Improve author detection for compact declarations and dotted author names (#4229)

AUTHORS.rst

@@ -27,6 +27,7 @@ The following organizations or individuals have contributed to ScanCode:
 - Daniel Eder @daniel-eder
 - Dan Kegel @dankegel
 - Dennis Clark @DennisClark
+- Dhruv @DHRUV-1-star
 - Divyansh Sharma @Divyansh2512
 - Duncan Howe @Duncan-Howe
 - Felix Kauselmann @selmf

CHANGELOG.rst

@@ -10,6 +10,9 @@ Next release
 
 - Improve copyright detection for statements with parens or trailing "authors"
 
+- Improve author detection for compact declarations and dotted author names.
+  https://github.com/aboutcode-org/scancode-toolkit/issues/4229
+
 - Add experimental option for using cached results during scan time. When the
   ``--use-cached-results`` option is enabled in the ScanCode CLI, during scan
   time for a given Resource, we iterate through the active scanners and see if we

src/cluecode/copyrights.py

@@ -431,6 +431,21 @@ def get_tokens(numbered_lines, splitter=re.compile(r'[\t =;]+').split):
         # space before in some cases: exclude (digit , (s , (c , (-
         # this allows to recover from words like KISA(Korean
         line = re.sub(pattern=r'(\([^rsc\-\d])', repl=r' \g<1>', string=line)
+
+        # Split compact author prefix such as Author:Frankie.Chu while
+        # preserving colons elsewhere (e.g. in URLs).
+        line = re.sub(
+            pattern=r'(^|\s)([Aa]uthors?):(?=\S)',
+            repl=r'\1\2: ',
+            string=line,
+        )
+
+        # Normalize dotted names only when they follow an Author tag.
+        line = re.sub(
+            pattern=r'(^|\s)([Aa]uthors?):\s+([A-Z][a-z]+)\.([A-Z][a-z]+)(?=\b|[,;])',
+            repl=r'\1\2 \3 \4',
+            string=line,
+        )
         for tok in splitter(line):
             # strip trailing quotes+comma
             if tok.endswith("',"):

tests/cluecode/test_copyrights_basic.py

@@ -319,6 +319,38 @@ def test_detect_with_lines_only_holders(self):
         ))
         assert results == expected
 
+    def test_detect_author_without_space_and_with_dotted_name(self):
+        numbered_lines = [
+            (1, '//  Date:9 April,2012'),
+            (2, '//  Author:Frankie.Chu'),
+        ]
+        expected = [
+            copyrights.AuthorDetection('Frankie Chu', 2, 2),
+        ]
+        results = list(copyrights.detect_copyrights_from_lines(
+            numbered_lines,
+            include_copyrights=False,
+            include_holders=False,
+            include_authors=True,
+        ))
+        assert results == expected
+
+    def test_detect_author_with_space_and_with_dotted_name(self):
+        numbered_lines = [
+            (1, '//  Date:9 April,2012'),
+            (2, '//  Author: Frankie.Chu'),
+        ]
+        expected = [
+            copyrights.AuthorDetection('Frankie Chu', 2, 2),
+        ]
+        results = list(copyrights.detect_copyrights_from_lines(
+            numbered_lines,
+            include_copyrights=False,
+            include_holders=False,
+            include_authors=True,
+        ))
+        assert results == expected
+
 
 def check_full_detections(expected, test_file):
     """