add support to store packages/archives locally

scancodeio/settings.py

@@ -20,6 +20,7 @@
 # ScanCode.io is a free software code scanning tool from nexB Inc. and others.
 # Visit https://github.com/aboutcode-org/scancode.io for support and download.
 
+import os
 import sys
 import tempfile
 from pathlib import Path
@@ -367,6 +368,15 @@
     PROJECT_DIR("static"),
 ]
 
+# Media files (Uploaded package archives, etc.)
+
+MEDIA_URL = "/media/"
+MEDIA_ROOT = os.path.join(str(ROOT_DIR), "media")
+
+# Package storage settings
+
+ENABLE_PACKAGE_STORAGE = env.bool("ENABLE_PACKAGE_STORAGE", default=False)
+
 # Third-party apps
 
 CRISPY_TEMPLATE_PACK = "bootstrap3"

scancodeio/urls.py

@@ -21,6 +21,7 @@
 # Visit https://github.com/aboutcode-org/scancode.io for support and download.
 
 from django.conf import settings
+from django.conf.urls.static import static
 from django.contrib.auth import views as auth_views
 from django.urls import include
 from django.urls import path
@@ -54,6 +55,8 @@
     path("", RedirectView.as_view(url="project/")),
 ]
 
+urlpatterns += static(settings.MEDIA_URL, document_root=settings.MEDIA_ROOT)
+
 
 if settings.SCANCODEIO_ENABLE_ADMIN_SITE:
     urlpatterns.append(path("admin/", admin_site.urls))

scanpipe/forms.py

@@ -165,6 +165,7 @@ class Meta:
             "pipeline",
             "execute_now",
             "selected_groups",
+            "use_local_storage",
         ]
 
     def __init__(self, *args, **kwargs):
@@ -178,6 +179,11 @@ def __init__(self, *args, **kwargs):
         pipeline_choices = scanpipe_app.get_pipeline_choices(include_addon=False)
         self.fields["pipeline"].choices = pipeline_choices
 
+        self.fields["use_local_storage"].label = "Store packages locally"
+        self.fields["use_local_storage"].help_text = "If checked, " \
+        "packages will be stored on the local filesystem."
+        self.fields["use_local_storage"].widget.attrs.update({"class": "checkbox"})
+
     def clean_name(self):
         return " ".join(self.cleaned_data["name"].split())
 

scanpipe/migrations/0068_packagearchive_downloadedpackage.py

@@ -0,0 +1,44 @@
+# Generated by Django 5.1.1 on 2025-05-10 06:55
+
+import django.db.models.deletion
+import uuid
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ('scanpipe', '0067_discoveredpackage_notes'),
+    ]
+
+    operations = [
+        migrations.CreateModel(
+            name='PackageArchive',
+            fields=[
+                ('uuid', models.UUIDField(db_index=True, default=uuid.uuid4, editable=False, primary_key=True, serialize=False, verbose_name='UUID')),
+                ('checksum_sha256', models.CharField(db_index=True, help_text='SHA256 checksum of the package archive file.', max_length=64, unique=True)),
+                ('storage_path', models.CharField(help_text='Path to the stored archive file (e.g., file:///path/to/file).', max_length=1024)),
+                ('created_date', models.DateTimeField(auto_now_add=True, help_text='Date when the archive was added to storage.')),
+            ],
+            options={
+                'indexes': [models.Index(fields=['checksum_sha256'], name='checksum_idx')],
+            },
+        ),
+        migrations.CreateModel(
+            name='DownloadedPackage',
+            fields=[
+                ('uuid', models.UUIDField(db_index=True, default=uuid.uuid4, editable=False, primary_key=True, serialize=False, verbose_name='UUID')),
+                ('url', models.URLField(blank=True, db_index=True, help_text='URL from which the package was downloaded, if applicable.', max_length=1024)),
+                ('filename', models.CharField(help_text='Name of the package file.', max_length=255)),
+                ('download_date', models.DateTimeField(auto_now_add=True, help_text='Date when the package was downloaded or added.')),
+                ('scan_log', models.TextField(blank=True, help_text='Log output from scanning the package.')),
+                ('scan_date', models.DateTimeField(blank=True, help_text='Date when the package was scanned.', null=True)),
+                ('project', models.ForeignKey(editable=False, on_delete=django.db.models.deletion.CASCADE, related_name='downloadedpackages', to='scanpipe.project')),
+                ('package_archive', models.ForeignKey(help_text='The stored archive file associated with this package.', on_delete=django.db.models.deletion.CASCADE, to='scanpipe.packagearchive')),
+            ],
+            options={
+                'indexes': [models.Index(fields=['url'], name='url_idx')],
+                'constraints': [models.UniqueConstraint(condition=models.Q(('url__gt', '')), fields=('url', 'project'), name='scanpipe_downloadedpackage_unique_url_project')],
+            },
+        ),
+    ]

scanpipe/migrations/0069_packagearchive_package_file_and_more.py

@@ -0,0 +1,23 @@
+# Generated by Django 5.1.1 on 2025-05-12 09:41
+
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ('scanpipe', '0068_packagearchive_downloadedpackage'),
+    ]
+
+    operations = [
+        migrations.AddField(
+            model_name='packagearchive',
+            name='package_file',
+            field=models.FileField(blank=True, help_text='The actual package archive file (e.g., ZIP or TAR).', null=True, upload_to='packages/'),
+        ),
+        migrations.AlterField(
+            model_name='packagearchive',
+            name='storage_path',
+            field=models.CharField(blank=True, help_text='Path to the stored archive file (e.g., file:///path/to/file).', max_length=1024),
+        ),
+    ]

scanpipe/migrations/0070_project_use_local_storage_and_more.py

@@ -0,0 +1,28 @@
+# Generated by Django 5.1.1 on 2025-05-26 09:19
+
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ('scanpipe', '0069_packagearchive_package_file_and_more'),
+    ]
+
+    operations = [
+        migrations.AddField(
+            model_name='project',
+            name='use_local_storage',
+            field=models.BooleanField(default=False, help_text='Store packages locally if enabled.'),
+        ),
+        migrations.AlterField(
+            model_name='packagearchive',
+            name='package_file',
+            field=models.FileField(blank=True, help_text='The actual package archive file ( ZIP or TAR).', null=True, upload_to='packages/'),
+        ),
+        migrations.AlterField(
+            model_name='packagearchive',
+            name='storage_path',
+            field=models.CharField(blank=True, help_text='Path to the stored archive file', max_length=1024),
+        ),
+    ]

scanpipe/models.py

@@ -585,18 +585,9 @@ class Project(UUIDPKModel, ExtraDataFieldMixin, UpdateMixin, models.Model):
     )
     notes = models.TextField(blank=True)
     settings = models.JSONField(default=dict, blank=True)
-    labels = TaggableManager(through=UUIDTaggedItem, ordering=["name"])
-    purl = models.CharField(
-        max_length=2048,
-        blank=True,
-        help_text=_(
-            "Package URL (PURL) for the project, required for pushing the project's "
-            "scan result to FederatedCode. For example, if the input is an input URL "
-            "like https://registry.npmjs.org/lodash/-/lodash-4.17.21.tgz, the "
-            "corresponding PURL would be pkg:npm/[email protected]."
-        ),
-    )
-
+    labels = TaggableManager(through=UUIDTaggedItem)
+    use_local_storage = models.BooleanField(default=False,
+                         help_text="Store packages locally if enabled.")
     objects = ProjectQuerySet.as_manager()
 
     class Meta:
@@ -4386,6 +4377,102 @@ def success(self):
         return self.response_status_code in (200, 201, 202)
 
 
+class PackageArchive(UUIDPKModel):
+    """
+    Stores metadata about a package archive file stored in the project's storage.
+    Each archive is uniquely identified by its SHA256 checksum.
+    """
+
+    checksum_sha256 = models.CharField(
+        max_length=64,
+        unique=True,
+        db_index=True,
+        help_text=_("SHA256 checksum of the package archive file."),
+    )
+    storage_path = models.CharField(
+        max_length=1024,
+        blank=True,
+        help_text=_("Path to the stored archive file"),
+    )
+    package_file = models.FileField(
+        upload_to="packages/",
+        null=True,
+        blank=True,
+        help_text=_("The actual package archive file ( ZIP or TAR)."),
+    )
+    created_date = models.DateTimeField(
+        auto_now_add=True,
+        help_text=_("Date when the archive was added to storage."),
+    )
+
+    class Meta:
+        indexes = [
+            models.Index(fields=["checksum_sha256"], name="checksum_idx"),
+        ]
+
+    def __str__(self):
+        return f"Archive {self.checksum_sha256[:8]} at {self.storage_path
+        or self.package_file.name}"
+
+
+class DownloadedPackage(UUIDPKModel):
+    """
+    Tracks packages downloaded or provided as input for a project, linked to a
+    PackageArchive. Each instance represents a package associated with a project,
+    including its source URL (if downloaded) and scan details.
+    """
+
+    project = models.ForeignKey(
+        Project,
+        related_name="downloadedpackages",
+        on_delete=models.CASCADE,
+        editable=False,
+    )
+    url = models.URLField(
+        max_length=1024,
+        db_index=True,
+        blank=True,
+        help_text=_("URL from which the package was downloaded, if applicable."),
+    )
+    filename = models.CharField(
+        max_length=255,
+        help_text=_("Name of the package file."),
+    )
+    download_date = models.DateTimeField(
+        auto_now_add=True,
+        help_text=_("Date when the package was downloaded or added."),
+    )
+    scan_log = models.TextField(
+        blank=True,
+        help_text=_("Log output from scanning the package."),
+    )
+    scan_date = models.DateTimeField(
+        null=True,
+        blank=True,
+        help_text=_("Date when the package was scanned."),
+    )
+    package_archive = models.ForeignKey(
+        PackageArchive,
+        on_delete=models.CASCADE,
+        help_text=_("The stored archive file associated with this package."),
+    )
+
+    class Meta:
+        indexes = [
+            models.Index(fields=["url"], name="url_idx"),
+        ]
+        constraints = [
+            models.UniqueConstraint(
+                fields=["url", "project"],
+                condition=Q(url__gt=""),
+                name="%(app_label)s_%(class)s_unique_url_project",
+            ),
+        ]
+
+    def __str__(self):
+        return f"{self.filename} for project {self.project.name}"
+
+
 @receiver(models.signals.post_save, sender=settings.AUTH_USER_MODEL)
 def create_auth_token(sender, instance=None, created=False, **kwargs):
     """Create an API key token on user creation, using the signal system."""

scanpipe/pipelines/analyze_docker.py

@@ -20,9 +20,15 @@
 # ScanCode.io is a free software code scanning tool from nexB Inc. and others.
 # Visit https://github.com/aboutcode-org/scancode.io for support and download.
 
-from scanpipe.pipelines.analyze_root_filesystem import RootFS
+import logging
+from pathlib import Path
+
+from scanpipe.pipelines import RootFS
 from scanpipe.pipes import docker
 from scanpipe.pipes import rootfs
+from scanpipe.pipes.fetch import store_package_archive
+
+logger = logging.getLogger(__name__)
 
 
 class Docker(RootFS):
@@ -36,6 +42,7 @@ def steps(cls):
             cls.find_images_os_and_distro,
             cls.collect_images_information,
             cls.collect_and_create_codebase_resources,
+            cls.store_package_archives,
             cls.collect_and_create_system_packages,
             cls.flag_uninteresting_codebase_resources,
             cls.flag_empty_files,
@@ -74,6 +81,38 @@ def collect_and_create_codebase_resources(self):
         """Collect and labels all image files as CodebaseResources."""
         for image in self.images:
             docker.create_codebase_resources(self.project, image)
+        self.package_files = []
+        for resource in self.project.codebaseresources.filter(extension=".deb"):
+            self.package_files.append(resource.path)
+            logger.debug(f"Found package file: {resource.path}")
+
+    def store_package_archives(self):
+        """Store identified package archives."""
+        if not self.project.use_local_storage:
+           logger.info(f"Local storage is disabled for project: {self.project.name}."
+                        "Skipping package storage.")
+           return []
+
+        logger.info(
+           f"Storing package archives for project: {self.project.name},"
+             "files: {self.package_files}"
+        )
+        stored_files = []
+        for package_path in self.package_files:
+            if not Path(package_path).exists():
+               logger.error(f"Invalid or missing package path: {package_path}")
+               continue
+            package_path_str = str(package_path)
+            logger.info(f"Storing package archive: {package_path_str}")
+            try:
+                result = store_package_archive(
+                  self.project, url=None, file_path=package_path_str
+                )
+                logger.info(f"Stored package archive {package_path_str}: {result}")
+                stored_files.append(result)
+            except Exception as e:
+                   logger.error(f"Failed to store {package_path_str}: {e}")
+        return stored_files
 
     def collect_and_create_system_packages(self):
         """Collect installed system packages for each layer based on the distro."""