fix: cap Retry-After at 60 seconds and trim trailing slashes from baseURL (#19)

Author: TheGreatAxios

README.md

@@ -53,24 +53,24 @@ console.log(vectors.length, vectors[0]?.length); // 2 768
 
 ### Config
 
-| Field            | Type                   | Description                                                                                 |
-| ---------------- | ---------------------- | ------------------------------------------------------------------------------------------- |
-| `baseURL`        | `string`               | Server root with its version path, such as `http://host:11434/v1`.                          |
-| `model`          | `string`               | Model name.                                                                                 |
-| `apiKey`         | `string?`              | Sent as a bearer token.                                                                     |
-| `dimensions`     | `number?`              | Requested output dimension. Sent only when set; a reply of any other dimension is rejected. |
-| `encodingFormat` | `"float" \| "base64"?` | Wire format. Defaults to `float`. Results are always `number[]`.                            |
-| `batchSize`      | `number?`              | Texts per request. Defaults to 32.                                                          |
-| `timeoutMs`      | `number?`              | Per-attempt timeout; a timed-out attempt is retried. Defaults to 30000.                     |
+| Field            | Type                   | Description                                                                                      |
+| ---------------- | ---------------------- | ------------------------------------------------------------------------------------------------ |
+| `baseURL`        | `string`               | Server root with its version path, such as `http://host:11434/v1`. Trailing slashes are ignored. |
+| `model`          | `string`               | Model name.                                                                                      |
+| `apiKey`         | `string?`              | Sent as a bearer token.                                                                          |
+| `dimensions`     | `number?`              | Requested output dimension. Sent only when set; a reply of any other dimension is rejected.      |
+| `encodingFormat` | `"float" \| "base64"?` | Wire format. Defaults to `float`. Results are always `number[]`.                                 |
+| `batchSize`      | `number?`              | Texts per request. Defaults to 32.                                                               |
+| `timeoutMs`      | `number?`              | Per-attempt timeout; a timed-out attempt is retried. Defaults to 30000.                          |
 
 ### Options
 
-| Field                 | Description                                                                     |
-| --------------------- | ------------------------------------------------------------------------------- |
-| `deps`                | `{ fetch, scheduler }`. Defaults to global `fetch` and Interchange's scheduler. |
-| `retryPolicy`         | Retry policy. Defaults to Interchange's policy.                                 |
-| `extractRetryAfterMs` | Reads the server's retry delay. Defaults to parsing `Retry-After`.              |
-| `signal`              | Aborts all pending requests.                                                    |
+| Field                 | Description                                                                              |
+| --------------------- | ---------------------------------------------------------------------------------------- |
+| `deps`                | `{ fetch, scheduler }`. Defaults to global `fetch` and Interchange's scheduler.          |
+| `retryPolicy`         | Retry policy. Defaults to Interchange's policy.                                          |
+| `extractRetryAfterMs` | Reads the server's retry delay. Defaults to parsing `Retry-After`, capped at 60 seconds. |
+| `signal`              | Aborts all pending requests.                                                             |
 
 ### Errors
 

e2e/retry.test.ts

@@ -165,3 +165,16 @@ test("reports a caller abort as aborted, without retrying", async () => {
     reason: { category: "aborted" },
   });
 });
+
+test("caps a Retry-After at 60 seconds", async () => {
+  harness = setupHarness({ enableInferenceTimers: true });
+  reply(harness, "rate limited", 429, { "retry-after": "86400" });
+  reply(harness, JSON.stringify({ data: [{ index: 0, embedding: [1] }] }), 200);
+
+  const pending = embedTexts(["a"], CONFIG, { deps: harness.deps });
+  await harness.run();
+
+  expect(await pending).toEqual([[1]]);
+  expect(harness.clock.now()).toBeGreaterThanOrEqual(60_000);
+  expect(harness.clock.now()).toBeLessThan(61_000);
+});

e2e/wire.test.ts

@@ -37,6 +37,25 @@ test("sends only configured fields to {baseURL}/embeddings", async () => {
   expect(await run(harness, pending)).toEqual([[1]]);
 });
 
+test("trims trailing slashes from baseURL", async () => {
+  harness = setupHarness();
+  const stream = harness.scenario.createStream();
+  harness.scenario.whenRequestMatches(
+    (req) => req.url === "https://embed.example/v1/embeddings",
+    stream,
+  );
+  stream.enqueueAll(
+    [new TextEncoder().encode('{"data":[{"index":0,"embedding":[1]}]}')],
+    { startAt: 1 },
+  );
+  const pending = embedTexts(
+    ["a"],
+    { ...CONFIG, baseURL: "https://embed.example/v1//" },
+    { deps: harness.deps },
+  );
+  expect(await run(harness, pending)).toEqual([[1]]);
+});
+
 test("sends a bearer token only when apiKey is set", async () => {
   harness = setupHarness();
   const stream = harness.scenario.createStream();

src/embed.ts

@@ -86,7 +86,7 @@ function buildRequest(
   }
 
   return {
-    url: `${config.baseURL}/embeddings`,
+    url: `${config.baseURL.replace(/\/+$/, "")}/embeddings`,
     headers,
     // Unset `dimensions` and `encoding_format` are dropped by JSON.stringify.
     body: JSON.stringify({

src/request.ts

@@ -63,12 +63,18 @@ type Attempt =
 
 export const DEFAULT_TIMEOUT_MS = 30_000;
 
+const MAX_RETRY_AFTER_MS = 60_000;
+
+const clampRetryAfter = (ms: number): number =>
+  Math.min(MAX_RETRY_AFTER_MS, Math.max(0, ms));
+
 /**
  * `Retry-After` in seconds or as an HTTP date; undefined when absent.
  *
- * Both branches clamp at zero. A server may name an instant that has already
- * passed, and `Retry-After: -5` is not unheard of; a negative delay would flow
- * into the retry policy as though it were a pacing hint.
+ * Both branches clamp to [0, 60s]. A server may name an instant that has
+ * already passed, and `Retry-After: -5` is not unheard of; a negative delay
+ * would flow into the retry policy as though it were a pacing hint. A day-long
+ * `Retry-After` would stall the caller for that long.
  *
  * The HTTP-date branch is wall-clock by necessity — the header names an
  * absolute instant, so it cannot be resolved against the harness scheduler's
@@ -79,10 +85,10 @@ export const extractRetryAfterMs: RetryAfterExtractor = (headers) => {
   if (header === undefined || header === "") return undefined;
 
   const seconds = Number(header);
-  if (Number.isFinite(seconds)) return Math.max(0, seconds * 1_000);
+  if (Number.isFinite(seconds)) return clampRetryAfter(seconds * 1_000);
 
   const at = Date.parse(header);
-  return Number.isNaN(at) ? undefined : Math.max(0, at - Date.now());
+  return Number.isNaN(at) ? undefined : clampRetryAfter(at - Date.now());
 };
 
 async function attemptOnce(