Refactor and address feedback

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

Author: AyanSinhaMahapatra

src/packagedcode/__init__.py

@@ -254,11 +254,6 @@ def get_package_handler(package_data):
     return ppc
 
 
-PACKAGE_DATA_CLASS_WITH_CUSTOM_LICENSE_DETECTION = [
-    maven.MavenPackageData
-]
-
 PACKAGE_DATA_CLASS_BY_DATASOURCE_ID = {
-    package_data_class.datasource_id: package_data_class
-    for package_data_class in PACKAGE_DATA_CLASS_WITH_CUSTOM_LICENSE_DETECTION
+    maven.MavenPackageData.datasource_id: maven.MavenPackageData
 }

src/packagedcode/jar_manifest.py

@@ -248,7 +248,7 @@ def dget(s):
     package['homepage_url'] = dget('Implementation-URL') or dget('Implementation-Url')
 
     # Bundle-DocURL: http://logging.apache.org/log4j/1.2
-    #package['documentation_url'] = dget('Bundle-DocURL')
+    doc_url = dget('Bundle-DocURL')
 
     # vendor/owner/contact
     #########################
@@ -339,7 +339,15 @@ def dget(s):
     # Deps:
     # Require-Bundle
 
-    #package['notes'] = dget('Comment')
+    comment = dget('Comment')
+
+    if comment or doc_url:
+        package["extra_data"] = {}
+        if comment:
+            package["extra_data"]['notes'] = comment
+        if doc_url:
+            package["extra_data"]['documentation_url'] = doc_url
+
     return package
 
 
@@ -378,16 +386,20 @@ def parse_scm_connection(scm_connection):
 
 
 def get_datasource_id(package_type):
+    """
+    Get the corresponding `datasource_id` for the given
+    `package_type`. This is a seperate function to avoid
+    cyclic imports.
+    """
     from packagedcode.maven import JavaJarManifestHandler
     from packagedcode.maven import JavaOSGiManifestHandler
 
-    if package_type == 'maven':
-        return JavaJarManifestHandler.datasource_id
-    elif package_type == 'jar':
+    if package_type in ['maven', 'jar']:
         return JavaJarManifestHandler.datasource_id
     elif package_type == 'osgi':
         return JavaOSGiManifestHandler.datasource_id
 
+
 def is_id(s):
     """
     Return True if `s` is some kind of id.

src/packagedcode/licensing.py

@@ -775,8 +775,8 @@ def get_license_detections_and_expression(
     if not extracted_license_statement:
         return detection_data, license_expression
 
-    if datasource_id in PACKAGE_DATA_CLASS_BY_DATASOURCE_ID:
-        package_data_class = PACKAGE_DATA_CLASS_BY_DATASOURCE_ID.get(datasource_id, None)
+    package_data_class = PACKAGE_DATA_CLASS_BY_DATASOURCE_ID.get(datasource_id)
+    if package_data_class:
         license_detections = package_data_class.get_license_detections_for_extracted_license_statement(
             extracted_license=extracted_license_statement,
             try_as_expression=try_as_expression,

src/packagedcode/maven.py

@@ -60,6 +60,9 @@ class MavenBasePackageHandler(models.DatafileHandler):
     @classmethod
     def assemble(cls, package_data, resource, codebase, package_adder=models.add_to_package):
         """
+        Assembles from codebases where both a pom.xml and a MANIFEST.MF is present,
+        otherwise uses the default assemble function and `assign_package_to_resources`
+        function from respective DatafileHandlers.
         """
         if codebase.has_single_resource:
             yield from models.DatafileHandler.assemble(package_data, resource, codebase)
@@ -1308,11 +1311,11 @@ def get_license_detections_for_extracted_license_statement(
 
         for license_entry in new_extracted_license:
             license_entry.pop("distribution")
-            if not license_entry.get("name", None):
+            if not license_entry.get("name"):
                 license_entry.pop("name")
-            if not license_entry.get("url", None):
+            if not license_entry.get("url"):
                 license_entry.pop("url")
-            if not license_entry.get("comments", None):
+            if not license_entry.get("comments"):
                 license_entry.pop("comments")
 
         extracted_license_statement = saneyaml.dump(new_extracted_license)

tests/packagedcode/data/maven_misc/assemble/johnzon-jsonb-1.2.11-expected.json

@@ -427,7 +427,9 @@
           "notice_text": null,
           "source_packages": [],
           "file_references": [],
-          "extra_data": {},
+          "extra_data": {
+            "documentation_url": "http://johnzon.apache.org/"
+          },
           "dependencies": [],
           "repository_homepage_url": null,
           "repository_download_url": null,

tests/packagedcode/data/maven_misc/manifest/10/MANIFEST.MF.package-data.json

@@ -15,5 +15,8 @@
     }
   ],
   "vcs_url": null,
-  "code_view_url": null
+  "code_view_url": null,
+  "extra_data": {
+    "documentation_url": "http://www.hibernate.org/orm/5.3"
+  }
 }

tests/packagedcode/data/maven_misc/manifest/24/MANIFEST.MF.package-data.json

@@ -15,5 +15,8 @@
     }
   ],
   "vcs_url": null,
-  "code_view_url": null
+  "code_view_url": null,
+  "extra_data": {
+    "documentation_url": "http://www.oracle.com"
+  }
 }

tests/packagedcode/data/maven_misc/manifest/4/MANIFEST.MF.package-data.json

@@ -15,5 +15,8 @@
     }
   ],
   "vcs_url": null,
-  "code_view_url": null
+  "code_view_url": null,
+  "extra_data": {
+    "documentation_url": "http://logging.apache.org/log4j/1.2"
+  }
 }

tests/packagedcode/data/maven_misc/manifest/5/MANIFEST.MF.package-data.json

@@ -15,5 +15,8 @@
     }
   ],
   "vcs_url": null,
-  "code_view_url": null
+  "code_view_url": null,
+  "extra_data": {
+    "documentation_url": "http://www.oracle.com"
+  }
 }

tests/packagedcode/data/maven_misc/manifest/7/MANIFEST.MF.package-data.json

@@ -15,5 +15,8 @@
     }
   ],
   "vcs_url": null,
-  "code_view_url": null
+  "code_view_url": null,
+  "extra_data": {
+    "documentation_url": "http://fusesource.com/"
+  }
 }