Modify license clue/false positive heuristics

Signed-off-by: Ayan Sinha Mahapatra <[email protected]>

Author: AyanSinhaMahapatra

src/licensedcode/data/rules/freeware_bare_word_only.RULE

@@ -1,7 +1,7 @@
 ---
 license_expression: unknown-license-reference
 is_license_clue: yes
-relevance: 100
+relevance: 60
 ---
 
 freeware

src/licensedcode/data/rules/gpl3_bare_word_only.RULE

@@ -1,7 +1,7 @@
 ---
 license_expression: gpl-3.0
 is_license_reference: yes
-relevance: 95
+relevance: 60
 ---
 
 GPL3

src/licensedcode/data/rules/lgpl_bare_single_word.RULE

@@ -1,7 +1,7 @@
 ---
 license_expression: lgpl-2.0-plus
 is_license_reference: yes
-relevance: 75
+relevance: 60
 ---
 
 LGPL

src/licensedcode/data/rules/public-domain_bare_words.RULE

@@ -1,6 +1,6 @@
 ---
 license_expression: public-domain
-is_license_reference: yes
+is_license_clue: yes
 is_continuous: yes
 relevance: 70
 ---

src/licensedcode/detection.py

@@ -1163,6 +1163,25 @@ def is_false_positive(license_matches, package_license=False):
     if package_license:
         return False
 
+    has_copyrights = all(
+        True
+        for license_match in license_matches
+        if "copyright" in license_match.matched_text().lower() 
+    )
+    has_full_relevance = all(
+        True
+        for license_match in license_matches
+        if license_match.rule.relevance == 100
+    )
+    if has_copyrights or has_full_relevance:
+        return False
+
+    has_low_relevance = all(
+        True
+        for license_match in license_matches
+        if license_match.rule.relevance < 60
+    )
+
     start_line_region = min(
         license_match.start_line for license_match in license_matches
     )
@@ -1194,13 +1213,13 @@ def is_false_positive(license_matches, package_license=False):
 
     is_single_match = len(license_matches) == 1
 
-    if is_single_match and is_bare_rule:
+    if is_single_match and is_bare_rule and has_low_relevance:
         return True
 
     if is_gpl and all_match_rule_length_one:
         return True
 
-    if start_line_region > FALSE_POSITIVE_START_LINE_THRESHOLD and any(
+    if has_low_relevance and start_line_region > FALSE_POSITIVE_START_LINE_THRESHOLD and any(
         match_rule_length_value <= FALSE_POSITIVE_RULE_LENGTH_THRESHOLD
         for match_rule_length_value in match_rule_length_values
     ):

tests/licensedcode/data/datadriven/lic4/memdump.cocci

@@ -0,0 +1,8 @@
+///
+// Confidence: High
+// Copyright: (C) 2010-2012 Nicolas Palix.  GPLv2.
+// Copyright: (C) 2010-2012 Julia Lawall, INRIA/LIP6.  GPLv2.
+// Copyright: (C) 2010-2012 Gilles Muller, INRIA/LiP6.  GPLv2.
+// URL: http://coccinelle.lip6.fr/
+// Comments:
+// Options: --no-includes --include-headers

tests/licensedcode/data/datadriven/lic4/memdump.cocci.yml

@@ -0,0 +1,2 @@
+license_expressions:
+  - gpl-2.0