From 76cd5b362d434b5855d0b3d178b40457a40e24bf Mon Sep 17 00:00:00 2001
From: Jono Yang <jyang@nexb.com>
Date: Mon, 16 Oct 2023 18:38:18 -0700
Subject: [PATCH] Create command to copy packages and related

    * Add minecode database

Signed-off-by: Jono Yang <jyang@nexb.com>
---
 .../commands/copy_legacy_package_data.py      | 71 +++++++++++++++++++
 purldb/settings.py                            | 11 ++-
 2 files changed, 81 insertions(+), 1 deletion(-)
 create mode 100644 packagedb/management/commands/copy_legacy_package_data.py

diff --git a/packagedb/management/commands/copy_legacy_package_data.py b/packagedb/management/commands/copy_legacy_package_data.py
new file mode 100644
index 00000000..5714c20d
--- /dev/null
+++ b/packagedb/management/commands/copy_legacy_package_data.py
@@ -0,0 +1,71 @@
+#
+# Copyright (c) nexB Inc. and others. All rights reserved.
+# purldb is a trademark of nexB Inc.
+# SPDX-License-Identifier: Apache-2.0
+# See http://www.apache.org/licenses/LICENSE-2.0 for the license text.
+# See https://github.com/nexB/purldb for support or download.
+# See https://aboutcode.org for more information about nexB OSS projects.
+#
+
+import logging
+import sys
+
+from minecode.management.commands import VerboseCommand
+from packagedb.models import Package, DependentPackage, Party
+
+TRACE = False
+
+logger = logging.getLogger(__name__)
+logging.basicConfig(stream=sys.stdout)
+logger.setLevel(logging.INFO)
+
+
+class Command(VerboseCommand):
+    def handle(self, *args, **options):
+        packages = Package.objects.using('minecode').all()
+        package_count = packages.count()
+        iterator = packages.iterator(chunk_size=2000)
+        unsaved_packages = []
+        unsaved_dependencies = []
+        unsaved_parties = []
+
+        i = 0
+        for package in iterator:
+            if Package.objects.filter(download_url=package.download_url).exists():
+                continue
+            if not (i % 100) and unsaved_packages:
+                Package.objects.bulk_create(
+                    unsaved_packages
+                )
+                DependentPackage.objects.bulk_create(
+                    unsaved_dependencies
+                )
+                Party.objects.bulk_create(
+                    unsaved_parties
+                )
+                unsaved_packages = []
+                unsaved_dependencies = []
+                unsaved_parties = []
+                print(f"  {i:,} / {package_count:,} saved")
+            else:
+                unsaved_packages.append(package)
+                dependencies = package.dependencies.all()
+                unsaved_dependencies.extend(list(dependencies))
+                parties = package.parties.all()
+                unsaved_parties.extend(list(parties))
+                i += 1
+
+        if unsaved_packages:
+            Package.objects.bulk_create(
+                unsaved_packages
+            )
+            DependentPackage.objects.bulk_create(
+                unsaved_dependencies
+            )
+            Party.objects.bulk_create(
+                unsaved_parties
+            )
+            unsaved_packages = []
+            unsaved_dependencies = []
+            unsaved_parties = []
+            print(f"  {i:,} / {package_count:,} saved")
diff --git a/purldb/settings.py b/purldb/settings.py
index 02992fd3..56846a41 100644
--- a/purldb/settings.py
+++ b/purldb/settings.py
@@ -110,7 +110,16 @@
         'PASSWORD': env.str('PACKAGEDB_DB_PASSWORD', 'packagedb'),
         'PORT': env.str('PACKAGEDB_DB_PORT', '5432'),
         'ATOMIC_REQUESTS': True,
-    }
+    },
+    'minecode': {
+        'ENGINE': env.str('PACKAGEDB_DB_ENGINE', 'django.db.backends.postgresql'),
+        'HOST': env.str('PACKAGEDB_DB_HOST', 'localhost'),
+        'NAME': 'minecode',
+        'USER': 'minecode',
+        'PASSWORD': 'minecode',
+        'PORT': env.str('PACKAGEDB_DB_PORT', '5432'),
+        'ATOMIC_REQUESTS': True,
+    },
 }
 
 DEFAULT_AUTO_FIELD = "django.db.models.AutoField"