Fix unknown license clues bug

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

Author: AyanSinhaMahapatra

etc/scripts/licenses/buildrules-template.txt

@@ -1,18 +1,22 @@
 ----------------------------------------
-license_expression: 
-relevance: 100
-minimum_coverage: 90
+license_expression: apache-2.0
 is_license_notice: yes
-is_license_text: yes
-is_license_reference: yes
-is_license_tag: yes
-is_false_positive: yes
-is_license_intro: yes
-is_license_clue: yes
 referenced_filenames:
-notes: 
+    - ASL2.0
+notes: seen in woodstox
 ---
+This copy of is licensed under the
+{{Apache (Software) License, version 2.0 ("the License")}}.
+See the License for details about distribution rights, and the
+specific rights regarding derivate works.
+
+You may obtain a copy of the License at:
+
+http://www.apache.org/licenses/
 
+A copy is also included in the downloadable source code package
+containing , in file {{"ASL2.0"}}, under the same directory
+as this file.
 ----------------------------------------
 license_expression: 
 relevance: 100

src/licensedcode/data/rules/license-clue_1.RULE

@@ -0,0 +1,7 @@
+---
+license_expression: unknown-license-reference
+is_license_clue: yes
+notes: Seen in woodstox
+---
+
+This product currently only contains code developed by authors of specific components, as identified by the source code files.

src/licensedcode/data/rules/unknown-license-reference_354.RULE

@@ -1,6 +1,7 @@
 ---
 license_expression: unknown-license-reference
 is_license_reference: yes
+is_deprecated: yes
 notes: Seen in woodstox
 ---
 

src/licensedcode/detection.py

@@ -927,10 +927,21 @@ def is_undetected_license_matches(license_matches):
         return True
 
 
+def is_correct_detection_non_unknown(license_matches):
+    """
+    Return True if all the matches in ``license_matches`` List of LicenseMatch
+    are correct/perfect license detections and also there aren't any unknowns.
+    """
+    return (
+        is_correct_detection(license_matches)
+        and not has_unknown_matches(license_matches)
+    )
+
+
 def is_correct_detection(license_matches):
     """
     Return True if all the matches in ``license_matches`` List of LicenseMatch
-    are correct license detections.
+    are perfect license detections.
     """
     matchers = (license_match.matcher for license_match in license_matches)
     is_match_coverage_perfect = [
@@ -940,7 +951,7 @@ def is_correct_detection(license_matches):
 
     return (
         all(matcher in ("1-hash", "1-spdx-id", "2-aho") for matcher in matchers)
-        and all(is_match_coverage_perfect) and not has_unknown_matches(license_matches)
+        and all(is_match_coverage_perfect)
     )
 
 
@@ -1088,6 +1099,14 @@ def is_unknown_intro(license_match):
     )
 
 
+def has_correct_license_clue_matches(license_matches):
+    """
+    Return True if all the matches in ``license_matches`` List of LicenseMatch
+    has True for the `is_license_clue` rule attribute.
+    """
+    return is_correct_detection(license_matches) and all(match.rule.is_license_clue for match in license_matches)
+
+
 def is_license_clues(license_matches):
     """
     Return True if the license_matches are not part of a correct
@@ -1480,9 +1499,12 @@ def analyze_detection(license_matches, package_license=False):
         return DetectionCategory.FALSE_POSITVE.value
 
     # Case where all matches have `matcher` as `1-hash` or `4-spdx-id`
-    elif is_correct_detection(license_matches=license_matches):
+    elif is_correct_detection_non_unknown(license_matches=license_matches):
         return DetectionCategory.PERFECT_DETECTION.value
 
+    elif has_correct_license_clue_matches(license_matches=license_matches):
+        return DetectionCategory.LICENSE_CLUES.value
+
     # Case where even though the matches have perfect coverage, they have
     # matches with `unknown` rule identifiers
     elif has_unknown_matches(license_matches=license_matches):

tests/licensedcode/data/plugin_license/clues/woodstox.expected.json

@@ -0,0 +1,6 @@
+
+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+
+   

tests/licensedcode/data/plugin_license/clues/woodstox/ASL2.0

@@ -0,0 +1,12 @@
+This copy of Woodstox XML processor is licensed under the
+Apache (Software) License, version 2.0 ("the License").
+See the License for details about distribution rights, and the
+specific rights regarding derivate works.
+
+You may obtain a copy of the License at:
+
+http://www.apache.org/licenses/
+
+A copy is also included in the downloadable source code package
+containing Woodstox, in file "ASL2.0", under the same directory
+as this file.

tests/licensedcode/data/plugin_license/clues/woodstox/LICENSE

@@ -0,0 +1,11 @@
+Manifest-Version: 1.0
+Ant-Version: Apache Ant 1.6.5
+Created-By: 1.5.0_14-b03 (Sun Microsystems Inc.)
+Built-By: tatu
+Specification-Title: StAX 1.0 API
+Specification-Version: 1.0
+Specification-Vendor: http://jcp.org/en/jsr/detail?id=173
+Implementation-Title: WoodSToX XML-processor
+Implementation-Version: 3.2.8
+Implementation-Vendor: woodstox.codehaus.org
+

tests/licensedcode/data/plugin_license/clues/woodstox/META-INF/MANIFEST.MF

@@ -0,0 +1,8 @@
+This product currently only contains code developed by authors
+of specific components, as identified by the source code files.
+
+Since product implements StAX API, it has dependencies to StAX API
+classes.
+
+For additional credits (generally to people who reported problems)
+see CREDITS file.

tests/licensedcode/data/plugin_license/clues/woodstox/NOTICE

@@ -74,6 +74,25 @@ def test_license_match_unknown_license_intro_with_dual_license():
     check_json_scan(test_loc, result_file, regen=REGEN_TEST_FIXTURES)
 
 
+def test_license_match_unknown_clues_is_not_in_expression():
+    test_dir = test_env.get_test_loc('plugin_license/clues/woodstox/', copy=True)
+    result_file = test_env.get_temp_file('json')
+    args = [
+        '--license',
+        '--license-text',
+        '--license-text-diagnostics',
+        '--license-diagnostics',
+        '--license-references',
+        '--strip-root',
+        '--verbose',
+        '--json', result_file,
+        test_dir,
+    ]
+    run_scan_click(args)
+    test_loc = test_env.get_test_loc('plugin_license/clues/woodstox.expected.json')
+    check_json_scan(test_loc, result_file, regen=REGEN_TEST_FIXTURES)
+
+
 def test_license_match_unknown_license_intro_eclipse_foundation():
     test_dir = test_env.get_test_loc('plugin_license/unknown_intro/scan-unknown-intro-eclipse-foundation/', copy=True)
     result_file = test_env.get_temp_file('json')