t/perf: add infrastructure for measuring sizes

The main objective of scripts in the perf framework is to
run "test_perf", which measures the time it takes to run
some operation. However, it can also be interesting to see
the change in the output size of certain operations.

This patch introduces test_size, which records a single
numeric output from the test and shows it in the aggregated
output (with pretty printing and relative size comparison).

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

t/perf/README

@@ -168,3 +168,28 @@ that
   While we have tried to make sure that it can cope with embedded
   whitespace and other special characters, it will not work with
   multi-line data.
+
+Rather than tracking the performance by run-time as `test_perf` does, you
+may also track output size by using `test_size`. The stdout of the
+function should be a single numeric value, which will be captured and
+shown in the aggregated output. For example:
+
+	test_perf 'time foo' '
+		./foo >foo.out
+	'
+
+	test_size 'output size'
+		wc -c <foo.out
+	'
+
+might produce output like:
+
+	Test                origin           HEAD
+	-------------------------------------------------------------
+	1234.1 time foo     0.37(0.79+0.02)  0.26(0.51+0.02) -29.7%
+	1234.2 output size             4.3M             3.6M -14.7%
+
+The item being measured (and its units) is up to the test; the context
+and the test title should make it clear to the user whether bigger or
+smaller numbers are better. Unlike test_perf, the test code will only be
+run once, since output sizes tend to be more deterministic than timings.

t/perf/aggregate.perl

@@ -13,10 +13,16 @@ sub get_times {
 	my $line = <$fh>;
 	return undef if not defined $line;
 	close $fh or die "cannot close $name: $!";
-	$line =~ /^(?:(\d+):)?(\d+):(\d+(?:\.\d+)?) (\d+(?:\.\d+)?) (\d+(?:\.\d+)?)$/
-		or die "bad input line: $line";
-	my $rt = ((defined $1 ? $1 : 0.0)*60+$2)*60+$3;
-	return ($rt, $4, $5);
+	# times
+	if ($line =~ /^(?:(\d+):)?(\d+):(\d+(?:\.\d+)?) (\d+(?:\.\d+)?) (\d+(?:\.\d+)?)$/) {
+		my $rt = ((defined $1 ? $1 : 0.0)*60+$2)*60+$3;
+		return ($rt, $4, $5);
+	# size
+	} elsif ($line =~ /^\d+$/) {
+		return $&;
+	} else {
+		die "bad input line: $line";
+	}
 }
 
 sub relative_change {
@@ -32,9 +38,15 @@ sub relative_change {
 
 sub format_times {
 	my ($r, $u, $s, $firstr) = @_;
+	# no value means we did not finish the test
 	if (!defined $r) {
 		return "<missing>";
 	}
+	# a single value means we have a size, not times
+	if (!defined $u) {
+		return format_size($r, $firstr);
+	}
+	# otherwise, we have real/user/system times
 	my $out = sprintf "%.2f(%.2f+%.2f)", $r, $u, $s;
 	$out .= ' ' . relative_change($r, $firstr) if defined $firstr;
 	return $out;
@@ -54,6 +66,25 @@ sub usage {
 	exit(1);
 }
 
+sub human_size {
+	my $n = shift;
+	my @units = ('', qw(K M G));
+	while ($n > 900 && @units > 1) {
+		$n /= 1000;
+		shift @units;
+	}
+	return $n unless length $units[0];
+	return sprintf '%.1f%s', $n, $units[0];
+}
+
+sub format_size {
+	my ($size, $first) = @_;
+	# match the width of a time: 0.00(0.00+0.00)
+	my $out = sprintf '%15s', human_size($size);
+	$out .= ' ' . relative_change($size, $first) if defined $first;
+	return $out;
+}
+
 my (@dirs, %dirnames, %dirabbrevs, %prefixes, @tests,
     $codespeed, $sortby, $subsection, $reponame);
 
@@ -184,7 +215,14 @@ sub print_default_results {
 		my $firstr;
 		for my $i (0..$#dirs) {
 			my $d = $dirs[$i];
-			$times{$prefixes{$d}.$t} = [get_times("$resultsdir/$prefixes{$d}$t.times")];
+			my $base = "$resultsdir/$prefixes{$d}$t";
+			$times{$prefixes{$d}.$t} = [];
+			foreach my $type (qw(times size)) {
+				if (-e "$base.$type") {
+					$times{$prefixes{$d}.$t} = [get_times("$base.$type")];
+					last;
+				}
+			}
 			my ($r,$u,$s) = @{$times{$prefixes{$d}.$t}};
 			my $w = length format_times($r,$u,$s,$firstr);
 			$colwidth[$i] = $w if $w > $colwidth[$i];

t/perf/perf-lib.sh

@@ -231,6 +231,19 @@ test_perf () {
 	test_wrapper_ test_perf_ "$@"
 }
 
+test_size_ () {
+	say >&3 "running: $2"
+	if test_eval_ "$2" 3>"$base".size; then
+		test_ok_ "$1"
+	else
+		test_failure_ "$@"
+	fi
+}
+
+test_size () {
+	test_wrapper_ test_size_ "$@"
+}
+
 # We extend test_done to print timings at the end (./run disables this
 # and does it after running everything)
 test_at_end_hook_ () {