Fix IntegrityError on rerun of collect_and_create_codebase_resources

scanpipe/pipes/__init__.py

@@ -31,6 +31,7 @@
 from itertools import islice
 from pathlib import Path
 
+from django.db import IntegrityError
 from django.db.models import Count
 
 from scanpipe.models import AbstractTaskFieldsModel
@@ -139,12 +140,17 @@ def collect_and_create_codebase_resources(project, batch_size=5000):
     """
     model_class = CodebaseResource
     objs = yield_resources_from_codebase(project)
-
-    while True:
-        batch = list(islice(objs, batch_size))
-        if not batch:
-            break
-        model_class.objects.bulk_create(batch, batch_size)
+    try:
+        while True:
+            batch = list(islice(objs, batch_size))
+            if not batch:
+                break
+            model_class.objects.bulk_create(batch, batch_size)
+    except IntegrityError as e:
+        raise IntegrityError(
+            "Codebase resources already exist for this project. "
+            "Reset the project before re-running."
+        ) from e
 
 
 def update_or_create_resource(project, resource_data):

scanpipe/tests/pipes/test_pipes.py

@@ -25,6 +25,7 @@
 from pathlib import Path
 from unittest import mock
 
+from django.db import IntegrityError
 from django.test import TestCase
 from django.test import TransactionTestCase
 
@@ -448,3 +449,20 @@ def test_scanpipe_pipes_collect_and_create_codebase_resources(self):
         self.assertEqual("from", from_resource.tag)
         to_resource = p1.codebaseresources.get(path="to/a.txt")
         self.assertEqual("to", to_resource.tag)
+
+    def test_scanpipe_pipes_collect_and_create_codebase_resources_duplicate_run(self):
+        p1 = Project.objects.create(name="Analysis")
+        input_location = self.data / "codebase" / "a.txt"
+        to_dir = p1.codebase_path / "to"
+        to_dir.mkdir()
+        from_dir = p1.codebase_path / "from"
+        from_dir.mkdir()
+        copy_input(input_location, to_dir)
+        copy_input(input_location, from_dir)
+
+        pipes.collect_and_create_codebase_resources(p1)
+        self.assertEqual(4, p1.codebaseresources.count())
+
+        with self.assertRaises(IntegrityError) as ctx:
+            pipes.collect_and_create_codebase_resources(p1)
+        self.assertIn("Reset", str(ctx.exception))