Do not consider issue text in __eq__ method

Issue text is not supposed to be consistent between banndit versions. It
was changed few times in the past. That's why we don't need to match text too
between issues.

This issue is only valid if we compare old bandit baseline generated
with the old version with a new one generated using the last bandit.

bandit/core/issue.py

@@ -36,10 +36,9 @@ def __str__(self):
                                   self.confidence, self.fname, self.lineno)
 
     def __eq__(self, other):
-        # if the issue text, severity, confidence, and filename match, it's
+        # if the issue severity, confidence, and filename match, it's
         # the same issue from our perspective
-        match_types = ['text', 'severity', 'confidence', 'fname', 'test',
-                       'test_id']
+        match_types = ['severity', 'confidence', 'fname', 'test', 'test_id']
         return all(getattr(self, field) == getattr(other, field)
                    for field in match_types)
 

tests/unit/core/test_issue.py

@@ -93,9 +93,6 @@ def test_matches_issue(self):
         # confidence doesn't match
         self.assertNotEqual(issue_a, issue_c)
 
-        # text doesn't match
-        self.assertNotEqual(issue_a, issue_d)
-
         # filename doesn't match
         self.assertNotEqual(issue_a, issue_e)