adding code for analyzing overlap

1 files changed, 27 insertions, 6 deletions

diff --git a/monitor/josef_logreader.py b/monitor/josef_logreader.py
index 5f6f768..2b3a473 100755
--- a/monitor/josef_logreader.py
+++ b/monitor/josef_logreader.py
@@ -5,6 +5,7 @@ import sys
 import time
 import datetime
 import os
+import json
 
 from monitor_conf import *
 
@@ -96,18 +97,38 @@ def print_errors(l):
             break # comment this line to print all errors ever
             pass
 
+def print_overlap(l):
+    log = read_loglog(l)
+    entry_count = {}
+    total = 0
+    for line in log:
+        # line = item[TIME_LEN:]
+        logs = json.loads(line.split("logs: ")[-1][:-1].replace("'", '"'))
+        for l in logs:
+            if l in entry_count:
+                entry_count[l] += 1
+            else:
+                entry_count[l] = 1
+        total += 1
+    
+    # print entry_count
+    for e in entry_count:
+        print e + ", " + str(entry_count[e]) + " (" + str(int(100 * float(entry_count[e])/float(total))) + "%)"
 
 if __name__ == "__main__":
+    OVERLAP_STR = "_overlap.log"
     logs = get_logs()
     for log in logs:
         if log == "monitor.log":
             pass
-        else:
-            print log
-            print_log_stats(log)
-            print_average_age(log)
-            print_errors(log)
-            print ""
+        elif log[- len(OVERLAP_STR):] == OVERLAP_STR:
+            print_overlap(log)
+        # else:
+        #     print log
+        #     print_log_stats(log)
+        #     print_average_age(log)
+        #     print_errors(log)
+        #     print ""