Feature: Document fuzzy match improvements (#12579)

Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>

src/documents/tests/test_management_fuzzy.py

@@ -1,4 +1,5 @@
 from io import StringIO
+from unittest.mock import patch
 
 import pytest
 from django.core.management import CommandError
@@ -6,12 +7,11 @@ from django.core.management import call_command
 from django.test import TestCase
 
 from documents.models import Document
+from documents.tests.factories import DocumentFactory
 
 
 @pytest.mark.management
 class TestFuzzyMatchCommand(TestCase):
-    MSG_REGEX = r"Document \d fuzzy match to \d \(confidence \d\d\.\d\d\d\)"
-
     def call_command(self, *args, **kwargs):
         stdout = StringIO()
         stderr = StringIO()
@@ -77,7 +77,7 @@ class TestFuzzyMatchCommand(TestCase):
             filename="other_test.pdf",
         )
         stdout, _ = self.call_command()
-        self.assertIn("No matches found", stdout)
+        self.assertIn("No duplicate documents found", stdout)
 
     def test_with_matches(self) -> None:
         """
@@ -106,7 +106,7 @@ class TestFuzzyMatchCommand(TestCase):
             filename="other_test.pdf",
         )
         stdout, _ = self.call_command("--processes", "1")
-        self.assertRegex(stdout, self.MSG_REGEX)
+        self.assertIn("Found 1 matching pair(s)", stdout)
 
     def test_with_3_matches(self) -> None:
         """
@@ -142,10 +142,8 @@ class TestFuzzyMatchCommand(TestCase):
             filename="final_test.pdf",
         )
         stdout, _ = self.call_command("--no-progress-bar", "--processes", "1")
-        lines = [x.strip() for x in stdout.splitlines() if x.strip()]
-        self.assertEqual(len(lines), 3)
-        for line in lines:
-            self.assertRegex(line, self.MSG_REGEX)
+        # 3 docs -> 3 unique pairs; summary confirms count and no duplication
+        self.assertIn("Found 3 matching pair(s)", stdout)
 
     def test_document_deletion(self) -> None:
         """
@@ -186,22 +184,47 @@ class TestFuzzyMatchCommand(TestCase):
 
         stdout, _ = self.call_command(
             "--delete",
+            "--yes",
             "--no-progress-bar",
             "--processes",
             "1",
         )
 
-        self.assertIn(
-            "The command is configured to delete documents.  Use with caution",
-            stdout,
-        )
-        self.assertRegex(stdout, self.MSG_REGEX)
-        self.assertIn("Deleting 1 documents based on ratio matches", stdout)
+        self.assertIn("Delete Mode", stdout)
+        self.assertIn("Found 1 matching pair(s)", stdout)
+        self.assertIn("Deleting 1 document(s)", stdout)
 
         self.assertEqual(Document.objects.count(), 2)
         self.assertIsNotNone(Document.objects.get(pk=1))
         self.assertIsNotNone(Document.objects.get(pk=2))
 
+    def test_document_deletion_cancelled(self) -> None:
+        """
+        GIVEN:
+            - 3 documents exist
+            - Document 1 to document 3 has a similarity over 85.0
+        WHEN:
+            - Command is called with --delete but user answers "n" at the prompt
+        THEN:
+            - No documents are deleted
+        """
+        DocumentFactory(content="first document scanned by bob")
+        DocumentFactory(content="second document scanned by alice")
+        DocumentFactory(content="first document scanned by pete")
+
+        self.assertEqual(Document.objects.count(), 3)
+
+        with patch("builtins.input", return_value="n"):
+            stdout, _ = self.call_command(
+                "--delete",
+                "--no-progress-bar",
+                "--processes",
+                "1",
+            )
+
+        self.assertIn("Deletion cancelled", stdout)
+        self.assertEqual(Document.objects.count(), 3)
+
     def test_empty_content(self) -> None:
         """
         GIVEN:
@@ -226,4 +249,30 @@ class TestFuzzyMatchCommand(TestCase):
             filename="other_test.pdf",
         )
         stdout, _ = self.call_command()
-        self.assertIn("No matches found", stdout)
+        self.assertIn("No duplicate documents found", stdout)
+
+
+@pytest.mark.management
+@pytest.mark.django_db
+@pytest.mark.parametrize(
+    ("content_a", "content_b"),
+    [
+        pytest.param("x" * 90 + "y" * 10, "x" * 100, id="yellow-90pct"),  # 88-92%
+        pytest.param("x" * 94 + "y" * 6, "x" * 100, id="red-94pct"),  # 92-97%
+        pytest.param("x" * 99 + "y", "x" * 100, id="bold-red-99pct"),  # ≥97%
+    ],
+)
+def test_similarity_color_band(content_a: str, content_b: str) -> None:
+    """Each parametrized case exercises one color branch in _render_results."""
+    DocumentFactory(content=content_a)
+    DocumentFactory(content=content_b)
+    stdout = StringIO()
+    call_command(
+        "document_fuzzy_match",
+        "--no-progress-bar",
+        "--processes",
+        "1",
+        stdout=stdout,
+        skip_checks=True,
+    )
+    assert "Found 1 matching pair(s)" in stdout.getvalue()