changeset 327:cb3b9cd1a7ad trim

refined a bit
author Henry S. Thompson <ht@inf.ed.ac.uk>
date Sat, 24 Jan 2026 12:01:43 +0000
parents f6643533e971
children 429125596c68
files lib/python/cc/lmh/miss_bogons.py
diffstat 1 files changed, 27 insertions(+), 11 deletions(-) [+]
line wrap: on
line diff
--- a/lib/python/cc/lmh/miss_bogons.py	Fri Jan 23 15:54:54 2026 +0000
+++ b/lib/python/cc/lmh/miss_bogons.py	Sat Jan 24 12:01:43 2026 +0000
@@ -6,18 +6,34 @@
   b'4312://5.160.123.10/' b'10,123,160,5)/'
 '''
 import sys, re
-kPat = re.compile("b'([0-9]*s?://)([^/]*)('$|/(.*)'$)")
-kuPat = re.compile("^([^/]*)/?($|([?].*)$)")
+cdbPat = re.compile('''b['"]([0-9]*s?://)([^/]*)(.*)['"]$''') # '
+cdbUriPat = re.compile("(/[^/]*)/?([?].*)?$")
+idxPat = re.compile('''b['"]([^)]*)[)](.*)['"]$''') # '
 
+i=1
 with open(sys.argv[1], 'r') as f:
   for l in f:
-    (key, url) = l.rstrip().split(' ')
-    if (m:=kPat.match(key)):
-      (k_d,k_h,_,k_u) = m.groups()
+    try:
+      (cdb, idx) = l.rstrip().split(' ')
+    except ValueError:
+      if l[0]!='b':
+        exit(0)
+      else:
+        print("can't split:%s:%s"%(i,l), file=sys.stderr)
+        exit(1)
+    if (m:=cdbPat.match(cdb)):
+      (cdb_date,cdb_host,cdb_uri) = m.groups()
     else:
-      print("bad key |%s|"%key,file=sys.stderr)
-      exit(1)
-    #print(key,"|%s|%s|%s|"%(k_d,k_h,k_u))
-    print(kuPat.sub('\\1\\2', k_u))
-
-    
+      print("bad cdb |%s|"%cdb,file=sys.stderr)
+      exit(2)
+    #print(cdb,"|%s|%s|%s|"%(cdb_date,cdb_host,cdb_uri),file=sys.stderr)
+    clean_cdb_uri = '/' if cdb_uri == '' else \
+                    cdbUriPat.sub('\\1\\2', cdb_uri).lower()
+    if (m:=idxPat.match(idx)):
+      idx_uri = m.group(2)
+    else:
+      print("bad idx |%s|"%idx, file=sys.stderr)
+      exit(3)
+    if clean_cdb_uri != idx_uri:
+      print("%s:\n %s\n %s"%(i, clean_cdb_uri, idx_uri))
+    i+=1