import {z} from 'zod' export type GetApiApiStatsPidSpiderSidJobsArgs = {page?: number | undefined, page_size?: number | undefined, start_date: string, end_date: string, offset?: number | undefined}; export type GetApiApiStatsPidSpiderSidJobsOptions = {baseUrl?: string | undefined, Basic?: {username: string, password: string} | undefined, fetch?: typeof fetch | undefined}; export const spiderJobArg = z.object({name: z.string().min(1).max(1000), value: z.string().min(1).max(1000)}); export const spiderJobEnvVar = z.object({evid: z.number().int().nullable().optional(), name: z.string().min(1).max(1000), value: z.string().min(1).max(12000), masked: z.boolean().optional()}); export const spiderJobTag = z.object({name: z.string().min(1).max(50)}); export const jobsStats = z.object({total_jobs: z.number().int().optional(), waiting_jobs: z.number().int().optional(), running_jobs: z.number().int().optional(), stopped_jobs: z.number().int().optional(), completed_jobs: z.number().int().optional(), in_queue_jobs: z.number().int().optional(), error_jobs: z.number().int().optional()}); export const pagesStats = z.object({total_pages: z.number().int().optional(), scraped_pages: z.number().int().optional(), missed_pages: z.number().int().optional()}); export const statusCodesStats = z.object({status_200: z.number().int().optional(), status_301: z.number().int().optional(), status_302: z.number().int().optional(), status_401: z.number().int().optional(), status_403: z.number().int().optional(), status_404: z.number().int().optional(), status_429: z.number().int().optional(), status_500: z.number().int().optional()}); export const logsStats = z.object({total_logs: z.number().int().optional(), debug_logs: z.number().int().optional(), info_logs: z.number().int().optional(), warning_logs: z.number().int().optional(), error_logs: z.number().int().optional(), critical_logs: z.number().int().optional()}); export const fieldCoverageStats = z.object({field_name: z.string().min(1).optional(), field_count: z.number().int().optional(), field_coverage: z.number().optional()}); export const coverageStats = z.object({total_items: z.number().int().optional(), total_items_coverage: z.number().optional(), fields: z.array(fieldCoverageStats).optional()}); export const stats = z.object({jobs: jobsStats.optional(), pages: pagesStats, items_count: z.number().int().optional(), runtime: z.number().optional(), status_codes: statusCodesStats, success_rate: z.number().optional(), logs: logsStats, coverage: coverageStats.optional()}); export const spiderJobStats = z.object({jid: z.number().int().optional(), spider: z.string().optional(), created: z.string().optional(), name: z.string().min(1).optional(), lifespan: z.number().optional(), total_response_bytes: z.number().int().gte(0).lte(18446744073709552000).optional(), item_count: z.number().int().gte(0).lte(18446744073709552000).optional(), request_count: z.number().int().gte(0).lte(18446744073709552000).optional(), args: z.array(spiderJobArg).optional(), env_vars: z.array(spiderJobEnvVar).optional(), tags: z.array(spiderJobTag).optional(), job_status: z.string().min(1).optional(), cronjob: z.number().int().nullable().optional(), data_expiry_days: z.number().int().gte(0).lte(65535).nullable().optional(), data_status: z.enum(["PERSISTENT", "PENDING", "DELETED"]).optional(), stats: stats}); export const jobsPagination = z.object({count: z.number().int(), next: z.string().nullable().optional(), previous: z.string().nullable().optional(), results: z.array(spiderJobStats)}); export const validateResponse = async ( schema: { readonly '~standard': { readonly validate: (value: unknown) => ({ readonly value: Output; readonly issues?: undefined } | { readonly issues: ReadonlyArray<{ readonly message: string; readonly path?: ReadonlyArray | undefined }> }) | Promise<{ readonly value: Output; readonly issues?: undefined } | { readonly issues: ReadonlyArray<{ readonly message: string; readonly path?: ReadonlyArray | undefined }> }> } }, value: unknown ): Promise => { const result = await schema['~standard'].validate(value) if (result.issues !== undefined) { const detail = result.issues .map(({ message, path }) => path?.length ? `${path.map((segment) => String(typeof segment === 'object' ? segment.key : segment)).join('.')}: ${message}` : message ) .join('; ') throw new Error(`Response does not match the schema: ${detail}`, { cause: { issues: result.issues } }) } return result.value }; /** * Retrieve all the jobs of a spider executed in a range of dates. * * GET /api/stats/{pid}/spider/{sid}/jobs */ export const getApiApiStatsPidSpiderSidJobs = async (args: GetApiApiStatsPidSpiderSidJobsArgs, options?: GetApiApiStatsPidSpiderSidJobsOptions) => { const {baseUrl = "http://127.0.0.1:8000", fetch: fetchFn = fetch, Basic} = options ?? {} const query = new URLSearchParams() const headers = new Headers() headers.set('accept', "application/json") if (args.page != null) query.append("page", String(args.page)) if (args.page_size != null) query.append("page_size", String(args.page_size)) if (args.start_date != null) query.append("start_date", String(args.start_date)) if (args.end_date != null) query.append("end_date", String(args.end_date)) if (args.offset != null) query.append("offset", String(args.offset)) if (Basic !== undefined) headers.set('authorization', `Basic ${btoa(`${Basic.username}:${Basic.password}`)}`) const url = `${baseUrl.replace(/\/+$/, '')}/api/stats/${encodeURIComponent(String(args.pid))}/spider/${encodeURIComponent(String(args.sid))}/jobs${query.size > 0 ? `?${query}` : ''}` const res = await fetchFn(url, {method: "GET", headers}) if (!res.ok) throw new Error(`Request failed with status ${res.status}`, {cause: {status: res.status, body: await res.text()}}) return validateResponse(jobsPagination, await res.json()) };