Merge branch 'master' into cvs

3 files changed, 318 insertions, 2 deletions

diff --git a/t/prune.t b/t/prune.t
new file mode 100755
index 000000000..8c3925e9e
--- /dev/null
+++ b/t/prune.t
@@ -0,0 +1,23 @@
+#!/usr/bin/perl
+use warnings;
+use strict;
+use Test::More tests => 6;
+use File::Path qw(make_path remove_tree);
+
+BEGIN { use_ok("IkiWiki"); }
+BEGIN { use_ok("IkiWiki::Render"); }
+
+%config=IkiWiki::defaultconfig();
+
+remove_tree("t/tmp");
+
+make_path("t/tmp/srcdir/a/b/c");
+make_path("t/tmp/srcdir/d/e/f");
+writefile("a/b/c/d.mdwn", "t/tmp/srcdir", "foo");
+writefile("d/e/f/g.mdwn", "t/tmp/srcdir", "foo");
+IkiWiki::prune("t/tmp/srcdir/d/e/f/g.mdwn");
+ok(-d "t/tmp/srcdir");
+ok(! -e "t/tmp/srcdir/d");
+IkiWiki::prune("t/tmp/srcdir/a/b/c/d.mdwn", "t/tmp/srcdir");
+ok(-d "t/tmp/srcdir");
+ok(! -e "t/tmp/srcdir/a");
diff --git a/t/syntax.t b/t/syntax.t
index a3760a2b2..b7c6efd58 100755
--- a/t/syntax.t
+++ b/t/syntax.t
@@ -5,8 +5,9 @@ use Test::More;
 
 my @progs="ikiwiki.in";
 my @libs="IkiWiki.pm";
-# monotone, external, amazon_s3 skipped since they need perl modules
-push @libs, map { chomp; $_ } `find IkiWiki -type f -name \\*.pm | grep -v monotone.pm | grep -v external.pm | grep -v amazon_s3.pm | grep -v po.pm`;
+# monotone, external, amazon_s3, po, and cvs
+# skipped since they need perl modules
+push @libs, map { chomp; $_ } `find IkiWiki -type f -name \\*.pm | grep -v monotone.pm | grep -v external.pm | grep -v amazon_s3.pm | grep -v po.pm | grep -v cvs.pm`;
 push @libs, 'IkiWiki/Plugin/skeleton.pm.example';
 
 plan(tests => (@progs + @libs));
diff --git a/t/trail.t b/t/trail.t
new file mode 100755
index 000000000..dce3b3c7e
--- /dev/null
+++ b/t/trail.t
@@ -0,0 +1,292 @@
+#!/usr/bin/perl
+use warnings;
+use strict;
+use Test::More 'no_plan';
+use IkiWiki;
+
+sub check_trail {
+	my $file=shift;
+	my $expected=shift;
+	my $trailname=shift || qr/\w+/;
+	my $blob=readfile("t/tmp/out/$file");
+	my ($trailline)=$blob=~/^trail=$trailname\s+(.*)$/m;
+	is($trailline, $expected, "expected $expected in $file");
+}
+
+sub check_no_trail {
+	my $file=shift;
+	my $trailname=shift || qr/\w+/;
+	my $blob=readfile("t/tmp/out/$file");
+	my ($trailline)=$blob=~/^trail=$trailname\s+(.*)$/m;
+	$trailline="" unless defined $trailline;
+	ok($trailline !~ /^trail=$trailname\s+/, "no trail $trailname in $file");
+}
+
+my $blob;
+
+ok(! system("rm -rf t/tmp"));
+ok(! system("mkdir t/tmp"));
+
+# Write files with a date in the past, so that when we refresh,
+# the update is detected.
+sub write_old_file {
+	my $name = shift;
+	my $content = shift;
+
+	writefile($name, "t/tmp/in", $content);
+	ok(utime(333333333, 333333333, "t/tmp/in/$name"));
+}
+
+# Use a rather stylized template to override the default rendering, to make
+# it easy to search for the desired results
+write_old_file("templates/trails.tmpl", <<EOF
+<TMPL_LOOP TRAILLOOP>
+<TMPL_IF __FIRST__><nav></TMPL_IF>
+<div>
+trail=<TMPL_VAR TRAILPAGE> n=<TMPL_VAR NEXTPAGE> p=<TMPL_VAR PREVPAGE>
+</div>
+<div>
+<TMPL_IF PREVURL>
+<a href="<TMPL_VAR PREVURL>">&lt; <TMPL_VAR PREVTITLE></a>
+</TMPL_IF> |
+<a href="<TMPL_VAR TRAILURL>">^ <TMPL_VAR TRAILTITLE> ^</a>
+| <TMPL_IF NEXTURL>
+<a href="<TMPL_VAR NEXTURL>"><TMPL_VAR NEXTTITLE> &gt;</a>
+</TMPL_IF>
+</div>
+<TMPL_IF __LAST__></nav></TMPL_IF>
+</TMPL_LOOP>
+EOF
+);
+write_old_file("badger.mdwn", "[[!meta title=\"The Breezy Badger\"]]\ncontent of badger");
+write_old_file("mushroom.mdwn", "content of mushroom");
+write_old_file("snake.mdwn", "content of snake");
+write_old_file("ratty.mdwn", "content of ratty");
+write_old_file("mr_toad.mdwn", "content of mr toad");
+write_old_file("add.mdwn", '[[!trailitems pagenames="add/a add/b add/c add/d add/e"]]');
+write_old_file("add/b.mdwn", "b");
+write_old_file("add/d.mdwn", "d");
+write_old_file("del.mdwn", '[[!trailitems pages="del/*" sort=title]]');
+write_old_file("del/a.mdwn", "a");
+write_old_file("del/b.mdwn", "b");
+write_old_file("del/c.mdwn", "c");
+write_old_file("del/d.mdwn", "d");
+write_old_file("del/e.mdwn", "e");
+write_old_file("self_referential.mdwn", '[[!trailitems pagenames="self_referential" circular=yes]]');
+write_old_file("sorting/linked.mdwn", "linked");
+write_old_file("sorting/a/b.mdwn", "a/b");
+write_old_file("sorting/a/c.mdwn", "a/c");
+write_old_file("sorting/z/a.mdwn", "z/a");
+write_old_file("sorting/beginning.mdwn", "beginning");
+write_old_file("sorting/middle.mdwn", "middle");
+write_old_file("sorting/end.mdwn", "end");
+write_old_file("sorting/new.mdwn", "new");
+write_old_file("sorting/old.mdwn", "old");
+write_old_file("sorting/ancient.mdwn", "ancient");
+# These three need to be in the appropriate age order
+ok(utime(333333333, 333333333, "t/tmp/in/sorting/new.mdwn"));
+ok(utime(222222222, 222222222, "t/tmp/in/sorting/old.mdwn"));
+ok(utime(111111111, 111111111, "t/tmp/in/sorting/ancient.mdwn"));
+write_old_file("sorting/linked2.mdwn", "linked2");
+# This initially uses the default sort order: age for the inline, and path
+# for trailitems. We change it later.
+write_old_file("sorting.mdwn",
+	'[[!traillink linked]] ' .
+	'[[!trailitems pages="sorting/z/a or sorting/a/b or sorting/a/c"]] ' .
+	'[[!trailitems pagenames="sorting/beginning sorting/middle sorting/end"]] ' .
+	'[[!inline pages="sorting/old or sorting/ancient or sorting/new" trail="yes"]] ' .
+	'[[!traillink linked2]]');
+write_old_file("limited/a.mdwn", "a");
+write_old_file("limited/b.mdwn", "b");
+write_old_file("limited/c.mdwn", "c");
+write_old_file("limited/d.mdwn", "d");
+write_old_file("limited.mdwn",
+	'[[!inline pages="limited/*" trail="yes" show=2 sort=title]]');
+write_old_file("untrail/a.mdwn", "a");
+write_old_file("untrail/b.mdwn", "b");
+write_old_file("untrail.mdwn", "[[!traillink a]] [[!traillink b]]");
+write_old_file("retitled/a.mdwn", "a");
+write_old_file("retitled.mdwn",
+	'[[!meta title="the old title"]][[!traillink a]]');
+
+write_old_file("meme.mdwn", <<EOF
+[[!trail]]
+* [[!traillink badger]]
+* [[!traillink badger text="This is a link to badger, with a title"]]
+* [[!traillink That_is_the_badger|badger]]
+* [[!traillink badger]]
+* [[!traillink mushroom]]
+* [[!traillink mushroom]]
+* [[!traillink snake]]
+* [[!traillink snake]]
+EOF
+);
+
+write_old_file("wind_in_the_willows.mdwn", <<EOF
+[[!trailoptions circular=yes sort=title]]
+[[!trailitems pages="ratty or badger or mr_toad"]]
+[[!trailitem moley]]
+EOF
+);
+
+ok(! system("make -s ikiwiki.out"));
+
+my $command = "perl -I. ./ikiwiki.out -set usedirs=0 -plugin trail -plugin inline -url=http://example.com -cgiurl=http://example.com/ikiwiki.cgi -rss -atom -underlaydir=underlays/basewiki -set underlaydirbase=underlays -templatedir=templates t/tmp/in t/tmp/out -verbose";
+
+ok(! system($command));
+
+ok(! system("$command -refresh"));
+
+$blob = readfile("t/tmp/out/meme.html");
+ok($blob =~ /<a href="(\.\/)?badger.html">badger<\/a>/m);
+ok($blob =~ /<a href="(\.\/)?badger.html">This is a link to badger, with a title<\/a>/m);
+ok($blob =~ /<a href="(\.\/)?badger.html">That is the badger<\/a>/m);
+
+check_trail("badger.html", "n=mushroom p=", "meme");
+check_trail("badger.html", "n=mr_toad p=ratty", "wind_in_the_willows");
+
+ok(! -f "t/tmp/out/moley.html");
+
+check_trail("mr_toad.html", "n=ratty p=badger", "wind_in_the_willows");
+check_no_trail("mr_toad.html", "meme");
+# meta title is respected for pages that have one
+$blob = readfile("t/tmp/out/mr_toad.html");
+ok($blob =~ /">&lt; The Breezy Badger<\/a>/m);
+# pagetitle for pages that don't
+ok($blob =~ /">ratty &gt;<\/a>/m);
+
+check_no_trail("ratty.html", "meme");
+check_trail("ratty.html", "n=badger p=mr_toad", "wind_in_the_willows");
+
+check_trail("mushroom.html", "n=snake p=badger", "meme");
+check_no_trail("mushroom.html", "wind_in_the_willows");
+
+check_trail("snake.html", "n= p=mushroom", "meme");
+check_no_trail("snake.html", "wind_in_the_willows");
+
+check_trail("self_referential.html", "n= p=", "self_referential");
+
+check_trail("add/b.html", "n=add/d p=", "add");
+check_trail("add/d.html", "n= p=add/b", "add");
+ok(! -f "t/tmp/out/add/a.html");
+ok(! -f "t/tmp/out/add/c.html");
+ok(! -f "t/tmp/out/add/e.html");
+
+check_trail("del/a.html", "n=del/b p=");
+check_trail("del/b.html", "n=del/c p=del/a");
+check_trail("del/c.html", "n=del/d p=del/b");
+check_trail("del/d.html", "n=del/e p=del/c");
+check_trail("del/e.html", "n= p=del/d");
+
+check_trail("sorting/linked.html", "n=sorting/a/b p=");
+check_trail("sorting/a/b.html", "n=sorting/a/c p=sorting/linked");
+check_trail("sorting/a/c.html", "n=sorting/z/a p=sorting/a/b");
+check_trail("sorting/z/a.html", "n=sorting/beginning p=sorting/a/c");
+check_trail("sorting/beginning.html", "n=sorting/middle p=sorting/z/a");
+check_trail("sorting/middle.html", "n=sorting/end p=sorting/beginning");
+check_trail("sorting/end.html", "n=sorting/new p=sorting/middle");
+check_trail("sorting/new.html", "n=sorting/old p=sorting/end");
+check_trail("sorting/old.html", "n=sorting/ancient p=sorting/new");
+check_trail("sorting/ancient.html", "n=sorting/linked2 p=sorting/old");
+check_trail("sorting/linked2.html", "n= p=sorting/ancient");
+
+# If the inline has a limited number of pages, the trail still contains
+# everything.
+$blob = readfile("t/tmp/out/limited.html");
+ok($blob =~ /<a href="(\.\/)?limited\/a.html">a<\/a>/m);
+ok($blob =~ /<a href="(\.\/)?limited\/b.html">b<\/a>/m);
+ok($blob !~ /<a href="(\.\/)?limited\/c.html">/m);
+ok($blob !~ /<a href="(\.\/)?limited\/d.html">/m);
+check_trail("limited/a.html", "n=limited/b p=");
+check_trail("limited/b.html", "n=limited/c p=limited/a");
+check_trail("limited/c.html", "n=limited/d p=limited/b");
+check_trail("limited/d.html", "n= p=limited/c");
+
+check_trail("untrail/a.html", "n=untrail/b p=");
+check_trail("untrail/b.html", "n= p=untrail/a");
+
+$blob = readfile("t/tmp/out/retitled/a.html");
+ok($blob =~ /\^ the old title \^/m);
+
+# Make some changes and refresh. These writefile calls don't set an
+# old mtime, so they're strictly newer than the "old" files.
+
+writefile("add/a.mdwn", "t/tmp/in", "a");
+writefile("add/c.mdwn", "t/tmp/in", "c");
+writefile("add/e.mdwn", "t/tmp/in", "e");
+ok(unlink("t/tmp/in/del/a.mdwn"));
+ok(unlink("t/tmp/in/del/c.mdwn"));
+ok(unlink("t/tmp/in/del/e.mdwn"));
+
+writefile("sorting.mdwn", "t/tmp/in",
+	readfile("t/tmp/in/sorting.mdwn") .
+	'[[!trailoptions sort="title" reverse="yes"]]'); 
+
+writefile("retitled.mdwn", "t/tmp/in",
+	'[[!meta title="the new title"]][[!traillink a]]');
+
+# If the inline has a limited number of pages, the trail still depends on
+# everything.
+writefile("limited.html", "t/tmp/out", "[this gets rebuilt]");
+writefile("limited/c.mdwn", "t/tmp/in", '[[!meta title="New C page"]]c');
+
+writefile("untrail.mdwn", "t/tmp/in", "no longer a trail");
+
+ok(! system("$command -refresh"));
+
+check_trail("add/a.html", "n=add/b p=");
+check_trail("add/b.html", "n=add/c p=add/a");
+check_trail("add/c.html", "n=add/d p=add/b");
+check_trail("add/d.html", "n=add/e p=add/c");
+check_trail("add/e.html", "n= p=add/d");
+
+check_trail("del/b.html", "n=del/d p=");
+check_trail("del/d.html", "n= p=del/b");
+ok(! -f "t/tmp/out/del/a.html");
+ok(! -f "t/tmp/out/del/c.html");
+ok(! -f "t/tmp/out/del/e.html");
+
+check_trail("sorting/old.html", "n=sorting/new p=");
+check_trail("sorting/new.html", "n=sorting/middle p=sorting/old");
+check_trail("sorting/middle.html", "n=sorting/linked2 p=sorting/new");
+check_trail("sorting/linked2.html", "n=sorting/linked p=sorting/middle");
+check_trail("sorting/linked.html", "n=sorting/end p=sorting/linked2");
+check_trail("sorting/end.html", "n=sorting/a/c p=sorting/linked");
+check_trail("sorting/a/c.html", "n=sorting/beginning p=sorting/end");
+check_trail("sorting/beginning.html", "n=sorting/a/b p=sorting/a/c");
+check_trail("sorting/a/b.html", "n=sorting/ancient p=sorting/beginning");
+check_trail("sorting/ancient.html", "n=sorting/z/a p=sorting/a/b");
+check_trail("sorting/z/a.html", "n= p=sorting/ancient");
+
+# If the inline has a limited number of pages, the trail still depends on
+# everything, so it gets rebuilt even though it doesn't strictly need it.
+# This means we could use it as a way to recompute the order of members
+# and the contents of their trail navbars, allowing us to fix the regression
+# described in [[bugs/trail excess dependencies]] without a full content
+# dependency.
+$blob = readfile("t/tmp/out/limited.html");
+ok($blob =~ /<a href="(\.\/)?limited\/a.html">a<\/a>/m);
+ok($blob =~ /<a href="(\.\/)?limited\/b.html">b<\/a>/m);
+ok($blob !~ /<a href="(\.\/)?limited\/c.html">/m);
+ok($blob !~ /<a href="(\.\/)?limited\/d.html">/m);
+check_trail("limited/a.html", "n=limited/b p=");
+check_trail("limited/b.html", "n=limited/c p=limited/a");
+check_trail("limited/c.html", "n=limited/d p=limited/b");
+check_trail("limited/d.html", "n= p=limited/c");
+# Also, b and d should pick up the change to c. This regressed with the
+# change to using a presence dependency.
+$blob = readfile("t/tmp/out/limited/b.html");
+ok($blob =~ /New C page &gt;/m);
+$blob = readfile("t/tmp/out/limited/d.html");
+ok($blob =~ /&lt; New C page/m);
+
+# Members of a retitled trail should pick up that change.
+# This regressed with the change to using a presence dependency.
+$blob = readfile("t/tmp/out/retitled/a.html");
+ok($blob =~ /\^ the new title \^/m);
+
+# untrail is no longer a trail, so these are no longer in it.
+check_no_trail("untrail/a.html");
+check_no_trail("untrail/b.html");
+
+ok(! system("rm -rf t/tmp"));