http-snapshotter 0.5.3-beta.1 → 0.5.3

package/README.md

@@ -62,6 +62,12 @@ Once you are done writing your tests, run your test runner on all your tests and
 
 The tests of this library uses this library itself, check the `tests/` directory and try the tests `npm ci; npm test`.
 
+## Limitations
+
+1. MSW interceptor cannot mock `GET` calls with body. So trying to mock elasticsearch / opensearch `GET /<index>/_search` calls with body breaks. I ended up using `POST` only when unit testing, as `GET` is the right way to use with AWS `AmazonOpenSearchServiceReadOnlyAccess` IAM permission 🤷‍♂️.
+
+2. Stripe SDK uses node http module in a way MSW interceptor cannot mock - [reference](https://github.com/mswjs/msw/issues/2259#issuecomment-2379379566). Solution mentioned in a comment there is to use Stripe SDK with `fetch` client - [reference](https://github.com/mswjs/msw/issues/2259#issuecomment-2422672039). Another workaround is to mock Stripe SDK's methods with your testing library of choice.
+
 ## About snapshot files and its names
 
 A snapshot file name uniquely identifies a request. By default it is a combination of HTTP method + URL + body that makes a request unique (headers are ignored).
@@ -200,3 +206,24 @@ WARNING: This module isn't concurrent or thread safe. Make sure that:
 1. within one worker only one test is being executed at a time. e.g. If you use `ava`, and you have multiple `test()` blocks in one file, you need to change it to run serially with `test.serial()`.
 
 2. parallel tests don't update the same snapshot file at the same time (i.e. while you run with SNAPSHOT=update). Regardless, updating snapshots of multiple tests at the same time is not a great idea in my opinion, because reviewing the snapshots files are a pain, escpecially if you have a shared snapshot files.
+
+## Ignoring requests from snapshots
+
+Sometimes you may want certain requests to not be snapshotted (e.g., file uploads to S3 / object storage) so that you can layer function / result snapshotting on top. Use `attachSnapshotIgnoreRules()` to provide a function that determines which requests should be ignored:
+
+```js
+import { attachSnapshotIgnoreRules, resetSnapshotIgnoreRules } from "http-snapshotter";
+
+// Ignore analytics and health check requests
+attachSnapshotIgnoreRules((request) => {
+  const url = new URL(request.url);
+  return url.hostname.includes('/upload');
+});
+
+// Reset to default (no requests ignored)
+resetSnapshotIgnoreRules();
+```
+
+**Behavior varies by mode:**
+- `SNAPSHOT=update/append`: Ignored requests make real network calls but don't create snapshots
+- `SNAPSHOT=read`: Ignored requests throw an error (tests shouldn't make real network calls)

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "http-snapshotter",
-  "version": "0.5.3-beta.1",
+  "version": "0.5.3",
   "description": "Snapshot HTTP requests for tests (node.js)",
   "main": "index.js",
   "types": "index.d.ts",