API now supports setting metadata when POSTing documents.

docs/api.rst

@@ -216,9 +216,7 @@ be instructed to consume the document from there.
 The endpoint supports the following optional form fields:
 
 *   ``title``: Specify a title that the consumer should use for the document.
-*   ``correspondent``: Specify a correspondent that the consumer should use for the document.
-    Case sensitive. If the specified correspondent does not exist, it will be created with this
-    name and default settings.
+*   ``correspondent``: Specify the ID of a correspondent that the consumer should use for the document.
 *   ``document_type``: Similar to correspondent.
 *   ``tags``: Similar to correspondent. Specify this multiple times to have multiple tags added
     to the document.

docs/changelog.rst

@@ -8,6 +8,9 @@ Changelog
 paperless-ng 0.9.5
 ##################
 
+Apart from the API, this finalizes the changes I wanted to get into paperless before 1.0. The next releases will
+focus on fixing bugs, minor changes to the UI, and possibly some changes to the API.
+
 * OCR
 
   * Paperless now uses `OCRmyPDF <https://github.com/jbarlow83/OCRmyPDF>`_ to perform OCR on documents.
@@ -33,10 +36,15 @@ paperless-ng 0.9.5
   * The endpoint for uploading documents now supports specifying custom titles, correspondents, tags and types.
     This can be used by clients to override the default behavior of paperless.
   * The document endpoint of API now serves document in this form:
-    * correspondents, document types and tags are referenced by their ID in the fields ``correspondent``, ``document_type`` and ``tags``. The ``*_id`` versions are gone. These fields are read/write.
-    * in addition to that, ``*_object`` fields serve nested objects. Read only. Don't rely on these, they will probably get removed once I figure out how to better handle asynchronous data in the front end.
 
-* Some minor improvements to the front end, such as document count in the document list, better visibility of the current view, and improvements to the filter behavior.
+    * correspondents, document types and tags are referenced by their ID in the fields ``correspondent``, ``document_type`` and ``tags``. The ``*_id`` versions are gone. These fields are read/write.
+    * paperless does not serve nested tags, correspondents or types anymore.
+
+* Front end
+
+  * Paperless does some basic caching of correspondents, tags and types and will only request them from the server when necessary or when entirely reloading the page.
+  * Document lists should be somewhat faster now, especially when lots of tags/correspondents where present.
+  * Some minor improvements to the front end, such as document count in the document list, better highlighting of the current page, and improvements to the filter behavior.
 
 * Fixes:
 

src/documents/serialisers.py

@@ -126,22 +126,26 @@ class PostDocumentSerializer(serializers.Serializer):
         required=False,
     )
 
-    correspondent = serializers.CharField(
+    correspondent = serializers.PrimaryKeyRelatedField(
+        queryset=Correspondent.objects.all(),
         label="Correspondent",
+        allow_null=True,
         write_only=True,
         required=False,
     )
 
-    document_type = serializers.CharField(
+    document_type = serializers.PrimaryKeyRelatedField(
+        queryset=DocumentType.objects.all(),
         label="Document type",
+        allow_null=True,
         write_only=True,
         required=False,
     )
 
-    tags = serializers.ListField(
-        child=serializers.CharField(),
+    tags = serializers.PrimaryKeyRelatedField(
+        many=True,
+        queryset=Tag.objects.all(),
         label="Tags",
-        source="tag",
         write_only=True,
         required=False,
     )
@@ -170,24 +174,19 @@ class PostDocumentSerializer(serializers.Serializer):
 
         correspondent = attrs.get('correspondent')
         if correspondent:
-            c, _ = Correspondent.objects.get_or_create(name=correspondent)
-            attrs['correspondent_id'] = c.id
+            attrs['correspondent_id'] = correspondent.id
         else:
             attrs['correspondent_id'] = None
 
         document_type = attrs.get('document_type')
         if document_type:
-            dt, _ = DocumentType.objects.get_or_create(name=document_type)
-            attrs['document_type_id'] = dt.id
+            attrs['document_type_id'] = document_type.id
         else:
             attrs['document_type_id'] = None
 
-        tags = attrs.get('tag')
+        tags = attrs.get('tags')
         if tags:
-            tag_ids = []
-            for tag in tags:
-                tag, _ = Tag.objects.get_or_create(name=tag)
-                tag_ids.append(tag.id)
+            tag_ids = [tag.id for tag in tags]
             attrs['tag_ids'] = tag_ids
         else:
             attrs['tag_ids'] = None

src/documents/tests/test_api.py

@@ -410,7 +410,7 @@ class TestDocumentApi(DirectoriesMixin, APITestCase):
     def test_upload_with_correspondent(self, async_task):
         c = Correspondent.objects.create(name="test-corres")
         with open(os.path.join(os.path.dirname(__file__), "samples", "simple.pdf"), "rb") as f:
-            response = self.client.post("/api/documents/post_document/", {"document": f, "correspondent": "test-corres"})
+            response = self.client.post("/api/documents/post_document/", {"document": f, "correspondent": c.id})
         self.assertEqual(response.status_code, 200)
 
         async_task.assert_called_once()
@@ -420,23 +420,18 @@ class TestDocumentApi(DirectoriesMixin, APITestCase):
         self.assertEqual(kwargs['override_correspondent_id'], c.id)
 
     @mock.patch("documents.views.async_task")
-    def test_upload_with_new_correspondent(self, async_task):
+    def test_upload_with_invalid_correspondent(self, async_task):
         with open(os.path.join(os.path.dirname(__file__), "samples", "simple.pdf"), "rb") as f:
-            response = self.client.post("/api/documents/post_document/", {"document": f, "correspondent": "test-corres2"})
-        self.assertEqual(response.status_code, 200)
+            response = self.client.post("/api/documents/post_document/", {"document": f, "correspondent": 3456})
+        self.assertEqual(response.status_code, 400)
 
-        async_task.assert_called_once()
-
-        args, kwargs = async_task.call_args
-
-        c = Correspondent.objects.get(name="test-corres2")
-        self.assertEqual(kwargs['override_correspondent_id'], c.id)
+        async_task.assert_not_called()
 
     @mock.patch("documents.views.async_task")
     def test_upload_with_document_type(self, async_task):
         dt = DocumentType.objects.create(name="invoice")
         with open(os.path.join(os.path.dirname(__file__), "samples", "simple.pdf"), "rb") as f:
-            response = self.client.post("/api/documents/post_document/", {"document": f, "document_type": "invoice"})
+            response = self.client.post("/api/documents/post_document/", {"document": f, "document_type": dt.id})
         self.assertEqual(response.status_code, 200)
 
         async_task.assert_called_once()
@@ -446,30 +441,37 @@ class TestDocumentApi(DirectoriesMixin, APITestCase):
         self.assertEqual(kwargs['override_document_type_id'], dt.id)
 
     @mock.patch("documents.views.async_task")
-    def test_upload_with_new_document_type(self, async_task):
+    def test_upload_with_invalid_document_type(self, async_task):
         with open(os.path.join(os.path.dirname(__file__), "samples", "simple.pdf"), "rb") as f:
-            response = self.client.post("/api/documents/post_document/", {"document": f, "document_type": "invoice2"})
-        self.assertEqual(response.status_code, 200)
+            response = self.client.post("/api/documents/post_document/", {"document": f, "document_type": 34578})
+        self.assertEqual(response.status_code, 400)
 
-        async_task.assert_called_once()
-
-        args, kwargs = async_task.call_args
-
-        dt = DocumentType.objects.get(name="invoice2")
-        self.assertEqual(kwargs['override_document_type_id'], dt.id)
+        async_task.assert_not_called()
 
     @mock.patch("documents.views.async_task")
     def test_upload_with_tags(self, async_task):
         t1 = Tag.objects.create(name="tag1")
+        t2 = Tag.objects.create(name="tag2")
         with open(os.path.join(os.path.dirname(__file__), "samples", "simple.pdf"), "rb") as f:
             response = self.client.post(
                 "/api/documents/post_document/",
-                {"document": f, "tags": ["tag1", "tag2"]})
+                {"document": f, "tags": [t2.id, t1.id]})
         self.assertEqual(response.status_code, 200)
 
         async_task.assert_called_once()
 
         args, kwargs = async_task.call_args
 
-        t2 = Tag.objects.get(name="tag2")
         self.assertCountEqual(kwargs['override_tag_ids'], [t1.id, t2.id])
+
+    @mock.patch("documents.views.async_task")
+    def test_upload_with_invalid_tags(self, async_task):
+        t1 = Tag.objects.create(name="tag1")
+        t2 = Tag.objects.create(name="tag2")
+        with open(os.path.join(os.path.dirname(__file__), "samples", "simple.pdf"), "rb") as f:
+            response = self.client.post(
+                "/api/documents/post_document/",
+                {"document": f, "tags": [t2.id, t1.id, 734563]})
+        self.assertEqual(response.status_code, 400)
+
+        async_task.assert_not_called()