = 'UTC'; + +my $digits = '10010260936330'; +my $ua = 'Pine.LNX.4.10'; +my $mid = "$ua.$digits.2460-100000\@penguin.transmeta.com"; +my $ibx = create_inbox 'git', indexlevel => 'full', tmpdir => "$tmpdir/1", sub { + my ($im) = @_; + # n.b. these headers are not properly RFC2047-encoded + $im->add(PublicInbox::Eml->new(< +From: Ævar Arnfjörð Bjarmason +To: git\@vger.kernel.org EOF -my $num = 0; -# nb. using internal API, fragile! -my $xdb = $rw->_xdb_acquire; -$xdb->begin_transaction; + $im->add(PublicInbox::Eml->new(<<"")) or BAIL_OUT; +Message-ID: +From: replier +In-Reply-To: <$mid> +Subject: mismatch + + $im->add(PublicInbox::Eml->new(<<'EOF')) or BAIL_OUT; +Subject: +Message-ID: +From: blank subject +To: git@vger.kernel.org + +EOF -foreach (reverse split(/\n\n/, $data)) { - $_ .= "\n"; - my $mime = Email::MIME->new(\$_); - my $bytes = bytes::length($mime->as_string); - my $doc_id = $rw->add_message($mime, $bytes, ++$num, 'ignored'); - my $mid = $mime->header('Message-Id'); - ok($doc_id, 'message added: '. $mid); -} + $im->add(PublicInbox::Eml->new(<<'EOF')) or BAIL_OUT; +Message-ID: +From: no subject at all +To: git@vger.kernel.org -$xdb->commit_transaction; -$rw = undef; +EOF +}; my $cfgpfx = "publicinbox.test"; -my $config = PublicInbox::Config->new({ - "$cfgpfx.address" => 'git@vger.kernel.org', - "$cfgpfx.mainrepo" => $git_dir, -}); -my $www = PublicInbox::WWW->new($config); +my $cfg = PublicInbox::Config->new(\<{inboxdir} +EOF +my $www = PublicInbox::WWW->new($cfg); test_psgi(sub { $www->call(@_) }, sub { my ($cb) = @_; - my $res; - $res = $cb->(GET('/test/?q=%C3%86var')); - my $html = $res->content; - like($html, qr/Ævar - /, 'HTML escaped in title'); - my @res = ($html =~ m/\?q=(.+var)\b/g); - ok(scalar(@res), 'saw query strings'); - my %uniq = map { $_ => 1 } @res; - is(1, scalar keys %uniq, 'all query values identical in HTML'); - is('%C3%86var', (keys %uniq)[0], 'matches original query'); - ok(index($html, 'by Ævar Arnfjörð Bjarmason') >= 0, - "displayed Ævar's name properly in HTML"); + my ($html, $res); + my $approxidate = 'now'; + for my $req ('/test/?q=%C3%86var', '/test/?q=%25C3%2586var') { + $res = $cb->(GET($req."+d:..$approxidate")); + $html = $res->content; + like($html, qr/<title>Ævar d:\.\.\Q$approxidate\E/, + 'HTML escaped in title, "d:..$APPROXIDATE" preserved'); + my @res = ($html =~ m/\?q=(.+var)\+d:\.\.\Q$approxidate\E/g); + ok(scalar(@res), 'saw query strings'); + my %uniq = map { $_ => 1 } @res; + is(1, scalar keys %uniq, 'all query values identical in HTML'); + is('%C3%86var', (keys %uniq)[0], 'matches original query'); + ok(index($html, 'by Ævar Arnfjörð Bjarmason') + >= 0, "displayed Ævar's name properly in HTML"); + like($html, qr/download mbox\.gz: .*?"full threads"/s, + '"full threads" download option shown'); + } + like($html, qr/Initial query\b.*?returned no.results, used:.*instead/s, + 'noted retry on double-escaped query {-uxs_retried}'); + + my $warn = []; + local $SIG{__WARN__} = sub { push @$warn, @_ }; + $res = $cb->(GET('/test/?q=s:test&l=5e')); + is($res->code, 200, 'successful search result'); + is_deeply([], $warn, 'no warnings from non-numeric comparison'); + + $res = $cb->(GET('/test/?&q=s:test')); + is($res->code, 200, 'successful search result'); + is_deeply([], $warn, 'no warnings from black parameter'); + + $res = $cb->(POST('/test/?q=s:bogus&x=m')); + is($res->code, 404, 'failed search result gives 404'); + like($res->content, qr/No results found/, "`No results' shown"); + is_deeply([], $warn, 'no warnings'); + + my $mid_re = qr/\Q$mid\E/o; + while (length($digits) > 8) { + $res = $cb->(GET("/test/$ua.$digits/")); + is($res->code, 300, 'partial match found while truncated'); + like($res->content, qr/\b1 partial match found\b/); + like($res->content, $mid_re, 'found mid in response'); + chop($digits); + } + $res = $cb->(GET("/test/$mid/")); + $html = $res->content; + like($html, qr/\bFrom: Ævar /, + "displayed Ævar's name properly in permalink From:"); + unlike($html, qr/Ã/, 'no raw octets in permalink HTML'); + + $res = $cb->(GET('/test/')); + $html = $res->content; + like($html, qr/\bhref="no-subject-at-all[^>]+>\(no subject\)</, + 'subject-less message linked from "/$INBOX/"'); + like($html, qr/\bhref="blank-subject[^>]+>\(no subject\)</, + 'blank subject message linked from "/$INBOX/"'); + like($html, qr/test Ævar/, + "displayed Ævar's name properly in topic view"); + + $res = $cb->(GET('/test/?q=tc:git')); + like($html, qr/\bhref="no-subject-at-all[^>]+>\(no subject\)</, + 'subject-less message linked from "/$INBOX/?q=..."'); + like($html, qr/\bhref="blank-subject[^>]+>\(no subject\)</, + 'blank subject message linked from "/$INBOX/?q=..."'); + $res = $cb->(GET('/test/no-subject-at-all@example.com/raw')); + like($res->header('Content-Disposition'), + qr/filename=no-subject\.txt/); + $res = $cb->(GET('/test/no-subject-at-all@example.com/t.mbox.gz')); + like($res->header('Content-Disposition'), + qr/filename=no-subject\.mbox\.gz/); + + # "full threads" mbox.gz download + $res = $cb->(POST("/test/?q=s:test+d:..$approxidate&x=m&t")); + is($res->code, 200, 'successful mbox download with threads'); + gunzip(\($res->content) => \(my $before)); + is_deeply([ "Message-ID: <$mid>\n", "Message-ID: <reply\@asdf>\n" ], + [ grep(/^Message-ID:/m, split(/^/m, $before)) ], + 'got full thread'); + + # clobber has_threadid to emulate old versions: + { + my $sidx = PublicInbox::SearchIdx->new($ibx, 0); + my $xdb = $sidx->idx_acquire; + $xdb->set_metadata('has_threadid', '0'); + $sidx->idx_release; + } + $cfg->each_inbox(sub { delete $_[0]->{search} }); + $res = $cb->(GET('/test/?q=s:test')); + is($res->code, 200, 'successful search w/o has_threadid'); + unlike($html, qr/download mbox\.gz: .*?"full threads"/s, + '"full threads" download option not shown w/o has_threadid'); + + # in case somebody uses curl to bypass <form> + $res = $cb->(POST("/test/?q=s:test+d:..$approxidate&x=m&t")); + is($res->code, 200, 'successful mbox download w/ threads'); + gunzip(\($res->content) => \(my $after)); + isnt($before, $after); }); done_testing(); - -1;