Add is_binary, is_text, and is_archive fields to the CodebaseResource model #75

CHANGELOG.rst

@@ -1,6 +1,12 @@
 // Release notes
 // -------------
 
+### unreleased
+
+- Add the is_binary, is_text, and is_archive fields to the CodebaseResource
+  model.
+  https://github.com/nexB/scancode.io/issues/75
+
 ### v21.5.12
 
 - Adds a new way to fetch docker images using skopeo provided as a

scanpipe/migrations/0007_resource_is_binary_is_text_is_archive.py

@@ -0,0 +1,28 @@
+# Generated by Django 3.2.1 on 2021-05-27 12:38
+
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ('scanpipe', '0006_codebaseresource_compliance_alert'),
+    ]
+
+    operations = [
+        migrations.AddField(
+            model_name='codebaseresource',
+            name='is_archive',
+            field=models.BooleanField(default=False),
+        ),
+        migrations.AddField(
+            model_name='codebaseresource',
+            name='is_binary',
+            field=models.BooleanField(default=False),
+        ),
+        migrations.AddField(
+            model_name='codebaseresource',
+            name='is_text',
+            field=models.BooleanField(default=False),
+        ),
+    ]

scanpipe/models.py

@@ -967,6 +967,9 @@ class Type(models.TextChoices):
         blank=True,
         help_text=_("Descriptive file type for this resource."),
     )
+    is_binary = models.BooleanField(default=False)
+    is_text = models.BooleanField(default=False)
+    is_archive = models.BooleanField(default=False)
 
     class Compliance(models.TextChoices):
         OK = "ok"

scanpipe/pipes/scancode.py

@@ -101,9 +101,6 @@ def get_resource_info(location):
 
     # Missing fields on CodebaseResource model returned by `get_file_info`.
     unsupported_fields = [
-        "is_binary",
-        "is_text",
-        "is_archive",
         "is_media",
         "is_source",
         "is_script",

scanpipe/tests/test_api.py

@@ -371,6 +371,6 @@ def test_scanpipe_api_serializer_get_model_serializer(self):
 
     def test_scanpipe_api_serializer_get_serializer_fields(self):
         self.assertEqual(28, len(get_serializer_fields(DiscoveredPackage)))
-        self.assertEqual(21, len(get_serializer_fields(CodebaseResource)))
+        self.assertEqual(24, len(get_serializer_fields(CodebaseResource)))
         with self.assertRaises(LookupError):
             get_serializer_fields(None)

scanpipe/tests/test_pipes.py

@@ -221,14 +221,16 @@ def test_scanpipe_pipes_scancode_get_resource_info(self):
             "type": CodebaseResource.Type.FILE,
             "name": "notice",
             "extension": ".NOTICE",
+            "is_text": True,
             "size": 1178,
             "sha1": "4bd631df28995c332bf69d9d4f0f74d7ee089598",
             "md5": "90cd416fd24df31f608249b77bae80f1",
             "sha256": sha256,
             "mime_type": "text/plain",
             "file_type": "ASCII text",
         }
-        self.assertEqual(expected, scancode.get_resource_info(input_location))
+        resource_info = scancode.get_resource_info(input_location)
+        self.assertEqual(expected, resource_info)
 
     def test_scanpipe_pipes_scancode_scan_file(self):
         input_location = str(self.data_location / "notice.NOTICE")