Test migration change (#1784)

* added tests to ensure migrations do not change w/ burnettk

* fixed tests w/ burnettk

* pyl w/ burnettk

---------

Co-authored-by: jasquat <jasquat@users.noreply.github.com>

spiffworkflow-backend/src/spiffworkflow_backend/constants.py

@@ -1 +1,10 @@
 SPIFFWORKFLOW_BACKEND_SERIALIZER_VERSION = "5"
+
+# so we can tell if a migration has changed or if there is a new one
+SPIFFWORKFLOW_BACKEND_DATA_MIGRATION_CHECKSUM = {
+    "version_1_3.py": "22636f5ffb8e6d56fa460d82f62a854c",
+    "version_2.py": "962b4fda4d466758bdbdc09d75099603",
+    "version_3.py": "0e7154d0575c54b59011e3acedebe8b5",
+    "version_4.py": "889399d1c37e230a669d099f3a485fd4",
+    "version_5.py": "a5a9e62798b51741d4dd9d8f69868ded",
+}

spiffworkflow-backend/src/spiffworkflow_backend/data_migrations/process_instance_migrator.py

@@ -1,3 +1,6 @@
+import glob
+import hashlib
+import os
 import time
 from typing import Any
 
@@ -11,6 +14,10 @@ from spiffworkflow_backend.models.db import db
 from spiffworkflow_backend.models.process_instance import ProcessInstanceModel
 
 
+class DataMigrationFilesNotFoundError(Exception):
+    pass
+
+
 # simple decorator to time the func
 # https://stackoverflow.com/a/11151365/6090676, thank you
 def benchmark_log_func(func: Any) -> Any:
@@ -70,3 +77,38 @@ class ProcessInstanceMigrator:
             process_instance.spiff_serializer_version = data_migration_version_class.version()
             db.session.add(process_instance)
             db.session.commit()
+
+    @classmethod
+    def get_migration_files(cls) -> list[str]:
+        file_glob = os.path.join(current_app.instance_path, "..", "spiffworkflow_backend", "data_migrations", "version_*.py")
+        files = sorted(glob.glob(file_glob))
+        if len(files) == 0:
+            raise DataMigrationFilesNotFoundError(f"Could not find data migration with expected glob: {file_glob}")
+        return files
+
+    # modified from https://gist.github.com/vinovator/d864555d9e82d25e52fd
+    @classmethod
+    def generate_md5_for_file(cls, file: str, chunk_size: int = 4096) -> str:
+        """
+        Function which takes a file name and returns md5 checksum of the file
+        """
+        hash = hashlib.md5()  # noqa: S324
+        with open(file, "rb") as f:
+            # Read the 1st block of the file
+            chunk = f.read(chunk_size)
+            # Keep reading the file until the end and update hash
+            while chunk:
+                hash.update(chunk)
+                chunk = f.read(chunk_size)
+
+        # Return the hex checksum
+        return hash.hexdigest()
+
+    @classmethod
+    def generate_migration_checksum(cls) -> dict[str, str]:
+        files = cls.get_migration_files()
+        md5checksums = {}
+        for file in files:
+            md5checksum = cls.generate_md5_for_file(file)
+            md5checksums[os.path.basename(file)] = md5checksum
+        return md5checksums

spiffworkflow-backend/tests/spiffworkflow_backend/unit/test_process_instance_migrator.py

@@ -5,6 +5,8 @@ import os
 from flask.app import Flask
 from flask.testing import FlaskClient
 from SpiffWorkflow.bpmn.serializer.migration.version_1_3 import update_data_objects  # type: ignore
+from spiffworkflow_backend.constants import SPIFFWORKFLOW_BACKEND_DATA_MIGRATION_CHECKSUM
+from spiffworkflow_backend.constants import SPIFFWORKFLOW_BACKEND_SERIALIZER_VERSION
 from spiffworkflow_backend.data_migrations.process_instance_migrator import ProcessInstanceMigrator
 from spiffworkflow_backend.data_migrations.version_1_3 import VersionOneThree
 from spiffworkflow_backend.data_migrations.version_4 import Version4
@@ -25,6 +27,29 @@ from tests.spiffworkflow_backend.helpers.test_data import load_test_spec
 
 
 class TestProcessInstanceMigrator(BaseTest):
+    def test_data_migrations_directory_has_not_changed(
+        self,
+        app: Flask,
+        client: FlaskClient,
+    ) -> None:
+        md5checksums = ProcessInstanceMigrator.generate_migration_checksum()
+        assert md5checksums == SPIFFWORKFLOW_BACKEND_DATA_MIGRATION_CHECKSUM, (
+            "Data migrations seem to have changed but checksum has not been updated. "
+            "Please update SPIFFWORKFLOW_BACKEND_DATA_MIGRATION_CHECKSUM"
+        )
+
+        highest_version = 0
+        for file in ProcessInstanceMigrator.get_migration_files():
+            current_version = os.path.basename(file).replace("version_", "").replace(".py", "").replace("_", ".")
+            if current_version == "1.3":
+                continue
+            if int(current_version) > highest_version:
+                highest_version = int(current_version)
+        assert highest_version == int(SPIFFWORKFLOW_BACKEND_SERIALIZER_VERSION), (
+            f"Highest migration version file is '{highest_version}' however "
+            f"SPIFFWORKFLOW_BACKEND_SERIALIZER_VERSION is '{SPIFFWORKFLOW_BACKEND_SERIALIZER_VERSION}'"
+        )
+
     def test_can_run_all_migrations(
         self,
         app: Flask,