Merge pull request #2666 from nexB/improve-copyright-detection

Do not mistake path for copyright year

Author: pombredanne

src/cluecode/copyrights.py

@@ -516,7 +516,7 @@ def from_node(
     # COPYRIGHT
     ############################################################################
 
-    # first some exceptions
+    # some exceptions
 
     # NOT a copyright Copyright.txt : treat as NN
     (r'^Copyright\.txt$', 'NN'),
@@ -581,6 +581,10 @@ def from_node(
     # JUNK are things to ignore
     ############################################################################
 
+    # path with trailing year-like are NOT a year as in
+    # Landroid/icu/impl/IDNA2003 : treat as JUNK
+    (r'^[^\\/]+[\\/][^\\/]+[\\/].*$', 'JUNK'),
+
     # Combo of many (3+) letters and punctuations groups without spaces is likely junk
     # "AEO>>,o>>'!xeoI?o?O1/4thuA/"
     # (r'((\w+\W+){3,})+', 'JUNK'),

tests/cluecode/data/copyrights/year-like-path.txt

@@ -0,0 +1,4 @@
+Landroid/icu/impl/IDNA2003;->startsWithPrefix(Ljava/lang/StringBuffer;)Z,BLACK
+Landroid/icu/impl/IDNA2003;->toASCIILower(C)C,BLACK
+Landroid/icu/impl/IDNA2003;->toASCIILower(Ljava/lang/CharSequence;)Ljava/lang/StringBuffer;,BLACK
+

tests/cluecode/data/copyrights/year-like-path.txt.yml

@@ -0,0 +1,2 @@
+what:
+  - copyrights