added support to parse labels in dockerfile

src/packagedcode/__init__.py

@@ -18,6 +18,7 @@
 from packagedcode import debian
 from packagedcode import debian_copyright
 from packagedcode import distro
+from packagedcode import dockerfile
 from packagedcode import conda
 from packagedcode import conan
 from packagedcode import cocoapods
@@ -95,6 +96,7 @@
     debian.DebianSourcePackageTarballHandler,
 
     distro.EtcOsReleaseHandler,
+    dockerfile.DockerfileHandler,
 
     freebsd.CompactManifestHandler,
 

src/packagedcode/dockerfile_ocilabels.py

@@ -0,0 +1,52 @@
+#
+# Copyright (c) nexB Inc. and others. All rights reserved.
+# ScanCode 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/scancode-toolkit for support or download.
+# See https://aboutcode.org for more information about nexB OSS projects.
+#
+
+
+
+import io
+from pathlib import Path
+from dockerfile_parse import DockerfileParser
+from packagedcode import models
+from packagedcode import utils
+from packagedcode.models import NonAssemblableDatafileHandler
+import fnmatch
+
+
+class DockerOCILabelsHandler(NonAssemblableDatafileHandler):
+    datasource_id = 'dockerfile_oci_labels'
+    path_patterns = ('Dockerfile', 'containerfile', '*.dockerfile')
+
+    @classmethod
+    def parse(cls, location, package_only=False):
+        """
+        Parse a Dockerfile and yield one or more PackageData objects with OCI labels and metadata.
+        """
+        labels = cls.extract_oci_labels_from_dockerfile(location)
+        package_data = {
+            'datasource_id': cls.datasource_id,
+            'type': cls.default_package_type,
+            'name': labels.get('name', 'None'),
+            'version': labels.get('version', 'None'),
+            'license_expression': labels.get('license'),
+            'labels': labels,
+        }
+
+        yield models.PackageData.from_data(package_data, package_only)
+
+    @classmethod
+    def extract_oci_labels_from_dockerfile(cls, dockerfile_path):
+        """
+        Extract OCI labels from the Dockerfile using DockerfileParser.
+        """
+        labels = {}
+        parser = DockerfileParser()
+        with open(dockerfile_path, 'r') as dockerfile:
+          parser.content = dockerfile.read() 
+          labels = parser.labels
+        return labels

tests/packagedcode/data/docker/test-containerfile/test.containerfile

@@ -0,0 +1,11 @@
+#Copied from https://github.com/kubernetes-sigs/blixt/blob
+
+FROM alpine
+
+WORKDIR /opt/blixt/
+
+LABEL org.opencontainers.image.source=https://github.com/kubernetes-sigs/blixt
+LABEL org.opencontainers.image.licenses=GPL-2.0-only,BSD-2-Clause
+
+COPY dataplane/LICENSE.GPL-2.0 /opt/blixt/LICENSE.GPL-2.0
+COPY dataplane/LICENSE.BSD-2-Clause /opt/blixt/LICENSE.BSD-2-Clause

tests/packagedcode/data/docker/test-containerfile/test.containerfile-expected.json

@@ -0,0 +1,13 @@
+[
+    {
+        "datasource_id": "dockerfile_oci_labels",
+        "type": "default",  
+        "name": "Unknown",  
+        "version": "Unknown",  
+        "license_expression": "GPL-2.0-only AND BSD-2-Clause",
+        "labels": {
+            "source": "https://github.com/kubernetes-sigs/blixt",
+            "licenses": "GPL-2.0-only,BSD-2-Clause"
+        }
+    }
+]

tests/packagedcode/data/docker/test-containerfile/test.containerfile-package.expected.json

@@ -0,0 +1,9 @@
+[
+  {
+    "datasource_id": "docker_oci_labels",
+    "labels": {
+      "org.opencontainers.image.source": "https://github.com/kubernetes-sigs/blixt",
+      "org.opencontainers.image.licenses": "GPL-2.0-only,BSD-2-Clause"
+    }
+  }
+]

tests/packagedcode/data/docker/test-containerfile/test.containerfile-scan.expected.json

@@ -0,0 +1,7 @@
+{
+    "datasource_id": "docker_oci_labels",
+    "labels": {
+        "org.opencontainers.image.source": "https://github.com/kubernetes-sigs/blixt",
+        "org.opencontainers.image.licenses": "GPL-2.0-only,BSD-2-Clause"
+    }
+}

tests/packagedcode/data/docker/test-dockerfile/test.dockerfile

@@ -0,0 +1,21 @@
+#Copied from https://github.com/kanboard/kanboard
+
+FROM alpine:3.21
+LABEL org.opencontainers.image.source=https://github.com/kanboard/kanboard
+LABEL org.opencontainers.image.title=Kanboard
+LABEL org.opencontainers.image.description="Kanboard is project management software that focuses on the Kanban methodology"
+LABEL org.opencontainers.image.vendor=Kanboard
+LABEL org.opencontainers.image.licenses=MIT
+LABEL org.opencontainers.image.url=https://kanboard.org
+LABEL org.opencontainers.image.documentation=https://docs.kanboard.org
+VOLUME /var/www/app/data
+VOLUME /var/www/app/plugins
+VOLUME /etc/nginx/ssl
+EXPOSE 80 443
+ARG VERSION
+RUN apk --no-cache --update add ...
+ADD . /var/www/app
+ADD docker/ /
+RUN rm -rf /var/www/app/docker && echo $VERSION > /var/www/app/app/version.txt
+ENTRYPOINT ["/usr/local/bin/entrypoint.sh"]
+CMD []

tests/packagedcode/data/docker/test-dockerfile/test.dockerfile-expected.json

@@ -0,0 +1,9 @@
+{
+  "org.opencontainers.image.source": "https://github.com/kanboard/kanboard",
+  "org.opencontainers.image.title": "Kanboard",
+  "org.opencontainers.image.description": "Kanboard is project management software that focuses on the Kanban methodology",
+  "org.opencontainers.image.vendor": "Kanboard",
+  "org.opencontainers.image.licenses": "MIT",
+  "org.opencontainers.image.url": "https://kanboard.org",
+  "org.opencontainers.image.documentation": "https://docs.kanboard.org"
+}

tests/packagedcode/data/docker/test-dockerfile/test.dockerfile-package.expected.json

@@ -0,0 +1,18 @@
+[
+  {
+    "datasource_id": "docker_oci_labels",
+    "type": "default",
+    "name": "Kanboard",
+    "version": "1.2.42",
+    "license_expression": "MIT",
+    "labels": {
+      "org.opencontainers.image.source": "https://github.com/kanboard/kanboard",
+      "org.opencontainers.image.title": "Kanboard",
+      "org.opencontainers.image.description": "Kanboard is project management software that focuses on the Kanban methodology",
+      "org.opencontainers.image.vendor": "Kanboard",
+      "org.opencontainers.image.licenses": "MIT",
+      "org.opencontainers.image.url": "https://kanboard.org",
+      "org.opencontainers.image.documentation": "https://docs.kanboard.org"
+    }
+  }
+]

tests/packagedcode/data/docker/test-dockerfile/test.dockerfile-scan.expected.json

@@ -0,0 +1,18 @@
+[
+  {
+    "datasource_id": "docker_oci_labels",
+    "type": "default",
+    "name": "Kanboard",
+    "version": "1.2.42",
+    "license_expression": "MIT",
+    "labels": {
+      "org.opencontainers.image.source": "https://github.com/kanboard/kanboard",
+      "org.opencontainers.image.title": "Kanboard",
+      "org.opencontainers.image.description": "Kanboard is project management software that focuses on the Kanban methodology",
+      "org.opencontainers.image.vendor": "Kanboard",
+      "org.opencontainers.image.licenses": "MIT",
+      "org.opencontainers.image.url": "https://kanboard.org",
+      "org.opencontainers.image.documentation": "https://docs.kanboard.org"
+    }
+  }
+]

tests/packagedcode/test_dockerfile_ocilabels.py

@@ -0,0 +1,66 @@
+#
+# Copyright (c) nexB Inc. and others. All rights reserved.
+# ScanCode 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/scancode-toolkit for support or download.
+# See https://aboutcode.org for more information about nexB OSS projects.
+#
+
+import pytest
+import json
+from commoncode.testcase import FileDrivenTesting
+from scancode.cli_test_utils import run_scan_click
+from packagedcode.models import DockerOCILabelsHandler
+
+class TestDockerOCILabelsHandler(FileDrivenTesting):
+
+    @pytest.mark.parametrize('test_file, expected', [
+        ('docker/test.dockerfile', True),
+        ('docker/test.containerfile', True),
+    ])
+    def test_is_datafile(self, test_file, expected):
+        test_file_path = self.get_test_loc(test_file)
+        assert DockerOCILabelsHandler.is_datafile(test_file_path) == expected
+
+    def test_parse_dockerfile(self):
+        test_files = [
+            ('test.dockerfile', 'test.dockerfile-package.expected.json'),
+            ('test.containerfile', 'test.containerfile-package.expected.json'),
+        ]
+        for dockerfile, expected in test_files:
+            test_file = self.get_test_loc(f'docker/{dockerfile}')
+            expected_loc = self.get_test_loc(f'docker/{expected}')
+            packages = list(DockerOCILabelsHandler.parse(test_file))
+            expected_packages = self.load_expected(expected_loc)
+            assert packages == expected_packages
+
+    def test_extract_oci_labels_from_dockerfile(self):
+        test_files = [
+            ('test.dockerfile', 'test.dockerfile-expected.json'),
+            ('test.containerfile', 'test.containerfile-expected.json'),
+        ]
+        for dockerfile, expected in test_files:
+            dockerfile_path = self.get_test_loc(f'docker/{dockerfile}')
+            labels = DockerOCILabelsHandler.extract_oci_labels_from_dockerfile(dockerfile_path)
+            expected_loc = self.get_test_loc(f'docker/{expected}')
+            expected_labels = self.load_expected(expected_loc)
+            assert labels == expected_labels
+
+    def test_full_scan_docker_oci_labels_containerfile(self):
+        test_file = self.get_test_loc('docker/test.containerfile')
+        result_file = self.get_temp_file('json')
+        run_scan_click(['--package', test_file, '--json-pp', result_file])
+        result = json.load(open(result_file))
+        package_data = result.get('package_data', [])
+        assert len(package_data) == 1
+        package = package_data[0]
+        assert package['datasource_id'] == 'docker_oci_labels'
+        assert package['labels'] == {
+            'org.opencontainers.image.source': 'https://github.com/kubernetes-sigs/blixt',
+            'org.opencontainers.image.licenses': 'GPL-2.0-only,BSD-2-Clause'
+        }
+
+    def load_expected(self, expected_file):
+        with open(expected_file) as f:
+            return json.load(f)