Add support to find source repo for a package

Signed-off-by: Tushar Goel <[email protected]>

minecode/tests/test_maven.py

@@ -13,6 +13,7 @@
 import re
 
 from mock import patch
+from unittest import mock
 
 from django.test import TestCase as DjangoTestCase
 
@@ -829,3 +830,28 @@ def test_merge_ancestors(self, regen=False):
         maven_visitor.merge_ancestors(ancestor_pom_texts, package)
         expected_after_loc = self.get_test_loc('maven/pom/pulsar-client-1x-2.5.1-package_after.json')
         self.check_expected_results(package.to_dict(), expected_after_loc, regen=regen)
+
+    @mock.patch("minecode.visitors.maven.get_pom_text")
+    def test_get_merged_ancestor_package_from_maven_package(self, get_pom_text_mock, regen=False):
+        get_pom_text_mock.return_value = ""
+        ancestor_pom_texts = []
+        with patch("minecode.visitors.maven.get_ancestry") as mock_get_ancestry:
+            for loc in [
+                self.get_test_loc('maven/pom/apache-18.pom'),
+                self.get_test_loc('maven/pom/pulsar-2.5.1.pom'),
+                self.get_test_loc('maven/pom/pulsar-client-1x-base-2.5.1.pom')
+            ]:
+                with open(loc) as f:
+                    pom_text = f.read()
+                ancestor_pom_texts.append(pom_text)
+            mock_get_ancestry.return_value = ancestor_pom_texts
+            db_package = packagedb.models.Package.objects.create(
+                name="pulsar-client",
+                namespace="org.apache.pulsar",
+                version="2.5.1",
+                type="maven",
+                download_url="https://repo1.maven.org/maven2/org/apache/pulsar/pulsar-client/2.5.1/pulsar-client-2.5.1.jar",
+            )
+            merged_package = maven_visitor.get_merged_ancestor_package_from_maven_package(package=db_package)
+            expected_loc = self.get_test_loc('maven/pom/pulsar-client-merged-ancestor-package.json')
+            self.check_expected_results(merged_package.to_dict(), expected_loc, regen=regen)

minecode/tests/testfiles/maven/pom/pulsar-client-merged-ancestor-package.json

@@ -0,0 +1,50 @@
+{
+  "type":"maven",
+  "namespace":"org.apache.pulsar",
+  "name":"pulsar-client",
+  "version":"2.5.1",
+  "qualifiers":"",
+  "subpath":"",
+  "package_sets":[],
+  "package_content":null,
+  "primary_language":null,
+  "description":null,
+  "release_date":null,
+  "parties":[],
+  "keywords":[],
+  "homepage_url":"https://www.apache.org/",
+  "download_url":"https://repo1.maven.org/maven2/org/apache/pulsar/pulsar-client/2.5.1/pulsar-client-2.5.1.jar",
+  "size":null,
+  "md5":null,
+  "sha1":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":null,
+  "license_detections":[],
+  "other_license_expression":null,
+  "other_license_expression_spdx":null,
+  "other_license_detections":[],
+  "extracted_license_statement":null,
+  "notice_text":null,
+  "source_packages":[],
+  "extra_data":{
+    "history":[
+      "Field `declared_license_expression` has been updated using values obtained from the parent POM pkg:maven/org.apache/apache@18",
+      "Field `homepage_url` has been updated using values obtained from the parent POM pkg:maven/org.apache/apache@18"
+    ]
+  },
+  "dependencies":[],
+  "package_uid":"pkg:maven/org.apache.pulsar/[email protected]?uuid=fixed-uid-done-for-testing-5642512d1758",
+  "datasource_id":null,
+  "purl":"pkg:maven/org.apache.pulsar/[email protected]",
+  "repository_homepage_url":null,
+  "repository_download_url":null,
+  "api_data_url":null,
+  "file_references":[]
+}

minecode/visitors/maven.py

@@ -13,6 +13,7 @@
 import io
 import json
 import logging
+from typing import Dict
 
 import arrow
 import requests
@@ -109,6 +110,8 @@ def get_pom_text(namespace, name, version, qualifiers={}):
     field arguments in a string.
     """
     # Create URLs using purl fields
+    if qualifiers and not isinstance(qualifiers, Dict):
+        return
     urls = get_urls(
         namespace=namespace,
         name=name,
@@ -152,6 +155,8 @@ def fetch_parent(pom_text):
     """
     Return the parent pom text of `pom_text`, or None if `pom_text` has no parent.
     """
+    if not pom_text:
+        return
     pom = get_maven_pom(text=pom_text)
     if (
         pom.parent
@@ -188,6 +193,26 @@ def get_ancestry(pom_text):
     return reversed(ancestors)
 
 
+def get_merged_ancestor_package_from_maven_package(package):
+    """
+    Merge package details of a package with its ancestor pom
+    and return the merged package.
+    """
+    if not package:
+        return
+    pom_text = get_pom_text(
+        name=package.name,
+        namespace=package.namespace,
+        version=package.version,
+        qualifiers=package.qualifiers,
+    )
+    merged_package = merge_ancestors(
+        ancestor_pom_texts=get_ancestry(pom_text),
+        package=package,
+    )
+    return merged_package
+
+
 def merge_parent(package, parent_package):
     """
     Merge `parent_package` data into `package` and return `package.