Skip to content

Supercharge your AI agents with data from the web and beyond. Building the library for superintelligence. 🔥

test(api): exercise both developer search routes in the snips suite

Karan Lokchandani3bf814df3587
diff --git a/apps/api/src/__tests__/snips/v2/developer.test.ts b/apps/api/src/__tests__/snips/v2/developer.test.ts
index e97cc480f..077a48e91 100644
--- a/apps/api/src/__tests__/snips/v2/developer.test.ts
+++ b/apps/api/src/__tests__/snips/v2/developer.test.ts
@@ -10,8 +10,9 @@ const KEYLESS_ENABLED =
process.env.KEYLESS_REQUESTS_PER_DAY !== undefined &&
process.env.KEYLESS_CREDITS_PER_DAY !== undefined;
-const CANONICAL_PATH = "/v2/search/developer";
-const UPSTREAM_MIRROR_PATH = "/v2/developer/search";
+const CANONICAL_PATH = "/v2/developer/search";
+const ALIAS_PATH = "/v2/search/developer";
+const SERVING_PATHS = [CANONICAL_PATH, ALIAS_PATH];
const COVERAGE_STATUSES = ["ok", "degraded", "unavailable", "skipped"];
@@ -33,73 +34,60 @@ async function waitForSingleRow<T>(
}
describeIf(HAS_RESEARCH)("Developer Search API", () => {
- it("serves developer search from the canonical mount", async () => {
- const identity = await idmux({
- name: "developer/canonical search",
- credits: 100,
- });
-
- const res = await researchRaw(
- CANONICAL_PATH,
- { query: "how do I configure retries", k: 3 },
- identity,
- );
-
- expect(res.statusCode).toBe(200);
- expect(res.body.success).toBe(true);
- expect(Array.isArray(res.body.results)).toBe(true);
- for (const type of ["doc", "issue", "pull_request", "readme"]) {
- expect(COVERAGE_STATUSES).toContain(res.body.coverage[type]);
- }
- expect(typeof res.body.reranked).toBe("boolean");
-
- for (const result of res.body.results) {
- expect(typeof result.id).toBe("string");
- expect(typeof result.url).toBe("string");
- expect(["doc", "issue", "pull_request", "readme"]).toContain(result.type);
- expect(Array.isArray(result.passages)).toBe(true);
- expect(result.license).toBeUndefined();
- }
- }, 120000);
-
- it("serves the same search from a POST body", async () => {
- const identity = await idmux({
- name: "developer/post body",
- credits: 100,
- });
+ describe.each(SERVING_PATHS)("developer search on %s", path => {
+ it("serves developer search", async () => {
+ const identity = await idmux({
+ name: `developer/get ${path}`,
+ credits: 100,
+ });
- const res = await researchPostRaw(
- CANONICAL_PATH,
- { query: "graceful shutdown", k: 2, types: ["issue", "readme"] },
- identity,
- );
+ const res = await researchRaw(
+ path,
+ { query: "how do I configure retries", k: 3 },
+ identity,
+ );
- expect(res.statusCode).toBe(200);
- expect(res.body.success).toBe(true);
- expect(Array.isArray(res.body.results)).toBe(true);
- for (const result of res.body.results) {
- expect(["issue", "readme"]).toContain(result.type);
- }
- expect(res.body.coverage.doc).toBe("skipped");
- expect(res.body.coverage.pull_request).toBe("skipped");
- }, 120000);
+ expect(res.statusCode).toBe(200);
+ expect(res.body.success).toBe(true);
+ expect(Array.isArray(res.body.results)).toBe(true);
+ for (const type of ["doc", "issue", "pull_request", "readme"]) {
+ expect(COVERAGE_STATUSES).toContain(res.body.coverage[type]);
+ }
+ expect(typeof res.body.reranked).toBe("boolean");
+
+ for (const result of res.body.results) {
+ expect(typeof result.id).toBe("string");
+ expect(typeof result.url).toBe("string");
+ expect(["doc", "issue", "pull_request", "readme"]).toContain(
+ result.type,
+ );
+ expect(Array.isArray(result.passages)).toBe(true);
+ expect(result.license).toBeUndefined();
+ }
+ }, 120000);
- it("serves developer search from the upstream-mirror mount", async () => {
- const identity = await idmux({
- name: "developer/mirror mount",
- credits: 100,
- });
+ it("serves the same search from a POST body", async () => {
+ const identity = await idmux({
+ name: `developer/post ${path}`,
+ credits: 100,
+ });
- const res = await researchRaw(
- UPSTREAM_MIRROR_PATH,
- { query: "connection pool sizing", k: 1 },
- identity,
- );
+ const res = await researchPostRaw(
+ path,
+ { query: "graceful shutdown", k: 2, types: ["issue", "readme"] },
+ identity,
+ );
- expect(res.statusCode).toBe(200);
- expect(res.body.success).toBe(true);
- expect(Array.isArray(res.body.results)).toBe(true);
- }, 120000);
+ expect(res.statusCode).toBe(200);
+ expect(res.body.success).toBe(true);
+ expect(Array.isArray(res.body.results)).toBe(true);
+ for (const result of res.body.results) {
+ expect(["issue", "readme"]).toContain(result.type);
+ }
+ expect(res.body.coverage.doc).toBe("skipped");
+ expect(res.body.coverage.pull_request).toBe("skipped");
+ }, 120000);
+ });
it("keeps the upstream casing on repo enrollment status", async () => {
const identity = await idmux({
@@ -226,8 +214,8 @@ describeIf(HAS_RESEARCH)("Developer Search API", () => {
}, 120000);
describeIf(KEYLESS_ENABLED)("keyless developer search", () => {
- it("permits keyless access on the canonical mount", async () => {
- const res = await researchRaw(CANONICAL_PATH, {
+ it("permits keyless access on the alias mount", async () => {
+ const res = await researchRaw(ALIAS_PATH, {
query: "retry backoff",
k: 1,
});
@@ -237,8 +225,8 @@ describeIf(HAS_RESEARCH)("Developer Search API", () => {
expect(Array.isArray(res.body.results)).toBe(true);
}, 120000);
- it("refuses keyless access on the upstream-mirror mount", async () => {
- const res = await researchRaw(UPSTREAM_MIRROR_PATH, {
+ it("refuses keyless access on the canonical mount", async () => {
+ const res = await researchRaw(CANONICAL_PATH, {
query: "retry backoff",
k: 1,
});