2 # Copyright (C) all contributors <meta@public-inbox.org>
3 # License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
5 use PublicInbox::TestCommon;
6 use IO::Uncompress::Gunzip qw(gunzip);
8 use PublicInbox::Inbox;
9 my @mods = qw(DBD::SQLite Xapian HTTP::Request::Common Plack::Test
10 URI::Escape Plack::Builder);
12 use_ok($_) for (qw(HTTP::Request::Common Plack::Test));
13 use_ok 'PublicInbox::WWW';
14 use_ok 'PublicInbox::SearchIdx';
15 my ($tmpdir, $for_destroy) = tmpdir();
16 local $ENV{TZ} = 'UTC';
18 my $digits = '10010260936330';
19 my $ua = 'Pine.LNX.4.10';
20 my $mid = "$ua.$digits.2460-100000\@penguin.transmeta.com";
21 my $ibx = create_inbox '26-git', indexlevel => 'full', tmpdir => "$tmpdir/1",
24 # n.b. these headers are not properly RFC2047-encoded
25 $im->add(PublicInbox::Eml->new(<<EOF)) or BAIL_OUT;
28 From: Ævar Arnfjörð Bjarmason <avarab\@example>
29 To: git\@vger.kernel.org
33 $im->add(PublicInbox::Eml->new(<<"")) or BAIL_OUT;
34 Message-ID: <reply\@asdf>
35 From: replier <r\@example.com>
39 $im->add(PublicInbox::Eml->new(<<'EOF')) or BAIL_OUT;
41 Message-ID: <blank-subject@example.com>
42 From: blank subject <blank-subject@example.com>
43 To: git@vger.kernel.org
47 $im->add(PublicInbox::Eml->new(<<'EOF')) or BAIL_OUT;
48 Message-ID: <no-subject-at-all@example.com>
49 From: no subject at all <no-subject-at-all@example.com>
50 To: git@vger.kernel.org
53 $im->add(PublicInbox::Eml->new(<<'EOF')) or BAIL_OUT;
54 Message-ID: <ampersand@example.com>
56 To: git@vger.kernel.org
57 Subject: git & ampersand
66 my $cfg = cfg_new $tmpdir, <<EOF;
68 address = git\@vger.kernel.org
69 inboxdir = $ibx->{inboxdir}
71 my $www = PublicInbox::WWW->new($cfg);
72 test_psgi(sub { $www->call(@_) }, sub {
75 my $approxidate = 'now';
76 for my $req ('/test/?q=%C3%86var', '/test/?q=%25C3%2586var') {
77 $res = $cb->(GET($req."+d:..$approxidate"));
78 $html = $res->content;
79 like($html, qr/<title>Ævar d:\.\.\Q$approxidate\E/,
80 'HTML escaped in title, "d:..$APPROXIDATE" preserved');
81 my @res = ($html =~ m/\?q=(.+var)\+d:\.\.\Q$approxidate\E/g);
82 ok(scalar(@res), 'saw query strings');
83 my %uniq = map { $_ => 1 } @res;
84 is(1, scalar keys %uniq, 'all query values identical in HTML');
85 is('%C3%86var', (keys %uniq)[0], 'matches original query');
86 ok(index($html, 'by Ævar Arnfjörð Bjarmason')
87 >= 0, "displayed Ævar's name properly in HTML");
88 like($html, qr/download mbox\.gz: .*?"full threads"/s,
89 '"full threads" download option shown');
91 like($html, qr/Initial query\b.*?returned no.results, used:.*instead/s,
92 'noted retry on double-escaped query {-uxs_retried}');
95 local $SIG{__WARN__} = sub { push @$warn, @_ };
96 $res = $cb->(GET('/test/?q=s:test&l=5e'));
97 is($res->code, 200, 'successful search result');
98 is_deeply([], $warn, 'no warnings from non-numeric comparison');
100 $res = $cb->(GET('/test/?&q=s:test'));
101 is($res->code, 200, 'successful search result');
102 is_deeply([], $warn, 'no warnings from black parameter');
104 $res = $cb->(POST('/test/?q=s:bogus&x=m'));
105 is($res->code, 404, 'failed search result gives 404');
106 like($res->content, qr/No results found/, "`No results' shown");
107 is_deeply([], $warn, 'no warnings');
109 my $mid_re = qr/\Q$mid\E/o;
110 while (length($digits) > 8) {
111 $res = $cb->(GET("/test/$ua.$digits/"));
112 is($res->code, 300, 'partial match found while truncated');
113 like($res->content, qr/\b1 partial match found\b/);
114 like($res->content, $mid_re, 'found mid in response');
117 $res = $cb->(GET("/test/$mid/"));
118 $html = $res->content;
119 like($html, qr/\bFrom: Ævar /,
120 "displayed Ævar's name properly in permalink From:");
121 unlike($html, qr/Ã/, 'no raw octets in permalink HTML');
123 $res = $cb->(GET('/test/'));
124 $html = $res->content;
125 like($html, qr/\bhref="no-subject-at-all[^>]+>\(no subject\)</,
126 'subject-less message linked from "/$INBOX/"');
127 like($html, qr/\bhref="blank-subject[^>]+>\(no subject\)</,
128 'blank subject message linked from "/$INBOX/"');
129 like($html, qr/test Ævar/,
130 "displayed Ævar's name properly in topic view");
132 $res = $cb->(GET('/test/?q=tc:git'));
133 like($html, qr/\bhref="no-subject-at-all[^>]+>\(no subject\)</,
134 'subject-less message linked from "/$INBOX/?q=..."');
135 like($html, qr/\bhref="blank-subject[^>]+>\(no subject\)</,
136 'blank subject message linked from "/$INBOX/?q=..."');
137 $res = $cb->(GET('/test/no-subject-at-all@example.com/raw'));
138 like($res->header('Content-Disposition'),
139 qr/filename=no-subject\.txt/);
140 $res = $cb->(GET('/test/no-subject-at-all@example.com/t.mbox.gz'));
141 like($res->header('Content-Disposition'),
142 qr/filename=no-subject\.mbox\.gz/);
144 # "full threads" mbox.gz download
145 $res = $cb->(POST("/test/?q=s:test+d:..$approxidate&x=m&t"));
146 is($res->code, 200, 'successful mbox download with threads');
147 gunzip(\($res->content) => \(my $before));
148 is_deeply([ "Message-ID: <$mid>\n", "Message-ID: <reply\@asdf>\n" ],
149 [ grep(/^Message-ID:/m, split(/^/m, $before)) ],
152 # clobber has_threadid to emulate old versions:
154 my $sidx = PublicInbox::SearchIdx->new($ibx, 0);
155 my $xdb = $sidx->idx_acquire;
156 $xdb->set_metadata('has_threadid', '0');
159 $cfg->each_inbox(sub { delete $_[0]->{search} });
160 $res = $cb->(GET('/test/?q=s:test'));
161 is($res->code, 200, 'successful search w/o has_threadid');
162 unlike($html, qr/download mbox\.gz: .*?"full threads"/s,
163 '"full threads" download option not shown w/o has_threadid');
165 # in case somebody uses curl to bypass <form>
166 $res = $cb->(POST("/test/?q=s:test+d:..$approxidate&x=m&t"));
167 is($res->code, 200, 'successful mbox download w/ threads');
168 gunzip(\($res->content) => \(my $after));
169 isnt($before, $after);
171 $res = $cb->(GET('/test/?q=git+%26+ampersand&x=A'));
172 is $res->code, 200, 'Atom hit with ampersand';
173 unlike $res->content, qr/git\+&\+ampersand/, '& is HTML-escaped';
175 $res = $cb->(GET('/test/?q=%22hi+%2b%2b%2b+b/foo%22&x=A'));
176 is $res->code, 200, 'slashes and plusses search hit';
177 like $res->content, qr!q=%22hi\+(?:%2[bB]){3}\+b/foo%22!,
178 '+ and " escaped, but slash not escaped in query';
180 $res = $cb->(GET(q{/test/?q=%22s'more%22&x=A}));
181 is $res->code, 200, 'single quote inside phrase';
182 # TODO: more tests and odd cases