issue SpamExperts#147 - Recognize GTUBE string as test spam

pyzor/digest.py

@@ -77,6 +77,9 @@ class DataDigester(object):
     # Note that an empty string will always be used to remove whitespace.
     unwanted_txt_repl = ''
 
+    # The GTUBE standard spam test string https://spamassassin.apache.org/gtube/
+    gtube_pattern = 'XJS*C4JDBQADN1.NSBN3*2IDNEN*GTUBE-STANDARD-ANTI-UBE-TEST-EMAIL*C.34X'
+
     def __init__(self, msg, spec=None):
         if spec is None:
             spec = digest_spec
@@ -87,6 +90,10 @@ def __init__(self, msg, spec=None):
         lines = []
         for payload in self.digest_payloads(msg):
             for line in payload.splitlines():
+                # Any email containing the GTUBE pattern will be forced to the same digest hash
+                if self.gtube_pattern in line:
+                    lines = [self.gtube_pattern.encode("utf8", "ignore")]
+                    break
                 norm = self.normalize(line)
                 if self.should_handle_line(norm):
                     try:

pyzor/server.py

@@ -321,6 +321,12 @@ def handle_check(self, digests):
         This command returns the spam/ham counts for the specified digest.
         """
         digest = digests[0]
+        # Special case the digest for the GTUBE spam test message
+        gtube_digest = 'c13086867f444d503829044f504826177e3eb438'
+        if digest == gtube_digest:
+            self.server.log.debug("Request to check digest for GTUBE %s", digest)
+            self.response["Count"] = "%d" % sys.maxsize
+            self.response["WL-Count"] = "%d" % 0
         try:
             record = self.server.database[digest]
         except KeyError: