From 14fbb1499be15165bb0dd1565941974809f207eb Mon Sep 17 00:00:00 2001
From: Rodrigo Arias Mallo <rodrigo.arias@bsc.es>
Date: Fri, 5 Mar 2021 18:29:43 +0100
Subject: [PATCH] ds: add perf stat parser

We can only read one output file by now, located at:
.garlic/perf.csv
---
 garlic/ds/index.nix     |  2 ++
 garlic/ds/perf/stat.nix | 23 +++++++++++++
 garlic/ds/perf/stat.py  | 71 +++++++++++++++++++++++++++++++++++++++++
 3 files changed, 96 insertions(+)
 create mode 100644 garlic/ds/perf/stat.nix
 create mode 100644 garlic/ds/perf/stat.py

diff --git a/garlic/ds/index.nix b/garlic/ds/index.nix
index c3c0454..1c2c28e 100644
--- a/garlic/ds/index.nix
+++ b/garlic/ds/index.nix
@@ -15,4 +15,6 @@
     latency = callPackage ./osu/latency.nix {};
     bw = callPackage ./osu/bw.nix {};
   };
+
+  perf.stat = callPackage ./perf/stat.nix {};
 }
diff --git a/garlic/ds/perf/stat.nix b/garlic/ds/perf/stat.nix
new file mode 100644
index 0000000..14b55c4
--- /dev/null
+++ b/garlic/ds/perf/stat.nix
@@ -0,0 +1,23 @@
+{
+  stdenv
+, python3
+, gzip
+}:
+
+resultTree:
+
+stdenv.mkDerivation {
+  name = "perf-stat.json.gz";
+  preferLocalBuild = true;
+  src = ./stat.py;
+  phases = [ "unpackPhase" "installPhase" ];
+
+  unpackPhase = ''
+    cp $src stat.py
+  '';
+
+  buildInputs = [ python3 gzip ];
+  installPhase = ''
+    python stat.py ${resultTree} | gzip > $out
+  '';
+}
diff --git a/garlic/ds/perf/stat.py b/garlic/ds/perf/stat.py
new file mode 100644
index 0000000..ac180fe
--- /dev/null
+++ b/garlic/ds/perf/stat.py
@@ -0,0 +1,71 @@
+import json, re, sys, os, glob
+from os import path
+
+def eprint(*args, **kwargs):
+	print(*args, file=sys.stderr, flush=True, **kwargs)
+
+def process_run(tree, runPath):
+	with open(".garlic/perf.csv", "r") as f:
+		lines = [line.strip() for line in f.readlines()]
+
+	perf_data = {}
+
+	for line in lines:
+		if len(line) == 0: continue
+		if line[0] == '#': continue
+
+		slices = line.split(',')
+		if len(slices) != 7:
+			print("error: mismatched columns")
+			exit(1)
+
+		name = slices[2].replace("-", "_")
+		value = float(slices[0])
+
+		perf_data[name] = value
+
+	tree['perf'] = perf_data
+
+	print(json.dumps(tree))
+
+def process_result_tree(resultTree):
+
+	eprint("processing resultTree: " + resultTree)
+
+	os.chdir(resultTree)
+
+	experiments = glob.glob(resultTree + "/*-experiment")
+
+	for exp in glob.glob("*-experiment"):
+		eprint("found experiment: " + exp)
+		expPath = path.join(resultTree, exp)
+		os.chdir(expPath)
+
+		for unit in glob.glob("*-unit"):
+			eprint("found unit: " + unit)
+			unitPath = path.join(resultTree, exp, unit)
+			os.chdir(unitPath)
+
+			with open('garlic_config.json') as json_file:
+				garlic_conf = json.load(json_file)
+
+			tree = {"exp":exp, "unit":unit, "config":garlic_conf}
+
+			for i in range(garlic_conf['loops']):
+				run = str(i + 1)
+				runPath = path.join(resultTree, exp, unit, run)
+				if path.isdir(runPath) == False:
+					eprint("missing run {}, aborting".format(run))
+					exit(1)
+
+				tree["run"] = run
+				os.chdir(runPath)
+
+				process_run(tree, runPath)
+
+
+if len(sys.argv) != 2:
+	eprint("usage: python {} <resultTree>".format(argv[0]))
+	exit(1)
+
+process_result_tree(sys.argv[1])