From d7fc4cc9cf695b17a40896c960f5635953202ea6 Mon Sep 17 00:00:00 2001
From: Ayan Sinha Mahapatra <ayansmahapatra@gmail.com>
Date: Thu, 14 Nov 2024 17:22:19 +0530
Subject: [PATCH] Fix package resource assign bug for jar manifests

Signed-off-by: Ayan Sinha Mahapatra <ayansmahapatra@gmail.com>
---
 src/packagedcode/maven.py                     |  11 +-
 src/packagedcode/models.py                    |   6 +
 ...ackson-dataformat-xml-2.13.5-expected.json |   4 +-
 .../spring-web-5.3.20-jar-expected.json       | 321 ++++++++++++++++++
 .../spring-web-5.3.20.jar                     | Bin 0 -> 180 bytes
 .../META-INF/MANIFEST.MF                      |   6 +
 .../META-INF/notice.txt                       |   6 +
 .../extracted-jar/hsqldb-2.4.0-expected.json  |  16 +-
 tests/packagedcode/test_maven.py              |   9 +-
 9 files changed, 369 insertions(+), 10 deletions(-)
 create mode 100644 tests/packagedcode/data/maven_misc/assemble/spring-web-5.3.20-jar-expected.json
 create mode 100644 tests/packagedcode/data/maven_misc/assemble/spring-web-5.3.20-jar/spring-web-5.3.20.jar
 create mode 100644 tests/packagedcode/data/maven_misc/assemble/spring-web-5.3.20-jar/spring-web-5.3.20.jar-extract/META-INF/MANIFEST.MF
 create mode 100644 tests/packagedcode/data/maven_misc/assemble/spring-web-5.3.20-jar/spring-web-5.3.20.jar-extract/META-INF/notice.txt

diff --git a/src/packagedcode/maven.py b/src/packagedcode/maven.py
index c2ad164a199..c3910f213b6 100644
--- a/src/packagedcode/maven.py
+++ b/src/packagedcode/maven.py
@@ -68,6 +68,9 @@ def assemble(cls, package_data, resource, codebase, package_adder=models.add_to_
             yield from models.DatafileHandler.assemble(package_data, resource, codebase)
             return
 
+        if not package_data.purl:
+            return
+
         datafile_path = resource.path
 
         # This order is important as we want pom.xml to be used for package
@@ -116,11 +119,11 @@ def assemble(cls, package_data, resource, codebase, package_adder=models.add_to_
                     parent_resource=parent_resource,
                 )
         elif manifests and not pom_xmls:
-            yield from JavaJarManifestHandlerMixin.assemble(package_data, resource, codebase)
+            yield from JavaJarManifestHandlerMixin.assemble(package_data, resource, codebase, package_adder)
         elif pom_xmls and not manifests:
-            yield from MavenPomXmlHandlerMixin.assemble(package_data, resource, codebase)
+            yield from MavenPomXmlHandlerMixin.assemble(package_data, resource, codebase, package_adder)
         else:
-            yield from models.DatafileHandler.assemble(package_data, resource, codebase)
+            yield from models.DatafileHandler.assemble(package_data, resource, codebase, package_adder)
 
 
 class JavaJarManifestHandler(MavenBasePackageHandler):
@@ -149,7 +152,7 @@ def assign_package_to_resources(cls, package, resource, codebase, package_adder)
         # we want to root of the jar, two levels up
         parent = resource.parent(codebase)
         if parent:
-            parent = resource.parent(codebase)
+            parent = parent.parent(codebase)
         if parent:
             models.DatafileHandler.assign_package_to_resources(
                 package,
diff --git a/src/packagedcode/models.py b/src/packagedcode/models.py
index 2224c2d5121..45a2f6d4de6 100644
--- a/src/packagedcode/models.py
+++ b/src/packagedcode/models.py
@@ -1199,6 +1199,12 @@ def assign_package_to_resources(cls, package, resource, codebase, package_adder=
         # NOTE: we do not attach files to the Package level. Instead we
         # update `for_packages` of a codebase resource.
         package_uid = package.package_uid
+        if resource.path.endswith("-extract"):
+            archive_resource_path = resource.path.removesuffix("-extract")
+            archive_resource = codebase.get_resource(archive_resource_path)
+            if archive_resource:
+                package_adder(package_uid, archive_resource, codebase)
+
         if resource and package_uid:
             package_adder(package_uid, resource, codebase)
             for res in resource.walk(codebase):
diff --git a/tests/packagedcode/data/maven_misc/assemble/jackson-dataformat-xml-2.13.5-expected.json b/tests/packagedcode/data/maven_misc/assemble/jackson-dataformat-xml-2.13.5-expected.json
index 969783dfcf1..aacbb9fa2e7 100644
--- a/tests/packagedcode/data/maven_misc/assemble/jackson-dataformat-xml-2.13.5-expected.json
+++ b/tests/packagedcode/data/maven_misc/assemble/jackson-dataformat-xml-2.13.5-expected.json
@@ -188,7 +188,9 @@
       "path": "jackson-dataformat-xml-2.13.5",
       "type": "directory",
       "package_data": [],
-      "for_packages": [],
+      "for_packages": [
+        "pkg:jar/Jackson-dataformat-XML@2.13.5?uuid=fixed-uid-done-for-testing-5642512d1758"
+      ],
       "detected_license_expression": null,
       "detected_license_expression_spdx": null,
       "license_detections": [],
diff --git a/tests/packagedcode/data/maven_misc/assemble/spring-web-5.3.20-jar-expected.json b/tests/packagedcode/data/maven_misc/assemble/spring-web-5.3.20-jar-expected.json
new file mode 100644
index 00000000000..a60e601249c
--- /dev/null
+++ b/tests/packagedcode/data/maven_misc/assemble/spring-web-5.3.20-jar-expected.json
@@ -0,0 +1,321 @@
+{
+  "packages": [
+    {
+      "type": "jar",
+      "namespace": null,
+      "name": "spring-web",
+      "version": "5.3.20",
+      "qualifiers": {},
+      "subpath": null,
+      "primary_language": null,
+      "description": null,
+      "release_date": null,
+      "parties": [],
+      "keywords": [],
+      "homepage_url": null,
+      "download_url": null,
+      "size": null,
+      "sha1": null,
+      "md5": null,
+      "sha256": null,
+      "sha512": null,
+      "bug_tracking_url": null,
+      "code_view_url": null,
+      "vcs_url": null,
+      "copyright": null,
+      "holder": null,
+      "declared_license_expression": "apache-2.0",
+      "declared_license_expression_spdx": "Apache-2.0",
+      "license_detections": [
+        {
+          "license_expression": "apache-2.0",
+          "license_expression_spdx": "Apache-2.0",
+          "matches": [
+            {
+              "license_expression": "apache-2.0",
+              "license_expression_spdx": "Apache-2.0",
+              "from_file": "spring-web-5.3.20-jar/spring-web-5.3.20.jar-extract/META-INF/notice.txt",
+              "start_line": 4,
+              "end_line": 6,
+              "matcher": "2-aho",
+              "score": 100.0,
+              "matched_length": 27,
+              "match_coverage": 100.0,
+              "rule_relevance": 100,
+              "rule_identifier": "apache-2.0_701.RULE",
+              "rule_url": "https://github.com/nexB/scancode-toolkit/tree/develop/src/licensedcode/data/rules/apache-2.0_701.RULE"
+            }
+          ],
+          "detection_log": [],
+          "identifier": "apache_2_0-dda7c7b6-8825-2c52-d55b-46f66d469baa"
+        }
+      ],
+      "other_license_expression": null,
+      "other_license_expression_spdx": null,
+      "other_license_detections": [],
+      "extracted_license_statement": null,
+      "notice_text": null,
+      "source_packages": [],
+      "is_private": false,
+      "is_virtual": false,
+      "extra_data": {},
+      "repository_homepage_url": null,
+      "repository_download_url": null,
+      "api_data_url": null,
+      "package_uid": "pkg:jar/spring-web@5.3.20?uuid=fixed-uid-done-for-testing-5642512d1758",
+      "datafile_paths": [
+        "spring-web-5.3.20-jar/spring-web-5.3.20.jar-extract/META-INF/MANIFEST.MF"
+      ],
+      "datasource_ids": [
+        "java_jar_manifest"
+      ],
+      "purl": "pkg:jar/spring-web@5.3.20"
+    }
+  ],
+  "dependencies": [],
+  "license_detections": [
+    {
+      "identifier": "apache_2_0-dda7c7b6-8825-2c52-d55b-46f66d469baa",
+      "license_expression": "apache-2.0",
+      "license_expression_spdx": "Apache-2.0",
+      "detection_count": 2,
+      "detection_log": [],
+      "reference_matches": [
+        {
+          "license_expression": "apache-2.0",
+          "license_expression_spdx": "Apache-2.0",
+          "from_file": "spring-web-5.3.20-jar/spring-web-5.3.20.jar-extract/META-INF/notice.txt",
+          "start_line": 4,
+          "end_line": 6,
+          "matcher": "2-aho",
+          "score": 100.0,
+          "matched_length": 27,
+          "match_coverage": 100.0,
+          "rule_relevance": 100,
+          "rule_identifier": "apache-2.0_701.RULE",
+          "rule_url": "https://github.com/nexB/scancode-toolkit/tree/develop/src/licensedcode/data/rules/apache-2.0_701.RULE"
+        }
+      ]
+    }
+  ],
+  "files": [
+    {
+      "path": "spring-web-5.3.20-jar",
+      "type": "directory",
+      "package_data": [],
+      "for_packages": [],
+      "detected_license_expression": null,
+      "detected_license_expression_spdx": null,
+      "license_detections": [],
+      "license_clues": [],
+      "percentage_of_license_text": 0,
+      "scan_errors": []
+    },
+    {
+      "path": "spring-web-5.3.20-jar/spring-web-5.3.20.jar",
+      "type": "file",
+      "package_data": [
+        {
+          "type": "jar",
+          "namespace": null,
+          "name": null,
+          "version": null,
+          "qualifiers": {},
+          "subpath": null,
+          "primary_language": null,
+          "description": null,
+          "release_date": null,
+          "parties": [],
+          "keywords": [],
+          "homepage_url": null,
+          "download_url": null,
+          "size": null,
+          "sha1": null,
+          "md5": null,
+          "sha256": null,
+          "sha512": null,
+          "bug_tracking_url": null,
+          "code_view_url": null,
+          "vcs_url": null,
+          "copyright": null,
+          "holder": null,
+          "declared_license_expression": null,
+          "declared_license_expression_spdx": null,
+          "license_detections": [],
+          "other_license_expression": null,
+          "other_license_expression_spdx": null,
+          "other_license_detections": [],
+          "extracted_license_statement": null,
+          "notice_text": null,
+          "source_packages": [],
+          "file_references": [],
+          "is_private": false,
+          "is_virtual": false,
+          "extra_data": {},
+          "dependencies": [],
+          "repository_homepage_url": null,
+          "repository_download_url": null,
+          "api_data_url": null,
+          "datasource_id": "java_jar",
+          "purl": null
+        }
+      ],
+      "for_packages": [
+        "pkg:jar/spring-web@5.3.20?uuid=fixed-uid-done-for-testing-5642512d1758"
+      ],
+      "detected_license_expression": null,
+      "detected_license_expression_spdx": null,
+      "license_detections": [],
+      "license_clues": [],
+      "percentage_of_license_text": 0,
+      "scan_errors": []
+    },
+    {
+      "path": "spring-web-5.3.20-jar/spring-web-5.3.20.jar-extract",
+      "type": "directory",
+      "package_data": [],
+      "for_packages": [
+        "pkg:jar/spring-web@5.3.20?uuid=fixed-uid-done-for-testing-5642512d1758"
+      ],
+      "detected_license_expression": null,
+      "detected_license_expression_spdx": null,
+      "license_detections": [],
+      "license_clues": [],
+      "percentage_of_license_text": 0,
+      "scan_errors": []
+    },
+    {
+      "path": "spring-web-5.3.20-jar/spring-web-5.3.20.jar-extract/META-INF",
+      "type": "directory",
+      "package_data": [],
+      "for_packages": [
+        "pkg:jar/spring-web@5.3.20?uuid=fixed-uid-done-for-testing-5642512d1758"
+      ],
+      "detected_license_expression": null,
+      "detected_license_expression_spdx": null,
+      "license_detections": [],
+      "license_clues": [],
+      "percentage_of_license_text": 0,
+      "scan_errors": []
+    },
+    {
+      "path": "spring-web-5.3.20-jar/spring-web-5.3.20.jar-extract/META-INF/MANIFEST.MF",
+      "type": "file",
+      "package_data": [
+        {
+          "type": "jar",
+          "namespace": null,
+          "name": "spring-web",
+          "version": "5.3.20",
+          "qualifiers": {},
+          "subpath": null,
+          "primary_language": null,
+          "description": "",
+          "release_date": null,
+          "parties": [],
+          "keywords": [],
+          "homepage_url": null,
+          "download_url": null,
+          "size": null,
+          "sha1": null,
+          "md5": null,
+          "sha256": null,
+          "sha512": null,
+          "bug_tracking_url": null,
+          "code_view_url": null,
+          "vcs_url": null,
+          "copyright": null,
+          "holder": null,
+          "declared_license_expression": "apache-2.0",
+          "declared_license_expression_spdx": "Apache-2.0",
+          "license_detections": [
+            {
+              "license_expression": "apache-2.0",
+              "license_expression_spdx": "Apache-2.0",
+              "matches": [
+                {
+                  "license_expression": "apache-2.0",
+                  "license_expression_spdx": "Apache-2.0",
+                  "from_file": "spring-web-5.3.20-jar/spring-web-5.3.20.jar-extract/META-INF/notice.txt",
+                  "start_line": 4,
+                  "end_line": 6,
+                  "matcher": "2-aho",
+                  "score": 100.0,
+                  "matched_length": 27,
+                  "match_coverage": 100.0,
+                  "rule_relevance": 100,
+                  "rule_identifier": "apache-2.0_701.RULE",
+                  "rule_url": "https://github.com/nexB/scancode-toolkit/tree/develop/src/licensedcode/data/rules/apache-2.0_701.RULE"
+                }
+              ],
+              "detection_log": [],
+              "identifier": "apache_2_0-dda7c7b6-8825-2c52-d55b-46f66d469baa"
+            }
+          ],
+          "other_license_expression": null,
+          "other_license_expression_spdx": null,
+          "other_license_detections": [],
+          "extracted_license_statement": null,
+          "notice_text": null,
+          "source_packages": [],
+          "file_references": [],
+          "is_private": false,
+          "is_virtual": false,
+          "extra_data": {},
+          "dependencies": [],
+          "repository_homepage_url": null,
+          "repository_download_url": null,
+          "api_data_url": null,
+          "datasource_id": "java_jar_manifest",
+          "purl": "pkg:jar/spring-web@5.3.20"
+        }
+      ],
+      "for_packages": [
+        "pkg:jar/spring-web@5.3.20?uuid=fixed-uid-done-for-testing-5642512d1758"
+      ],
+      "detected_license_expression": null,
+      "detected_license_expression_spdx": null,
+      "license_detections": [],
+      "license_clues": [],
+      "percentage_of_license_text": 0,
+      "scan_errors": []
+    },
+    {
+      "path": "spring-web-5.3.20-jar/spring-web-5.3.20.jar-extract/META-INF/notice.txt",
+      "type": "file",
+      "package_data": [],
+      "for_packages": [
+        "pkg:jar/spring-web@5.3.20?uuid=fixed-uid-done-for-testing-5642512d1758"
+      ],
+      "detected_license_expression": "apache-2.0",
+      "detected_license_expression_spdx": "Apache-2.0",
+      "license_detections": [
+        {
+          "license_expression": "apache-2.0",
+          "license_expression_spdx": "Apache-2.0",
+          "matches": [
+            {
+              "license_expression": "apache-2.0",
+              "license_expression_spdx": "Apache-2.0",
+              "from_file": "spring-web-5.3.20-jar/spring-web-5.3.20.jar-extract/META-INF/notice.txt",
+              "start_line": 4,
+              "end_line": 6,
+              "matcher": "2-aho",
+              "score": 100.0,
+              "matched_length": 27,
+              "match_coverage": 100.0,
+              "rule_relevance": 100,
+              "rule_identifier": "apache-2.0_701.RULE",
+              "rule_url": "https://github.com/nexB/scancode-toolkit/tree/develop/src/licensedcode/data/rules/apache-2.0_701.RULE"
+            }
+          ],
+          "detection_log": [],
+          "identifier": "apache_2_0-dda7c7b6-8825-2c52-d55b-46f66d469baa"
+        }
+      ],
+      "license_clues": [],
+      "percentage_of_license_text": 71.05,
+      "scan_errors": []
+    }
+  ]
+}
\ No newline at end of file
diff --git a/tests/packagedcode/data/maven_misc/assemble/spring-web-5.3.20-jar/spring-web-5.3.20.jar b/tests/packagedcode/data/maven_misc/assemble/spring-web-5.3.20-jar/spring-web-5.3.20.jar
new file mode 100644
index 0000000000000000000000000000000000000000..46ba8b458e9b899fb130668d12db60d1eea34ddb
GIT binary patch
literal 180
zcmWIWW@Zs#00E=kyhtzuN^mkLF!;KLIO=-(x#@?7@G`Lfcww3j!=)A642&!<nHd<s
jM1VIVlL#|zgJ1^#bp)|UFqV}Kq=69#!-2Frh{FH?EYBh~

literal 0
HcmV?d00001

diff --git a/tests/packagedcode/data/maven_misc/assemble/spring-web-5.3.20-jar/spring-web-5.3.20.jar-extract/META-INF/MANIFEST.MF b/tests/packagedcode/data/maven_misc/assemble/spring-web-5.3.20-jar/spring-web-5.3.20.jar-extract/META-INF/MANIFEST.MF
new file mode 100644
index 00000000000..0e95f20e340
--- /dev/null
+++ b/tests/packagedcode/data/maven_misc/assemble/spring-web-5.3.20-jar/spring-web-5.3.20.jar-extract/META-INF/MANIFEST.MF
@@ -0,0 +1,6 @@
+Manifest-Version: 1.0
+Implementation-Title: spring-web
+Automatic-Module-Name: spring.web
+Implementation-Version: 5.3.20
+Created-By: 1.8.0_332 (Oracle Corporation)
+
diff --git a/tests/packagedcode/data/maven_misc/assemble/spring-web-5.3.20-jar/spring-web-5.3.20.jar-extract/META-INF/notice.txt b/tests/packagedcode/data/maven_misc/assemble/spring-web-5.3.20-jar/spring-web-5.3.20.jar-extract/META-INF/notice.txt
new file mode 100644
index 00000000000..afbdf6fdd6a
--- /dev/null
+++ b/tests/packagedcode/data/maven_misc/assemble/spring-web-5.3.20-jar/spring-web-5.3.20.jar-extract/META-INF/notice.txt
@@ -0,0 +1,6 @@
+Spring Framework 5.3.20
+Copyright (c) 2002-2022 Pivotal, Inc.
+
+This product is licensed to you under the Apache License, Version 2.0
+(the "License"). You may not use this product except in compliance with
+the License.
diff --git a/tests/packagedcode/data/maven_misc/extracted-jar/hsqldb-2.4.0-expected.json b/tests/packagedcode/data/maven_misc/extracted-jar/hsqldb-2.4.0-expected.json
index c6b6284291a..df5ce0a6e5e 100644
--- a/tests/packagedcode/data/maven_misc/extracted-jar/hsqldb-2.4.0-expected.json
+++ b/tests/packagedcode/data/maven_misc/extracted-jar/hsqldb-2.4.0-expected.json
@@ -63,7 +63,9 @@
       "path": "hsqldb-2.4.0.jar-extract",
       "type": "directory",
       "package_data": [],
-      "for_packages": [],
+      "for_packages": [
+        "pkg:osgi/org.hsqldb.hsqldb@2.4.0?uuid=fixed-uid-done-for-testing-5642512d1758"
+      ],
       "scan_errors": []
     },
     {
@@ -156,21 +158,27 @@
       "path": "hsqldb-2.4.0.jar-extract/org",
       "type": "directory",
       "package_data": [],
-      "for_packages": [],
+      "for_packages": [
+        "pkg:osgi/org.hsqldb.hsqldb@2.4.0?uuid=fixed-uid-done-for-testing-5642512d1758"
+      ],
       "scan_errors": []
     },
     {
       "path": "hsqldb-2.4.0.jar-extract/org/hsqldb",
       "type": "directory",
       "package_data": [],
-      "for_packages": [],
+      "for_packages": [
+        "pkg:osgi/org.hsqldb.hsqldb@2.4.0?uuid=fixed-uid-done-for-testing-5642512d1758"
+      ],
       "scan_errors": []
     },
     {
       "path": "hsqldb-2.4.0.jar-extract/org/hsqldb/Database.class",
       "type": "file",
       "package_data": [],
-      "for_packages": [],
+      "for_packages": [
+        "pkg:osgi/org.hsqldb.hsqldb@2.4.0?uuid=fixed-uid-done-for-testing-5642512d1758"
+      ],
       "scan_errors": []
     }
   ]
diff --git a/tests/packagedcode/test_maven.py b/tests/packagedcode/test_maven.py
index f344f933e0d..f11491718c2 100644
--- a/tests/packagedcode/test_maven.py
+++ b/tests/packagedcode/test_maven.py
@@ -249,7 +249,7 @@ def test_maven_assembly_with_pom_and_jar_manifest(self):
         expected_file = self.get_test_loc('maven_misc/assemble/numbers-1.7.4-expected.json')
         run_scan_click(['--package', '--license', '--license-diagnostics', '--processes', '-1', test_dir, '--json', result_file])
         check_json_scan(expected_file, result_file, remove_uuid=True, regen=REGEN_TEST_FIXTURES)
-    
+
     def test_maven_unknown_reference_to_license_in_manifest(self):
         test_dir = self.get_test_loc('maven_misc/assemble/jackson-dataformat-xml-2.13.5')
         result_file = self.get_temp_file('json')
@@ -257,6 +257,13 @@ def test_maven_unknown_reference_to_license_in_manifest(self):
         run_scan_click(['--package', '--license', '--license-diagnostics', '--processes', '-1', test_dir, '--json', result_file])
         check_json_scan(expected_file, result_file, remove_uuid=True, regen=REGEN_TEST_FIXTURES)
 
+    def test_maven_package_assembly_jar_manifest(self):
+        test_dir = self.get_test_loc('maven_misc/assemble/spring-web-5.3.20-jar')
+        result_file = self.get_temp_file('json')
+        expected_file = self.get_test_loc('maven_misc/assemble/spring-web-5.3.20-jar-expected.json')
+        run_scan_click(['--package', '--license', '--license-diagnostics', '--processes', '-1', test_dir, '--json', result_file])
+        check_json_scan(expected_file, result_file, remove_uuid=True, regen=REGEN_TEST_FIXTURES)
+
     def test_package_dependency_not_missing(self):
         test_file = self.get_test_loc('maven2/log4j/log4j-pom.xml')
         self.check_parse_to_package(test_file, regen=REGEN_TEST_FIXTURES)