Merge pull request #2707 from nexB/copyright-fixes

Improve Copyright detection

Author: pombredanne

spdx-copy.txt

@@ -0,0 +1,7 @@
+#
+# SPDX-FileCopyrightText: Copyright 2019-2021 Test Inc.
+# SPDX-FileCopyrightText: Copyright (c) SOIM
+#
+# SPDX-License-Identifier: GPL-3.0
+
+import os

spdx-copy.txt.yml

@@ -0,0 +1,5 @@
+what:
+  - copyrights
+copyrights:
+  - Copyright 2019-2021 Test Inc.
+  - Copyright (c) SOIM

src/cluecode/copyrights.py

@@ -617,7 +617,7 @@ def from_node(
     (r'^Agreement$', 'JUNK'),
     (r'^Usage$', 'JUNK'),
     (r'^Please$', 'JUNK'),
-    (r'^Based$', 'JUNK'),
+    (r'^\(?Based$', 'JUNK'),
     (r'^Upstream$', 'JUNK'),
     (r'^Files?$', 'JUNK'),
     (r'^Filename:?$', 'JUNK'),
@@ -1140,6 +1140,7 @@ def from_node(
     (r'^The$', 'NN'),
     (r'^THE', 'NN'),
     (r'^These$', 'NN'),
+    (r'^[tT]here$', 'NN'),
     (r'^This$', 'NN'),
     (r'^THIS$', 'NN'),
     (r'^Those$', 'NN'),
@@ -2217,7 +2218,8 @@ def from_node(
     COPYRIGHT: {<COPY>+  <YR-RANGE>  <NN>  <NNP>} #22792
 
     # Copyright (c) 2017 odahcam
-    COPYRIGHT: {<COPY>+  <YR-RANGE>  <NN>+ <ALLRIGHTRESERVED>?} #22793
+    # or Copyright (c) 2019-2021, Open source contributors.
+    COPYRIGHT: {<COPY>+  <YR-RANGE>  <NN>+ <CONTRIBUTORS>? <ALLRIGHTRESERVED>?} #22793
 
     # Licensed material of Foobar Company, All Rights Reserved, (C) 2005
     COPYRIGHT: {<COMPANY>  <ALLRIGHTRESERVED>  <COPYRIGHT>} #22794

tests/cluecode/data/copyrights/complex-punctuation.txt

@@ -0,0 +1 @@
+// Copyright 2018-2019 @paritytech/substrate-light-ui authors & contributors

tests/cluecode/data/copyrights/complex-punctuation.txt.yml

@@ -0,0 +1,6 @@
+what:
+  - copyrights
+copyrights:
+  - Copyright 2018-2019 paritytech/substrate-light-ui authors & contributors
+
+

tests/cluecode/data/copyrights/contributors.txt

@@ -0,0 +1 @@
+Copyright (c) 2015, Contributors

tests/cluecode/data/copyrights/contributors.txt.yml

@@ -0,0 +1,6 @@
+what:
+  - copyrights
+copyrights:
+  - Copyright (c) 2015, Contributors
+
+

tests/cluecode/data/copyrights/copyright-discussion-false-positive.txt

@@ -0,0 +1,37 @@
+# order to be recognized.  For example, each of these is fine:
+#
+#   Copyright @copyright{} 1990-2005, 2007-2009 Free Software
+#   Foundation, Inc.
+#
+#   # Copyright (C) 1990-2005, 2007-2009 Free Software
+#   # Foundation, Inc.
+#
+#   /*
+#    * Copyright © 90,2005,2007-2009
+#    * Free Software Foundation, Inc.
+#    */
+#
+# However, the following format is not recognized because the line
+# prefix changes after the first line:
+#
+#   ## Copyright (C) 1990-2005, 2007-2009 Free Software
+#   #  Foundation, Inc.
+#
+# However, any correctly formatted copyright statement following
+# a non-matching copyright statements would be recognized.
+#
+#   2. Its format is "Copyright (C)", then a list of copyright years,
+#      and then the name of the copyright holder.
+#   3. The "(C)" takes one of the following forms or is omitted
+#      entirely:
+#
+#        A. (C)
+#        B. (c)
+#        C. @copyright{}
+#        D. ©
+#        E. ©
+#
+#   4. The "Copyright" appears at the beginning of a line, except that it
+#      may be prefixed by any sequence (e.g., a comment) of no more than
+#      5 characters -- including white space.
+

tests/cluecode/data/copyrights/copyright-discussion-false-positive.txt.yml

@@ -0,0 +1,10 @@
+what:
+  - copyrights
+copyrights:
+  - Copyright 1990-2005, 2007-2009 Free Software Foundation, Inc.
+  - Copyright (c) 1990-2005, 2007-2009 Free Software Foundation, Inc.
+  - Copyright (c) 90,2005,2007-2009 Free Software Foundation, Inc.
+  - Copyright (c) 1990-2005, 2007-2009 Free Software Foundation, Inc.
+  - A. (c) B. (c)
+  - C. copyright D. (c)
+

tests/cluecode/data/copyrights/misco/open-contributors.txt

@@ -0,0 +1 @@
+Copyright (c) 2019-2021, Open source contributors.