Merge

src/documents/bulk_edit.py

@@ -2,15 +2,20 @@ import hashlib
 import itertools
 import logging
 import os
+from typing import Optional
 
 from django.db.models import Q
 
+from documents.data_models import ConsumableDocument
+from documents.data_models import DocumentMetadataOverrides
+from documents.data_models import DocumentSource
 from documents.models import Correspondent
 from documents.models import Document
 from documents.models import DocumentType
 from documents.models import StoragePath
 from documents.permissions import set_permissions_for_object
 from documents.tasks import bulk_update_documents
+from documents.tasks import consume_file
 from documents.tasks import update_document_archive_file
 from paperless import settings
 
@@ -179,3 +184,67 @@ def rotate(doc_ids: list[int], degrees: int):
         bulk_update_documents.delay(document_ids=affected_docs)
 
     return "OK"
+
+
+def merge(doc_ids: list[int], metadata_document_id: Optional[int] = None):
+    qs = Document.objects.filter(id__in=doc_ids)
+    import pikepdf
+
+    merged_pdf = pikepdf.new()
+    # use doc_ids to preserve order
+    for doc_id in doc_ids:
+        doc = qs.get(id=doc_id)
+        if doc is None:
+            continue
+        path = os.path.join(settings.ORIGINALS_DIR, str(doc.filename))
+        try:
+            with pikepdf.open(path, allow_overwriting_input=True) as pdf:
+                merged_pdf.pages.extend(pdf.pages)
+        except Exception as e:
+            logger.exception(
+                f"Error merging document {doc.id}, it will not be included in the merge",
+                e,
+            )
+
+    filepath = os.path.join(
+        settings.CONSUMPTION_DIR,
+        f"merged_{('_'.join([str(doc_id) for doc_id in doc_ids]))[:100]}.pdf",
+    )
+    merged_pdf.save(filepath)
+
+    overrides = DocumentMetadataOverrides()
+
+    if metadata_document_id:
+        metadata_document = qs.get(id=metadata_document_id)
+        if metadata_document is not None:
+            overrides.title = metadata_document.title + " (merged)"
+            overrides.correspondent_id = (
+                metadata_document.correspondent.pk
+                if metadata_document.correspondent
+                else None
+            )
+            overrides.document_type_id = (
+                metadata_document.document_type.pk
+                if metadata_document.document_type
+                else None
+            )
+            overrides.storage_path_id = (
+                metadata_document.storage_path.pk
+                if metadata_document.storage_path
+                else None
+            )
+            overrides.tag_ids = list(
+                metadata_document.tags.values_list("id", flat=True),
+            )
+            # Include owner and permissions?
+
+    logger.info("Adding merged document to the task queue.")
+    consume_file.delay(
+        ConsumableDocument(
+            source=DocumentSource.ConsumeFolder,
+            original_file=filepath,
+        ),
+        overrides,
+    )
+
+    return "OK"

src/documents/serialisers.py

@@ -870,6 +870,7 @@ class BulkEditSerializer(DocumentListSerializer, SetPermissionsMixin):
             "redo_ocr",
             "set_permissions",
             "rotate",
+            "merge",
         ],
         label="Method",
         write_only=True,
@@ -909,6 +910,8 @@ class BulkEditSerializer(DocumentListSerializer, SetPermissionsMixin):
             return bulk_edit.set_permissions
         elif method == "rotate":
             return bulk_edit.rotate
+        elif method == "merge":
+            return bulk_edit.merge
         else:
             raise serializers.ValidationError("Unsupported method.")
 

src/documents/tests/test_api_bulk_edit.py

@@ -836,3 +836,26 @@ class TestBulkEditAPI(DirectoriesMixin, APITestCase):
         self.assertEqual(response.status_code, status.HTTP_400_BAD_REQUEST)
 
         m.assert_not_called()
+
+    @mock.patch("documents.serialisers.bulk_edit.merge")
+    def test_merge(self, m):
+        m.return_value = "OK"
+
+        response = self.client.post(
+            "/api/documents/bulk_edit/",
+            json.dumps(
+                {
+                    "documents": [self.doc2.id, self.doc3.id],
+                    "method": "merge",
+                    "parameters": {"metadata_document_id": self.doc3.id},
+                },
+            ),
+            content_type="application/json",
+        )
+
+        self.assertEqual(response.status_code, status.HTTP_200_OK)
+
+        m.assert_called_once()
+        args, kwargs = m.call_args
+        self.assertCountEqual(args[0], [self.doc2.id, self.doc3.id])
+        self.assertEqual(kwargs["metadata_document_id"], self.doc3.id)