extract-from-sitemap 0.0.4 → 0.0.5

package/cli.js

@@ -8,13 +8,19 @@ const http = require("http");
 const { URL, URLSearchParams } = require("url");
 const os = require("os");
 const { extractFromSitemap } = require("./mod.js");
+
+/**
+ * @typedef {Object} OriginConfig
+ * @property {string} origin - The origin URL to process
+ * @property {boolean} forceExtract - Whether to force extraction for this origin
+ */
+
 /**
  * @typedef {Object} Config
  * @property {string} outDir - Output directory for extracted files
- * @property {string[]} origins - Array of origin URLs to process
+ * @property {OriginConfig[]} origins - Array of origin configurations
  * @property {Array<{title: string, description: string, url: string}>} customUrls - Custom URLs to extract
  * @property {boolean} keepOriginalUrls - Whether to keep original URL structure
- * @property {boolean} forceExtract - Whether to force extraction even if files exist
  */
 
 /**
@@ -209,11 +215,14 @@ async function loadConfig() {
     console.log(
       JSON.stringify(
         {
+          $schema: "https://extract.llmtext.com/llmtext.schema.json",
           outDir: "./docs",
-          origins: ["https://docs.example.com"],
+          origins: [
+            { origin: "https://docs.parallel.ai", forceExtract: false },
+            { origin: "https://parallel.ai", forceExtract: true },
+          ],
           customUrls: [],
           keepOriginalUrls: false,
-          forceExtract: false,
         },
         null,
         2
@@ -230,10 +239,22 @@ async function loadConfig() {
     if (!Array.isArray(config.origins))
       throw new Error("origins must be an array");
 
+    // Validate origin objects
+    for (const [index, originConfig] of config.origins.entries()) {
+      if (typeof originConfig !== "object" || originConfig === null) {
+        throw new Error(`origins[${index}] must be an object`);
+      }
+      if (!originConfig.origin) {
+        throw new Error(`origins[${index}].origin is required`);
+      }
+      if (typeof originConfig.forceExtract !== "boolean") {
+        throw new Error(`origins[${index}].forceExtract must be a boolean`);
+      }
+    }
+
     // Set defaults
     config.customUrls = config.customUrls || [];
     config.keepOriginalUrls = config.keepOriginalUrls ?? false;
-    config.forceExtract = config.forceExtract ?? false;
 
     return config;
   } catch (error) {
@@ -370,10 +391,9 @@ function cleanupOldFiles(outDir, currentFiles, previousFiles) {
  * Process custom URLs through extraction API
  * @param {Array<{title: string, description: string, url: string}>} customUrls - Custom URLs to process
  * @param {string} apiKey - API key for authentication
- * @param {boolean} forceExtract - Whether to force extraction
  * @returns {Promise<Record<string, any>>} Extracted files
  */
-async function processCustomUrls(customUrls, apiKey, forceExtract) {
+async function processCustomUrls(customUrls, apiKey) {
   const files = {};
 
   for (const customUrl of customUrls) {
@@ -468,14 +488,16 @@ async function main() {
     let totalPages = 0;
     let totalErrors = 0;
 
-    // Process each origin
-    for (const origin of config.origins) {
-      console.log(`\n🌐 Processing origin: ${origin}`);
+    // Process each origin with its own forceExtract setting
+    for (const originConfig of config.origins) {
+      console.log(
+        `\n🌐 Processing origin: ${originConfig.origin} (forceExtract: ${originConfig.forceExtract})`
+      );
 
       try {
         const result = await extractFromSitemap(
-          origin,
-          config.forceExtract,
+          originConfig.origin,
+          originConfig.forceExtract,
           apiKey
         );
 
@@ -493,7 +515,9 @@ async function main() {
           if (!config.keepOriginalUrls) {
             // Create domain-specific subdirectory
             const domain = new URL(
-              origin.startsWith("http") ? origin : `https://${origin}`
+              originConfig.origin.startsWith("http")
+                ? originConfig.origin
+                : `https://${originConfig.origin}`
             ).hostname;
             const domainDir = path.join(config.outDir, domain);
             fs.mkdirSync(domainDir, { recursive: true });
@@ -519,7 +543,10 @@ async function main() {
         totalPages += result.totalPages;
         totalErrors += result.errors;
       } catch (error) {
-        console.error(`❌ Error processing ${origin}:`, error.message);
+        console.error(
+          `❌ Error processing ${originConfig.origin}:`,
+          error.message
+        );
         totalErrors++;
       }
     }
@@ -527,11 +554,7 @@ async function main() {
     // Process custom URLs
     if (config.customUrls.length > 0) {
       console.log(`\n📋 Processing ${config.customUrls.length} custom URLs...`);
-      const customFiles = await processCustomUrls(
-        config.customUrls,
-        apiKey,
-        config.forceExtract
-      );
+      const customFiles = await processCustomUrls(config.customUrls, apiKey);
 
       for (const [filename, file] of Object.entries(customFiles)) {
         const filePath = path.join(config.outDir, filename);

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "extract-from-sitemap",
   "bin": "cli.js",
-  "version": "0.0.4",
+  "version": "0.0.5",
   "main": "mod.js",
   "description": "A module and CLI that allows extracting all pages from a sitemap into markdown and a llms.txt, using Parallel.ai APIs.",
   "files": [