Fix incorrect copyright detection #379 and #84

 * also add test for #358

Signed-off-by: Philippe Ombredanne <pombredanne@nexb.com>

src/cluecode/copyrights.py

@@ -198,6 +198,10 @@ def detect(location):
     (r'^AppPublisher$', 'JUNK'),
     (r'^DISCLAIMS?|SPECIFICALLY|WARRANT(Y|I)E?S?$', 'JUNK'),
     (r'^(hispagestyle|Generic|Change|Add|Generic|Average|Taken|LAWS\.?|design|Driver)$', 'JUNK'),
+    # Windows XP
+    (r'^(Windows|XP|SP1|SP2|SP3|SP4)$', 'JUNK'),
+
+    (r'^example\.com$', 'JUNK'),
 
     # various trailing words that are junk
     (r'^(?:Copyleft|LegalCopyright|AssemblyCopyright|Distributed|Report|'

tests/cluecode/test_copyrights.py

@@ -4100,7 +4100,25 @@ def test_copyright_copr5_correct(self):
         check_detection(expected, test_lines)
 
     def test_copyright_oracle(self):
-        test_lines =['Copyright (c) 1997-2015 Oracle and/or its affiliates. All rights reserved.']
+        test_lines = ['Copyright (c) 1997-2015 Oracle and/or its affiliates. All rights reserved.']
 
         expected = ['Copyright (c) 1997-2015 Oracle and/or its affiliates.']
         check_detection(expected, test_lines)
+
+    def test_copyright_windows(self):
+        test_lines = ['This release supports NT-based Windows releases like Windows 2000 SP4, Windows XP, and Windows 2003.']
+
+        expected = []
+        check_detection(expected, test_lines)
+
+    def test_copyright_in_binary_sql_server(self):
+        test_lines = ['2005charchar? 7 DDLSQL Server 2005smalldatetimedatetimeLDDDDDD7']
+
+        expected = []
+        check_detection(expected, test_lines)
+
+    def test_copyright_with_example_com_url(self):
+        test_lines = ['"domain": function(c) { assert.equal(c.domain, "example.com") },']
+
+        expected = []
+        check_detection(expected, test_lines)