From ded0eeabeb758db282c31a29eb014ddcbd00240e Mon Sep 17 00:00:00 2001
From: zackcglee <leechanggong@gmail.com>
Date: Tue, 31 Dec 2024 11:00:46 +0900
Subject: [PATCH] [bin] add stat script

---
 bin/print_bench_stat.py | 48 +++++++++++++++++++++++++++++++++++++++++
 1 file changed, 48 insertions(+)
 create mode 100755 bin/print_bench_stat.py

diff --git a/bin/print_bench_stat.py b/bin/print_bench_stat.py
new file mode 100755
index 0000000..13bcb6a
--- /dev/null
+++ b/bin/print_bench_stat.py
@@ -0,0 +1,48 @@
+#!/usr/bin/env python3
+
+import os, json, csv
+
+FILE_PATH=os.path.dirname(os.path.abspath(__file__))
+ROOT=os.path.abspath(os.path.join(FILE_PATH, ".."))
+BENCHMARK_DIR=os.path.join(ROOT, "benchmark")
+
+def get_benchmark_data():
+  out = os.path.join(ROOT, "stat.tsv")
+  projects = os.listdir(BENCHMARK_DIR)
+  stat = dict()
+  for project in projects:
+      stat[project] = dict()
+      versions = os.listdir(os.path.join(BENCHMARK_DIR, project))
+      for version in versions:
+          stat[project][version] = dict()
+          label_file = os.path.join(BENCHMARK_DIR, project, version, "label.json")
+          if not os.path.exists(label_file):
+              stat[project][version]['bug_trace'] = '0'
+              stat[project][version]['bugs'] = '0'
+              continue
+          with open(label_file, 'r') as f:
+              data = json.load(f)
+              stat[project][version]['bugs'] = len(data)
+              stat[project][version]['bug_trace'] = 0
+              for bug in data:
+                  if 'bug-trace' in bug.keys() and bug['bug-trace'] != []:
+                      stat[project][version]['bug_trace'] += 1
+  return stat
+
+def print_stat(stat):
+    stat = get_benchmark_data()
+    with open(os.path.join(ROOT, "stat.tsv"), 'w') as f:
+        writer = csv.writer(f, delimiter='\t')
+        writer.writerow(['Project', 'Version', '# Bugs', '# Bug Trace'])
+        f.flush()
+        for project in stat:
+            for version in stat[project]:
+                writer.writerow([project, version, stat[project][version]['bugs'], stat[project][version]['bug_trace']])
+                f.flush()
+    f.close()     
+def main():
+    stat = get_benchmark_data()
+    print_stat(stat)            
+
+if __name__ == "__main__":
+    main()
\ No newline at end of file