Benchmark encoding translated messages as binary, too.

Results for all of my trace files:

> ./compare_sizes.py ~/ownCloud/OpenXC\ Traces/peplin/**/*.json       1
For the 89 trace files given...
Total transferred raw JSON size is 13.3MB
Total transferred raw binary size is 2.1MB
Total transferred translated JSON size is 791.0MB
Total transferred translated binary size is 318.6MB
Total transferred JSON size is 804.3MB
Total transferred binary size is 320.6MB
Binary encoding is 84.328494% smaller than JSON for raw messages
Binary encoding is 59.727489% smaller than JSON for translated messages
Binary encoding is 60.133196% smaller than JSON overall

diff --git a/benchmark/proto/compare_sizes.py b/benchmark/proto/compare_sizes.py
old mode 100644 (file)
new mode 100755 (executable)
index ce6b25b..b8b79bb
--- a/benchmark/proto/compare_sizes.py
+++ b/benchmark/proto/compare_sizes.py
@@ -2,6 +2,7 @@
 
 from __future__ import division
 import sys
+import numbers
 
 import openxc_pb2
 import json
@@ -12,20 +13,53 @@ def sizeof_fmt(num):
             return "%3.1f%s" % (num, unit)
         num /= 1024.0
 
-total_json_size = 0
-total_binary_size = 0
+total_raw_json_size = 0
+total_raw_binary_size = 0
+total_translated_json_size = 0
+total_translated_binary_size = 0
 
-trace_file = sys.argv[1]
-for line in open(trace_file):
-    raw_message = json.loads(line)
-    total_json_size += len(json.dumps(raw_message))
-    binary_message = openxc_pb2.RawMessage()
-    binary_message.message_id = raw_message['id']
-    binary_message.data = int(raw_message['data'], 0)
-    total_binary_size += len(binary_message.SerializeToString())
+for trace_file in sys.argv[1:]:
+    for line in open(trace_file):
+        try:
+            json_message = json.loads(line)
+        except ValueError:
+            continue
 
-print("For the trace file %s..." % trace_file)
+        if 'id' and 'data' in json_message:
+            total_raw_json_size += len(line)
+            binary_message = openxc_pb2.RawMessage()
+            binary_message.message_id = json_message['id']
+            binary_message.data = int(json_message['data'], 0)
+            total_raw_binary_size += len(binary_message.SerializeToString())
+        else:
+            if isinstance(json_message['value'], bool):
+                binary_message = openxc_pb2.TranslatedBooleanMessage()
+            elif isinstance(json_message['value'], numbers.Number):
+                binary_message = openxc_pb2.TranslatedNumericMessage()
+            else:
+                binary_message = openxc_pb2.TranslatedStringMessage()
+            binary_message.name = json_message['name']
+            binary_message.value = json_message['value']
+            total_translated_json_size += len(line)
+            total_translated_binary_size += len(binary_message.SerializeToString())
+
+
+print("For the %d trace files given..." % len(sys.argv[1:]))
+print("Total transferred raw JSON size is %s" % sizeof_fmt(total_raw_json_size))
+print("Total transferred raw binary size is %s" % sizeof_fmt(total_raw_binary_size))
+print("Total transferred translated JSON size is %s" %
+        sizeof_fmt(total_translated_json_size))
+print("Total transferred translated binary size is %s" %
+        sizeof_fmt(total_translated_binary_size))
+
+total_json_size = total_raw_json_size + total_translated_json_size
 print("Total transferred JSON size is %s" % sizeof_fmt(total_json_size))
+total_binary_size = total_raw_binary_size + total_translated_binary_size
 print("Total transferred binary size is %s" % sizeof_fmt(total_binary_size))
+
+print("Binary encoding is %f%% smaller than JSON for raw messages" % (
+        100 - (total_raw_binary_size / total_raw_json_size * 100)))
+print("Binary encoding is %f%% smaller than JSON for translated messages" % (
+        100 - (total_translated_binary_size / total_translated_json_size * 100)))
 print("Binary encoding is %f%% smaller than JSON overall" % (
         100 - (total_binary_size / total_json_size * 100)))

diff --git a/benchmark/proto/openxc.proto b/benchmark/proto/openxc.proto
index 0af663c..0e39bb6 100644 (file)
--- a/benchmark/proto/openxc.proto
+++ b/benchmark/proto/openxc.proto
@@ -19,3 +19,6 @@ message TranslatedBooleanMessage {
     optional string name = 1;
     optional bool value = 2;
 }
+
+// TODO we should also consider having an enum type, h aving each specific
+// message defined as a protobuf