Spaces:
Sleeping
Sleeping
Download apps/api/src/controllers/v1/deep-research.ts from s-eto-hub/firecrawl-api: direct link, hf CLI and curl.
- Browser
- Download file 4.11 kB
-
https://huggingface.co/spaces/s-eto-hub/firecrawl-api/resolve/main/apps/api/src/controllers/v1/deep-research.ts
- Command line
-
hf download hf://spaces/s-eto-hub/firecrawl-api/apps/api/src/controllers/v1/deep-research.ts
-
curl -L -o deep-research.ts https://huggingface.co/spaces/s-eto-hub/firecrawl-api/resolve/main/apps/api/src/controllers/v1/deep-research.ts
4.11 kB
| import { v7 as uuidv7 } from "uuid"; | |
| import { Request, Response } from "express"; | |
| import { ErrorResponse, extractOptions, RequestWithAuth } from "./types"; | |
| import { getDeepResearchQueue } from "../../services/queue-service"; | |
| import * as Sentry from "@sentry/node"; | |
| import { saveDeepResearch } from "../../lib/deep-research/deep-research-redis"; | |
| import { z } from "zod"; | |
| import { logRequest } from "../../services/logging/log_job"; | |
| import { getScrapeZDR } from "../../lib/zdr-helpers"; | |
| const deepResearchRequestSchema = z | |
| .object({ | |
| query: z.string().describe("The query or topic to search for").optional(), | |
| maxDepth: z | |
| .number() | |
| .min(1) | |
| .max(12) | |
| .default(7) | |
| .describe("Maximum depth of research iterations"), | |
| maxUrls: z | |
| .number() | |
| .min(1) | |
| .max(1000) | |
| .default(20) | |
| .describe("Maximum number of URLs to analyze"), | |
| timeLimit: z | |
| .number() | |
| .min(30) | |
| .max(600) | |
| .default(300) | |
| .describe("Time limit in seconds"), | |
| analysisPrompt: z | |
| .string() | |
| .describe("The prompt to use for the final analysis") | |
| .optional(), | |
| systemPrompt: z | |
| .string() | |
| .describe("The system prompt to use for the research agent") | |
| .optional(), | |
| formats: z.array(z.enum(["markdown", "json"])).default(["markdown"]), | |
| // @deprecated Use query instead | |
| topic: z.string().describe("The topic or question to research").optional(), | |
| jsonOptions: extractOptions.optional(), | |
| }) | |
| .refine(data => data.query || data.topic, { | |
| message: "Either query or topic must be provided", | |
| }) | |
| .refine( | |
| obj => { | |
| const hasJsonFormat = obj.formats?.includes("json"); | |
| const hasJsonOptions = obj.jsonOptions !== undefined; | |
| return ( | |
| (hasJsonFormat && hasJsonOptions) || (!hasJsonFormat && !hasJsonOptions) | |
| ); | |
| }, | |
| { | |
| message: | |
| "When 'json' format is specified, jsonOptions must be provided, and vice versa", | |
| }, | |
| ) | |
| .transform(data => ({ | |
| ...data, | |
| query: data.topic || data.query, // Use topic as query if provided | |
| })); | |
| type DeepResearchRequest = z.infer<typeof deepResearchRequestSchema>; | |
| type DeepResearchResponse = | |
| | ErrorResponse | |
| | { | |
| success: boolean; | |
| id: string; | |
| }; | |
| /** | |
| * Initiates a deep research job based on the provided topic. | |
| * @param req - The request object containing authentication and research parameters. | |
| * @param res - The response object to send the research job ID. | |
| * @returns A promise that resolves when the research job is queued. | |
| */ | |
| export async function deepResearchController( | |
| req: RequestWithAuth<{}, DeepResearchResponse, DeepResearchRequest>, | |
| res: Response<DeepResearchResponse>, | |
| ) { | |
| if (getScrapeZDR(req.acuc?.flags) === "forced") { | |
| return res.status(400).json({ | |
| success: false, | |
| error: | |
| "Your team has zero data retention enabled. This is not supported on deep research. Please contact support@firecrawl.com to unblock this feature.", | |
| }); | |
| } | |
| req.body = deepResearchRequestSchema.parse(req.body); | |
| const researchId = uuidv7(); | |
| await logRequest({ | |
| id: researchId, | |
| kind: "deep_research", | |
| api_version: "v1", | |
| team_id: req.auth.team_id, | |
| origin: "api", | |
| target_hint: req.body.query ?? "", | |
| zeroDataRetention: false, // not supported for deep research | |
| api_key_id: req.acuc?.api_key_id ?? null, | |
| }); | |
| const jobData = { | |
| request: req.body, | |
| teamId: req.auth.team_id, | |
| subId: req.acuc?.sub_id ?? undefined, | |
| apiKeyId: req.acuc?.api_key_id ?? null, | |
| researchId, | |
| }; | |
| await saveDeepResearch(researchId, { | |
| id: researchId, | |
| team_id: req.auth.team_id, | |
| createdAt: Date.now(), | |
| status: "processing", | |
| currentDepth: 0, | |
| maxDepth: req.body.maxDepth, | |
| completedSteps: 0, | |
| totalExpectedSteps: req.body.maxDepth * 5, // 5 steps per depth level | |
| findings: [], | |
| sources: [], | |
| activities: [], | |
| summaries: [], | |
| }); | |
| await getDeepResearchQueue().add(researchId, jobData, { | |
| jobId: researchId, | |
| }); | |
| return res.status(200).json({ | |
| success: true, | |
| id: researchId, | |
| }); | |
| } | |