--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/bin/nogood.py	Mon Apr 13 14:12:12 2020 +0100
@@ -0,0 +1,20 @@
+#!/lustre/sw/miniconda3/bin/python3
+'''Detect bad diffs between warc.sh log file and existing extract_...tar
+Usage: nogood.py segid numChanged
+Note the slightly counter-intuitive value: 0 if the input is no good'''
+import sys,re
+LONG_ADD=re.compile('[1-9][0-9]*a[1-9][0-9]*,[1-9][0-9]*')
+(segment,n)=map(int,sys.argv[1:])
+c=0
+for l in sys.stdin:
+    (i,d)=l.rstrip().split(':')
+    if LONG_ADD.fullmatch(d):
+        c+=1
+        continue
+    else:
+        print("non-simple diff at %s: %s for segment %s"%(i,d,segment),file=sys.stderr)
+        exit(0)
+if c!=n:
+    print("too many diffs? (%s<>%s) for segment %s"%(c,n,segment),file=sys.stderr)
+    exit(0)
+exit(1)