Improve PR Jira issue matcher

sync2jira/intermediary.py

@@ -17,6 +17,7 @@
 #
 # Authors:  Ralph Bean <[email protected]>
 import re
+from typing import Optional
 
 
 class Issue(object):
@@ -277,33 +278,37 @@ def map_fixVersion(mapping, issue):
         issue["milestone"] = fixVersion_map.replace("XXX", issue["milestone"])
 
 
-def matcher(content, comments):
+JIRA_REFERENCE = re.compile(r"Relates +to +JIRA:\s*(\w+-\d+)")
+
+
+def matcher(content: Optional[str], comments: list[dict[str, str]]) -> str:
     """
     Helper function to match to a JIRA
 
+    Extract the Jira ticket reference from the first instance of the magic
+    cookie (e.g., "Relates to JIRA: FACTORY-1234") found when searching
+    through the comments in reverse order.  If no reference is found in the
+    comments, then look in the PR description.  This ordering allows later
+    comments to override earlier ones as well as any reference in the
+    description.
+
     :param String content: PR description
     :param List comments: Comments
     :return: JIRA match or None
     :rtype: Bool
     """
-    # Build out a string with all comments and initial_comment
-    all_data = " "
+
+    def find_it(input_str: str) -> str:
+        match = JIRA_REFERENCE.search(input_str)
+        return match.group(1) if match else None
+
     for comment in reversed(comments):
-        all_data += f" {comment['body']}"
-    if content:
-        all_data += content
-    if all_data:
-        # Parse to extract the JIRA information. 2 types of matches:
-        # 1 - To match to JIRA issue (i.e. Relates to JIRA: FACTORY-1234)
-        # 2 - To match to upstream issue (i.e. Relates to Issue: !5)
-        match_jira = re.findall(r"Relates to JIRA: ([\w]*-[\d]*)", all_data)
-        if match_jira:
-            for match in match_jira:
-                # Assert that the match was correct
-                if re.match(r"[\w]*-[\d]*", match):
-                    return match
-        else:
-            return None
+        match_str = find_it(comment["body"])
+        if match_str:
+            break
+    else:
+        match_str = find_it(content)
+    return match_str
 
 
 def trim_string(content):

tests/test_intermediary.py

@@ -229,18 +229,84 @@ def test_from_github_pr_reopen(self, mock_matcher):
 
     def test_matcher(self):
         """This tests the matcher function"""
-        # Positive case
-        content = "Relates to JIRA: XYZ-5678"
-        comments = [{"body": "Relates to JIRA: ABC-1234"}]
-        expected = True
-        actual = bool(i.matcher(content, comments))
-        assert expected == actual
+        # Found in content, no comments
+        expected = "XYZ-5678"
+        content = f"Relates to JIRA: {expected}"
+        comments = []
+        actual = i.matcher(content, comments)
+        self.assertEqual(expected, actual)
+
+        # Found in comment, no content
+        expected = "XYZ-5678"
+        content = None
+        comments = [{"body": f"Relates to JIRA: {expected}"}]
+        actual = i.matcher(content, comments)
+        self.assertEqual(expected, actual)
+
+        # Found in content, not spanning comments
+        expected = "XYZ-5678"
+        content = f"Relates to JIRA: {expected}"
+        comments = [
+            {"body": "ABC-1234"},
+            {"body": "JIRA:"},
+            {"body": "to"},
+            {"body": "Relates"},
+        ]
+        actual = i.matcher(content, comments)
+        self.assertEqual(expected, actual)
+
+        # Found in comment, not contents
+        expected = "XYZ-5678"
+        content = "Nothing here"
+        comments = [
+            {"body": "Relates"},
+            {"body": f"Relates to JIRA: {expected}"},
+            {"body": "stuff"},
+        ]
+        actual = i.matcher(content, comments)
+        self.assertEqual(expected, actual)
+
+        # Overridden in comment
+        expected = "XYZ-5678"
+        content = "Relates to JIRA: ABC-1234"
+        comments = [
+            {"body": "Relates"},
+            {"body": f"Relates to JIRA: {expected}"},
+            {"body": "stuff"},
+        ]
+        actual = i.matcher(content, comments)
+        self.assertEqual(expected, actual)
+
+        # Overridden twice in comments
+        expected = "XYZ-5678"
+        content = "Relates to JIRA: ABC-1234"
+        comments = [
+            {"body": "Relates to JIRA: ABC-1235"},
+            {"body": f"Relates to JIRA: {expected}"},
+            {"body": "stuff"},
+        ]
+        actual = i.matcher(content, comments)
+        self.assertEqual(expected, actual)
+
+        # Funky spacing
+        expected = "XYZ-5678"
+        content = f"Relates  to  JIRA:   {expected}"
+        comments = []
+        actual = i.matcher(content, comments)
+        self.assertEqual(expected, actual)
+
+        # Funkier spacing
+        expected = "XYZ-5678"
+        content = f"Relates to JIRA:{expected}"
+        comments = []
+        actual = i.matcher(content, comments)
+        self.assertEqual(expected, actual)
 
         # Negative case
         content = "No JIRAs here..."
         comments = [{"body": "... nor here"}]
-        expected = False
-        actual = bool(i.matcher(content, comments))
-        assert expected == actual
+        expected = None
+        actual = i.matcher(content, comments)
+        self.assertEqual(expected, actual)
 
     # TODO: Add new tests from PR