From 67c955936286b2e51a96af654f998e49512c44c4 Mon Sep 17 00:00:00 2001
From: shamoon <4887959+shamoon@users.noreply.github.com>
Date: Wed, 5 Mar 2025 10:05:35 -0800
Subject: [PATCH] Feature: allow using archive version in merge

---
 src/documents/bulk_edit.py            | 10 ++++++--
 src/documents/serialisers.py          |  5 ++++
 src/documents/tests/test_bulk_edit.py | 37 +++++++++++++++++++++++++++
 3 files changed, 50 insertions(+), 2 deletions(-)

diff --git a/src/documents/bulk_edit.py b/src/documents/bulk_edit.py
index be4608e36..54f5c1a33 100644
--- a/src/documents/bulk_edit.py
+++ b/src/documents/bulk_edit.py
@@ -318,6 +318,7 @@ def merge(
     *,
     metadata_document_id: int | None = None,
     delete_originals: bool = False,
+    archive_fallback: bool = False,
     user: User | None = None,
 ) -> Literal["OK"]:
     logger.info(
@@ -333,7 +334,12 @@ def merge(
     for doc_id in doc_ids:
         doc = qs.get(id=doc_id)
         try:
-            with pikepdf.open(str(doc.source_path)) as pdf:
+            doc_path = (
+                doc.archive_path
+                if archive_fallback and doc.mime_type != "application/pdf"
+                else doc.source_path
+            )
+            with pikepdf.open(str(doc_path)) as pdf:
                 version = max(version, pdf.pdf_version)
                 merged_pdf.pages.extend(pdf.pages)
             affected_docs.append(doc.id)
@@ -349,7 +355,7 @@ def merge(
         Path(
             tempfile.mkdtemp(dir=settings.SCRATCH_DIR),
         )
-        / f"{'_'.join([str(doc_id) for doc_id in doc_ids])[:100]}_merged.pdf"
+        / f"{'_'.join([str(doc_id) for doc_id in affected_docs])[:100]}_merged.pdf"
     )
     merged_pdf.remove_unreferenced_resources()
     merged_pdf.save(filepath, min_version=version)
diff --git a/src/documents/serialisers.py b/src/documents/serialisers.py
index c0487b7b8..50cd23918 100644
--- a/src/documents/serialisers.py
+++ b/src/documents/serialisers.py
@@ -1446,6 +1446,11 @@ class BulkEditSerializer(
                 raise serializers.ValidationError("delete_originals must be a boolean")
         else:
             parameters["delete_originals"] = False
+        if "archive_fallback" in parameters:
+            if not isinstance(parameters["archive_fallback"], bool):
+                raise serializers.ValidationError("archive_fallback must be a boolean")
+        else:
+            parameters["archive_fallback"] = False
 
     def validate(self, attrs):
         method = attrs["method"]
diff --git a/src/documents/tests/test_bulk_edit.py b/src/documents/tests/test_bulk_edit.py
index 4a7145d34..dd59a6217 100644
--- a/src/documents/tests/test_bulk_edit.py
+++ b/src/documents/tests/test_bulk_edit.py
@@ -514,12 +514,23 @@ class TestPDFActions(DirectoriesMixin, TestCase):
             Path(__file__).parent / "samples" / "simple.jpg",
             img_doc,
         )
+        img_doc_archive = self.dirs.archive_dir / "sample_image.pdf"
+        shutil.copy(
+            Path(__file__).parent
+            / "samples"
+            / "documents"
+            / "originals"
+            / "0000001.pdf",
+            img_doc_archive,
+        )
         self.img_doc = Document.objects.create(
             checksum="D",
             title="D",
             filename=img_doc,
             mime_type="image/jpeg",
         )
+        self.img_doc.archive_filename = img_doc_archive
+        self.img_doc.save()
 
     @mock.patch("documents.tasks.consume_file.s")
     def test_merge(self, mock_consume_file):
@@ -605,6 +616,32 @@ class TestPDFActions(DirectoriesMixin, TestCase):
             doc_ids,
         )
 
+    @mock.patch("documents.tasks.consume_file.s")
+    def test_merge_with_archive_fallback(self, mock_consume_file):
+        """
+        GIVEN:
+            - Existing documents
+        WHEN:
+            - Merge action is called with 2 documents, one of which is an image and archive_fallback is set to True
+        THEN:
+            - Image document should be included
+        """
+        doc_ids = [self.doc2.id, self.img_doc.id]
+
+        result = bulk_edit.merge(doc_ids, archive_fallback=True)
+        self.assertEqual(result, "OK")
+
+        expected_filename = (
+            f"{'_'.join([str(doc_id) for doc_id in doc_ids])[:100]}_merged.pdf"
+        )
+
+        mock_consume_file.assert_called()
+        consume_file_args, _ = mock_consume_file.call_args
+        self.assertEqual(
+            Path(consume_file_args[0].original_file).name,
+            expected_filename,
+        )
+
     @mock.patch("documents.tasks.consume_file.delay")
     @mock.patch("pikepdf.open")
     def test_merge_with_errors(self, mock_open_pdf, mock_consume_file):