feat: config schema validation

Author: adityak74

.env.example

@@ -1,3 +1,4 @@
 API_PORT=5000
 API_HOST=localhost
+MAX_PAGES_TO_CRAWL=45
 NODE_ENV=development

package-lock.json

@@ -15,6 +15,7 @@
     "express-fileupload": "^1.4.3",
     "glob": "^10.3.10",
     "inquirer": "^9.2.12",
+    "joi": "^17.11.0",
     "playwright": "*",
     "prettier": "^3.1.0"
   },

package.json

@@ -1,4 +1,8 @@
 import type { Page } from "playwright";
+import Joi from "joi";
+import { configDotenv } from "dotenv";
+
+configDotenv();
 
 export type Config = {
   /**
@@ -39,3 +43,17 @@ export type Config = {
   /** Optional timeout for waiting for a selector to appear */
   waitForSelectorTimeout?: number;
 };
+
+export const ConfigSchema = Joi.object({
+  url: Joi.string().uri().required(),
+  match: Joi.alternatives(Joi.string().required(), Joi.array().items(Joi.string().required())).required(),
+  selector: Joi.string().default(''),
+  maxPagesToCrawl: Joi.number().integer().default(process.env.MAX_PAGES_TO_CRAWL || 50),
+  outputFileName: Joi.string().default('output.json'),
+  cookie: Joi.object({
+    name: Joi.string().required(),
+    value: Joi.string().required(),
+  }).optional(),
+  onVisitPage: Joi.function().optional(),
+  waitForSelectorTimeout: Joi.number().optional(),
+});

src/config.ts

@@ -2,7 +2,7 @@ import express from 'express';
 import cors from 'cors';
 import { readFile } from 'fs/promises';
 import { crawl, write } from "./core.js";
-import { Config } from './config.js';
+import { Config, ConfigSchema } from './config.js';
 import { configDotenv } from 'dotenv';
 
 configDotenv();
@@ -18,9 +18,10 @@ app.use(express.json());
 app.post('/crawl', async (req, res) => {
     const config: Config = req.body;
     try {
-        await crawl(config);
-        await write(config);
-        const outputFileContent = await readFile(config.outputFileName, 'utf-8');
+        const validatedConfig = ConfigSchema.validate(config).value;
+        await crawl(validatedConfig);
+        await write(validatedConfig);
+        const outputFileContent = await readFile(validatedConfig.outputFileName, 'utf-8');
         res.contentType('application/json');
         return res.send(outputFileContent);
     } catch (error) {