From d0755b0e3693b8497e7232cb31a53686a3181eb8 Mon Sep 17 00:00:00 2001
From: Eric Wong <normalperson@yhbt.net>
Date: Sat, 5 Apr 2014 01:11:08 +0000
Subject: view: implement quote folding and flesh out tests

Unfortunately, quoting is often excessive, so hide multi-line quotes
by default and provide anchored links to full messages instead.
---
 lib/PublicInbox/View.pm | 61 +++++++++++++++++++++++++++++++++++++++++++------
 t/view.t                | 57 +++++++++++++++++++++++++++++++++++++++++++--
 2 files changed, 109 insertions(+), 9 deletions(-)

diff --git a/lib/PublicInbox/View.pm b/lib/PublicInbox/View.pm
index 08849a92..3d3b6b5e 100644
--- a/lib/PublicInbox/View.pm
+++ b/lib/PublicInbox/View.pm
@@ -9,17 +9,17 @@ use Encode::MIME::Header;
 
 # only one public function:
 sub as_html {
-	my ($class, $mime) = @_;
+	my ($class, $mime, $full_pfx) = @_;
 
 	headers_to_html_header($mime) .
-		multipart_text_as_html($mime) .
+		multipart_text_as_html($mime, $full_pfx) .
 		"</pre>\n";
 }
 
 # only private functions below.
 
 sub multipart_text_as_html {
-	my ($mime) = @_;
+	my ($mime, $full_pfx) = @_;
 	my $rv = "";
 	my $part_nr = 0;
 
@@ -33,15 +33,19 @@ sub multipart_text_as_html {
 			my $fn = $part->filename;
 
 			if ($part_nr > 0) {
-				defined($fn) or $fn = "part #$part_nr";
+				defined($fn) or $fn = "part #" . ($part_nr + 1);
 				$rv .= add_filename_line($fn);
 			}
 
-			# n.b. $part->body should already be decoded if text
-			$rv .= escapeHTML($part->body);
+			if (defined $full_pfx) {
+				$rv .= add_text_body_short($part, $part_nr,
+							$full_pfx);
+			} else {
+				$rv .= add_text_body_full($part, $part_nr);
+			}
 			$rv .= "\n" unless $rv =~ /\n\z/s;
 		} else {
-			$rv .= "-- part #$part_nr ";
+			$rv .= "-- part #" . ($part_nr + 1) . " ";
 			$rv .= escapeHTML($part_type);
 			$rv .= " skipped\n";
 		}
@@ -60,6 +64,49 @@ sub add_filename_line {
 	"$pad " . escapeHTML($fn) . " $pad\n";
 }
 
+sub add_text_body_short {
+	my ($part, $part_nr, $full_pfx) = @_;
+	my $n = 0;
+	my $s = escapeHTML($part->body);
+	$s =~ s!^((?:(?:&gt;[^\n]+)\n)+)!
+		my $cur = $1;
+		my @lines = split(/\n/, $cur);
+		if (@lines > 1) {
+			# show a short snippet of quoted text
+			$cur = join(' ', @lines);
+			$cur =~ s/&gt; ?//g;
+
+			my @sum = split(/\s+/, $cur);
+			$cur = '';
+			do {
+				$cur .= shift(@sum) . ' ';
+			} while (@sum && length($cur) < 68);
+			$cur=~ s/ \z/ .../;
+			"&gt; &lt;<a href=${full_pfx}#q${part_nr}_" . $n++ .
+				">$cur<\/a>&gt;";
+		} else {
+			$cur;
+		}
+	!emg;
+	$s;
+}
+
+sub add_text_body_full {
+	my ($part, $part_nr) = @_;
+	my $n = 0;
+	my $s = escapeHTML($part->body);
+	$s =~ s!^((?:(?:&gt;[^\n]+)\n)+)!
+		my $cur = $1;
+		my @lines = split(/\n/, $cur);
+		if (@lines > 1) {
+			"<a name=q${part_nr}_" . $n++ . ">$cur</a>";
+		} else {
+			$cur;
+		}
+	!emg;
+	$s;
+}
+
 sub trim_message_id {
 	my ($mid) = @_;
 	$mid =~ tr/<>//d;
diff --git a/t/view.t b/t/view.t
index df79bd49..fbdbd884 100644
--- a/t/view.t
+++ b/t/view.t
@@ -8,6 +8,17 @@ use PublicInbox::View;
 
 # plain text
 {
+	my $body = <<EOF;
+So and so wrote:
+> keep this inline
+
+OK
+
+> Long and wordy reply goes here and it is split across multiple lines.
+> We generate links to a separate full page where quoted-text is inline.
+
+hello world
+EOF
 	my $s = Email::Simple->create(
 		header => [
 			From => 'a@example.com',
@@ -16,7 +27,7 @@ use PublicInbox::View;
 			'Message-ID' => '<hello@example.com>',
 			Subject => 'this is a subject',
 		],
-		body => "hello world\n",
+		body => $body,
 	);
 	$s = Email::MIME->new($s->as_string);
 	my $html = PublicInbox::View->as_html($s);
@@ -24,6 +35,18 @@ use PublicInbox::View;
 	# ghetto
 	like($html, qr/<a href="?hello%40/s, "MID link present");
 	like($html, qr/hello world\b/, "body present");
+	like($html, qr/&gt; keep this inline/, "short quoted text is inline");
+	like($html, qr/<a name=[^>]+>&gt; Long and wordy/,
+		"long quoted text is anchored");
+
+	# short page
+	my $pfx = "http://example.com/test/full";
+	my $short = PublicInbox::View->as_html($s, $pfx);
+	like($short, qr/\n&gt; keep this inline/,
+		"short quoted text is inline");
+	like($short, qr/<a href=\Q$pfx\E#[^>]+>Long and wordy/,
+		"long quoted text is made into a link");
+	ok(length($short) < length($html), "short page is shorter");
 }
 
 # multipart crap
@@ -49,7 +72,37 @@ use PublicInbox::View;
 	);
 
 	my $html = PublicInbox::View->as_html($mime);
-	print $html;
+	like($html, qr/hi\n-+ part #2 -+\nbye\n/, "multipart split");
+}
+
+# multipart email with attached patch
+{
+	my $parts = [
+		Email::MIME->create(
+			attributes => { content_type => 'text/plain', },
+			body => 'hi, see attached patch',
+		),
+		Email::MIME->create(
+			attributes => {
+				content_type => 'text/plain',
+				filename => "foo.patch",
+			},
+			body => "--- a/file\n+++ b/file\n" .
+			        "@@ -49, 7 +49,34 @@\n",
+		),
+	];
+	my $mime = Email::MIME->create(
+		header_str => [
+			From => 'a@example.com',
+			Subject => '[PATCH] asdf',
+			'Message-ID' => '<patch@example.com>',
+			],
+		parts => $parts,
+	);
+
+	my $html = PublicInbox::View->as_html($mime);
+	like($html, qr!see attached patch\n-+ foo\.patch -+\n--- a/file\n!,
+		"parts split with filename");
 }
 
 done_testing();
-- 
cgit v1.2.3-24-ge0c7