sadpandajoe commented on code in PR #45595:
URL: https://github.com/apache/superset/pull/45595#discussion_r4352503541


##########
superset-frontend/spec/helpers/swcJestTransformer.js:
##########
@@ -0,0 +1,87 @@
+/**
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements.  See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership.  The ASF licenses this file
+ * to you under the Apache License, Version 2.0 (the
+ * "License"); you may not use this file except in compliance
+ * with the License.  You may obtain a copy of the License at
+ *
+ *   http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing,
+ * software distributed under the License is distributed on an
+ * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+ * KIND, either express or implied.  See the License for the
+ * specific language governing permissions and limitations
+ * under the License.
+ */
+import { createTransformer as createSwcTransformer } from '@swc/jest';
+
+/**
+ * Jest transformer: @swc/jest plus one post-processing step.
+ *
+ * SWC's CommonJS output defines every ESM export as a lazy, non-configurable
+ * getter, which is why `jest.spyOn(module, 'name')` and
+ * `jest.replaceProperty(module, 'name', value)` throw "Cannot redefine
+ * property" on anything SWC transformed. Babel's output defined plain
+ * properties, so the suite relies on spying on module exports everywhere.
+ * This wrapper marks those getters `configurable: true` so the spies keep
+ * working, while leaving the getters lazy (an eager rewrite would break
+ * modules that import each other in a cycle).
+ *
+ * The second difference is what a module in an import cycle sees when it
+ * reads an export that has not been initialised yet. Babel assigns
+ * `exports.name = void 0` up front, so the read yields `undefined`; SWC's
+ * getter returns the live binding and throws a ReferenceError (temporal dead
+ * zone) instead. Several suites spread `jest.requireActual()` of a module
+ * that is still mid-evaluation, so the getters are made lenient to match
+ * Babel: an uninitialised export reads as `undefined`.
+ */
+// SWC keeps a leading doc comment of a default export inside the call, as
+// `Object.defineProperty(exports, /** ... */ "default", {`, hence the optional
+// block comment before the name.
+const EXPORT_GETTER_DESCRIPTOR =
+  /(Object\.defineProperty\((?:target, name|to, 
k|exports,\s*(?:\/\*[\s\S]*?\*\/\s*)?"[^"]+"), \{\s*enumerable: 
true,)(\s*)(get:)/g;
+// `get name () { return binding; }` inside `_export(exports, {...})`, and the
+// `get: function() { return binding; }` form SWC uses for a single export.
+const EXPORT_GETTER_BODY =
+  /(get(?: \w+ \(\)|: function\(\)) \{\s*)return (\w+|from\[k\]);/g;
+
+function mirrorBabelExportSemantics(code) {
+  return code
+    .replace(EXPORT_GETTER_DESCRIPTOR, '$1$2configurable: true,$2$3')
+    .replace(
+      EXPORT_GETTER_BODY,
+      '$1try { return $2; } catch (e) { return undefined; }',

Review Comment:
   This replacement runs over the whole generated file and catches every 
exception, so it also rewrites ordinary object getters (e.g. `get value() { 
return later; }`) and re-export getters (`from[k]`). A getter that throws a 
genuine error, or reads a real TDZ binding in application code, now silently 
returns `undefined` under Jest while production still throws, so a test could 
pass on a bug. Could the rewrite be limited to the `_export(exports, 
{...})`/`Object.defineProperty(exports, ...)` descriptors and only swallow 
`ReferenceError`?



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to