Austin/firecrawl
1
0
Fork 0
mirror of https://github.com/mendableai/firecrawl.git synced 2024-11-16 11:42:24 +08:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

tests

Browse Source
This commit is contained in:
rafaelmmiller 2024-11-14 17:03:54 -03:00
parent 45091430ab
commit be5e6da5ff
2 changed files with 133 additions and 124 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

248
apps/api/src/__tests__/e2e_extract/index.test.ts
View File

@ -10,142 +10,142 @@ dotenv.config();
const TEST_URL = "http://127.0.0.1:3002"; const TEST_URL = "http://127.0.0.1:3002";
describe("E2E Tests for Extract API Routes", () => { describe("E2E Tests for Extract API Routes", () => {
describe("POST /v1/extract", () => { it.concurrent("should return authors of blog posts on firecrawl.dev", async () => {
it.concurrent("should return authors of blog posts on firecrawl.dev", async () => { const response = await request(TEST_URL)
const response = await request(TEST_URL) .post("/v1/extract")
.post("/v1/extract") .set("Authorization", `Bearer ${process.env.TEST_API_KEY}`)
.set("Authorization", `Bearer ${process.env.TEST_API_KEY}`) .set("Content-Type", "application/json")
.set("Content-Type", "application/json") .send({
.send({ urls: ["https://firecrawl.dev"],
urls: ["https://firecrawl.dev"], prompt: "Who are the authors of the blog posts?",
prompt: "Who are the authors of the blog posts?", schema: {
schema: { type: "object",
type: "object", properties: { authors: { type: "array", items: { type: "string" } } },
properties: { authors: { type: "array", items: { type: "string" } } }, },
}, });
});
expect(response.statusCode).toBe(200);
expect(response.body).toHaveProperty("data");
expect(response.body.data).toHaveProperty("founders");
let gotItRight = 0; console.log(response.body);
for (const author of response.body.data?.authors) { expect(response.statusCode).toBe(200);
if (author.includes("Caleb Peffer")) gotItRight++; expect(response.body).toHaveProperty("data");
if (author.includes("Gergő Móricz")) gotItRight++; expect(response.body.data).toHaveProperty("authors");
if (author.includes("Eric Ciarla")) gotItRight++;
if (author.includes("Nicolas Camara")) gotItRight++;
}
expect(gotItRight).toBeGreaterThan(3); let gotItRight = 0;
}, 60000); for (const author of response.body.data?.authors) {
if (author.includes("Caleb Peffer")) gotItRight++;
if (author.includes("Gergő Móricz")) gotItRight++;
if (author.includes("Eric Ciarla")) gotItRight++;
if (author.includes("Nicolas Camara")) gotItRight++;
}
it.concurrent("should return founders of firecrawl.dev (allowExternalLinks = true)", async () => { expect(gotItRight).toBeGreaterThan(1);
const response = await request(TEST_URL) }, 60000);
.post("/v1/extract")
.set("Authorization", `Bearer ${process.env.TEST_API_KEY}`)
.set("Content-Type", "application/json")
.send({
urls: ["mendable.ai"],
prompt: "Who are the founders of the company?",
allowExternalLinks: true,
schema: {
type: "object",
properties: { founders: { type: "array", items: { type: "string" } } },
},
});
expect(response.statusCode).toBe(200);
expect(response.body).toHaveProperty("data");
expect(response.body.data).toHaveProperty("founders");
let gotItRight = 0; it.concurrent("should return founders of firecrawl.dev (allowExternalLinks = true)", async () => {
for (const founder of response.body.data?.founders) { const response = await request(TEST_URL)
if (founder.includes("Caleb")) gotItRight++; .post("/v1/extract")
if (founder.includes("Eric")) gotItRight++; .set("Authorization", `Bearer ${process.env.TEST_API_KEY}`)
if (founder.includes("Nicolas")) gotItRight++; .set("Content-Type", "application/json")
} .send({
urls: ["mendable.ai"],
prompt: "Who are the founders of the company?",
allowExternalLinks: true,
schema: {
type: "object",
properties: { founders: { type: "array", items: { type: "string" } } },
},
});
expect(response.statusCode).toBe(200);
expect(response.body).toHaveProperty("data");
expect(response.body.data).toHaveProperty("founders");
expect(gotItRight).toBe(3); let gotItRight = 0;
}, 60000); for (const founder of response.body.data?.founders) {
if (founder.includes("Caleb")) gotItRight++;
if (founder.includes("Eric")) gotItRight++;
if (founder.includes("Nicolas")) gotItRight++;
}
it.concurrent("should return hiring opportunities on firecrawl.dev (allowExternalLinks = true)", async () => { expect(gotItRight).toBe(3);
const response = await request(TEST_URL) }, 60000);
.post("/v1/extract")
.set("Authorization", `Bearer ${process.env.TEST_API_KEY}`)
.set("Content-Type", "application/json")
.send({
urls: ["https://firecrawl.dev"],
prompt: "What are they hiring for?",
allowExternalLinks: true,
schema: {
type: "array",
items: {
type: "string"
}
},
});
expect(response.statusCode).toBe(200);
expect(response.body).toHaveProperty("data");
console.log(response.body.data);
let gotItRight = 0; it.concurrent("should return hiring opportunities on firecrawl.dev (allowExternalLinks = true)", async () => {
for (const hiring of response.body.data?.items) { const response = await request(TEST_URL)
if (hiring.includes("Developer Relations Specialist")) gotItRight++; .post("/v1/extract")
if (hiring.includes("Web Automation Engineer")) gotItRight++; .set("Authorization", `Bearer ${process.env.TEST_API_KEY}`)
if (hiring.includes("Developer Experience Engineer")) gotItRight++; .set("Content-Type", "application/json")
if (hiring.includes("Developer Support Engineer")) gotItRight++; .send({
if (hiring.includes("Dev Ops Engineer")) gotItRight++; urls: ["https://firecrawl.dev"],
if (hiring.includes("Founding Web Automation Engineer")) gotItRight++; prompt: "What are they hiring for?",
} allowExternalLinks: true,
schema: {
type: "array",
items: {
type: "string"
}
},
});
expect(response.statusCode).toBe(200);
expect(response.body).toHaveProperty("data");
console.log(response.body.data);
expect(gotItRight).toBeGreaterThan(5); let gotItRight = 0;
}, 60000); for (const hiring of response.body.data?.items) {
if (hiring.includes("Developer Relations Specialist")) gotItRight++;
if (hiring.includes("Web Automation Engineer")) gotItRight++;
if (hiring.includes("Developer Experience Engineer")) gotItRight++;
if (hiring.includes("Developer Support Engineer")) gotItRight++;
if (hiring.includes("Dev Ops Engineer")) gotItRight++;
if (hiring.includes("Founding Web Automation Engineer")) gotItRight++;
}
it.concurrent("should return PCI DSS compliance for Fivetran", async () => { expect(gotItRight).toBeGreaterThan(5);
const response = await request(TEST_URL) }, 60000);
.post("/v1/extract")
.set("Authorization", `Bearer ${process.env.TEST_API_KEY}`) it.concurrent("should return PCI DSS compliance for Fivetran", async () => {
.set("Content-Type", "application/json") const response = await request(TEST_URL)
.send({ .post("/v1/extract")
urls: ["fivetran.com"], .set("Authorization", `Bearer ${process.env.TEST_API_KEY}`)
prompt: "Does Fivetran have PCI DSS compliance?", .set("Content-Type", "application/json")
allowExternalLinks: true, .send({
schema: { urls: ["fivetran.com"],
prompt: "Does Fivetran have PCI DSS compliance?",
allowExternalLinks: true,
schema: {
type: "object",
properties: {
pciDssCompliance: { type: "boolean" }
}
},
});
expect(response.statusCode).toBe(200);
expect(response.body).toHaveProperty("data");
expect(response.body.data?.pciDssCompliance).toBe(true);
}, 60000);
it.concurrent("should return Azure Data Connectors for Fivetran", async () => {
const response = await request(TEST_URL)
.post("/v1/extract")
.set("Authorization", `Bearer ${process.env.TEST_API_KEY}`)
.set("Content-Type", "application/json")
.send({
urls: ["fivetran.com"],
prompt: "What are the Azure Data Connectors they offer?",
schema: {
type: "array",
items: {
type: "object", type: "object",
properties: { properties: {
pciDssCompliance: { type: "boolean" } connector: { type: "string" },
} description: { type: "string" },
}, supportsCaptureDelete: { type: "boolean" }
});
expect(response.statusCode).toBe(200);
expect(response.body).toHaveProperty("data");
expect(response.body.data?.pciDssCompliance).toBe(true);
}, 60000);
it.concurrent("should return Azure Data Connectors for Fivetran", async () => {
const response = await request(TEST_URL)
.post("/v1/extract")
.set("Authorization", `Bearer ${process.env.TEST_API_KEY}`)
.set("Content-Type", "application/json")
.send({
urls: ["fivetran.com"],
prompt: "What are the Azure Data Connectors they offer?",
schema: {
type: "array",
items: {
type: "object",
properties: {
connector: { type: "string" },
description: { type: "string" },
supportsCaptureDelete: { type: "boolean" }
}
} }
} }
}) }
})
console.log(response.body); console.log(response.body);
// expect(response.statusCode).toBe(200); // expect(response.statusCode).toBe(200);
// expect(response.body).toHaveProperty("data"); // expect(response.body).toHaveProperty("data");
// expect(response.body.data?.pciDssCompliance).toBe(true); // expect(response.body.data?.pciDssCompliance).toBe(true);
}, 60000); }, 60000);
});
}); });

9
apps/api/src/types.ts
View File

@ -106,6 +106,15 @@ export interface FirecrawlCrawlStatusResponse {
error?: string; error?: string;
} }
export interface FirecrawlExtractResponse {
statusCode: number;
body: {
success: boolean;
data: any[];
};
error?: string;
}
export enum RateLimiterMode { export enum RateLimiterMode {
Crawl = "crawl", Crawl = "crawl",
CrawlStatus = "crawlStatus", CrawlStatus = "crawlStatus",