@paywalls-net/filter 1.2.0 → 1.2.2

package/README.md

@@ -1,6 +1,6 @@
 # paywalls-net/filter
 
-SDK for integrating paywalls.net authorization services with CDN or edge environments.
+SDK for paywalls.net authorization and real-time licensing services. For use with CDN or edge environments.
 
 ## Install
 
@@ -10,19 +10,26 @@ npm install @paywalls-net/filter
 
 ## Environment Variables
 - `PAYWALLS_PUBLISHER_ID`: The unique identifier for the publisher using paywalls.net services.
-- `PAYWALLS_CLOUD_API_HOST`: The host for the paywalls.net API. eg `https://cloud-api.paywalls.net`.
 - `PAYWALLS_CLOUD_API_KEY`: The API key for accessing paywalls.net services. NOTE: This key should be treated like a password and kept secret and stored in a secure secrets vault or environment variable.
 
 ## Usage
+The following is an example of using the SDK with Cloudflare Workers:
+
 ```javascript
 import { init } from '@paywalls-net/filter';
 
+// Initialize the paywalls.net handler for Cloudflare
 const handleRequest = await init('cloudflare');
 
 export default {
-  async fetch(request, env, ctx) {
-    return handleRequest(request, env, ctx);
-  }
+   async fetch(request, env, ctx) {
+       let pw_response = await handleRequest(request, env, ctx);
+       if (pw_response) {
+           // If the handler returns a response, return it
+           return pw_response;
+       }
+       return fetch(request); // Proceed to origin/CDN
+   }
 };
 ```
 

package/package.json

@@ -3,7 +3,7 @@
   "description": "Client SDK for integrating paywalls.net bot filtering and authorization services into your server or CDN.",
   "author": "paywalls.net",
   "license": "MIT",
-  "version": "1.2.0",
+  "version": "1.2.2",
   "publishConfig": {
     "access": "public"
   },

package/src/index.js

@@ -5,15 +5,26 @@
 
 import { classifyUserAgent, loadAgentPatterns } from './user-agent-classification.js';
 
+const PAYWALLS_CLOUD_API_HOST = "https://cloud-api.paywalls.net";
+
 async function logAccess(cfg, request, access) {
     // Separate html from the status in the access object.
     const { response, ...status } = access;
 
     // Get all headers as a plain object (name-value pairs)
     let headers = {};
-    for (const [key, value] of request.headers.entries()) {
-        headers[key] = value;
+    if (typeof request.headers.entries === "function") {
+        // Standard Headers object (e.g., in Cloudflare Workers)
+        for (const [key, value] of request.headers.entries()) {
+            headers[key] = value;
+        }
+    } else {
+        // CloudFront headers object
+        for (const key in request.headers) {
+            headers[key] = request.headers[key][0]?.value || "";
+        }
     }
+
     const url = new URL(request.url);
     let body = {
         account_id: cfg.paywallsPublisherId,
@@ -124,10 +135,14 @@ function isCloudflareKnownBot(request) {
 }
 
 function isTestBot(request) {
-    // check if the URL has a query parameter to always test as a bot
-    const url = new URL(request.url);
-    const uaParam = url.searchParams.get("user-agent");
-    return uaParam && uaParam.includes("bot");
+    try {
+        // check if the URL has a query parameter to always test as a bot
+        const url = new URL(request.url || request.uri);
+        const uaParam = url.searchParams.get("user-agent");
+        return uaParam && uaParam.includes("bot");
+    } catch (err) {
+        throw new Error(`test bot failed: ${request.url} | ${request.uri} | ${err.message}`);
+    }
 }
 async function isPaywallsKnownBot(cfg, request) {
     const userAgent = request.headers.get("User-Agent");
@@ -140,6 +155,11 @@ async function isRecognizedBot(cfg, request) {
 }
 
 
+/**
+ * Create response in format used by most CDNs
+ * @param {*} authz 
+ * @returns 
+ */
 function setHeaders(authz) {
     let headers = {
         "Content-Type": "text/html",
@@ -157,6 +177,11 @@ function setHeaders(authz) {
     });
 }
 
+/**
+ * Create CloudFront format response 
+ * @param {*} authz 
+ * @returns 
+ */
 function setCloudFrontHeaders(authz) {
     const headers = {};
 
@@ -200,7 +225,7 @@ async function cloudflare(config = null) {
 
     return async function handle(request, env, ctx) {
         const paywallsConfig = {
-            paywallsAPIHost: env.PAYWALLS_CLOUD_API_HOST,
+            paywallsAPIHost: env.PAYWALLS_CLOUD_API_HOST || PAYWALLS_CLOUD_API_HOST,
             paywallsAPIKey: env.PAYWALLS_CLOUD_API_KEY,
             paywallsPublisherId: env.PAYWALLS_PUBLISHER_ID
         };
@@ -214,7 +239,7 @@ async function cloudflare(config = null) {
             if (authz.access === 'deny') {
                 return setHeaders(authz);
             } else {
-                console.log("Bot-like request allowed. Proceeding to origin/CDN.");
+                // console.log("Bot-like request allowed. Proceeding to origin/CDN.");
             }
         }
     };
@@ -223,7 +248,7 @@ async function cloudflare(config = null) {
 
 async function fastly(config) {
     const paywallsConfig = {
-        paywallsAPIHost: config.get('PAYWALLS_CLOUD_API_HOST'),
+        paywallsAPIHost: config.get('PAYWALLS_CLOUD_API_HOST') || PAYWALLS_CLOUD_API_HOST,
         paywallsAPIKey: config.get('PAYWALLS_API_KEY'),
         paywallsPublisherId: config.get('PAYWALLS_PUBLISHER_ID')
     };
@@ -241,34 +266,64 @@ async function fastly(config) {
         }
     };
 }
+/**
+ * Adapt to CloudFront format
+ * Lambda@Edge events see https://docs.aws.amazon.com/AmazonCloudFront/latest/DeveloperGuide/lambda-event-structure.html#lambda-event-structure-request
+ * CloudFront events see https://docs.aws.amazon.com/AmazonCloudFront/latest/DeveloperGuide/functions-event-structure.html#functions-event-structure-example
+ * @param {*} request 
+ * @returns 
+ */
+function requestShim(request) {
+    if (!request.headers.get) {
+        // add get() to headers object to adapt to CloudFront
+        request.headers.get = (name) => {
+            const header = request.headers[name.toLowerCase()];
+            return header ? header[0].value : null;
+        };
+    }
+
+    // combine the CloudFront host, request.uri and request.querystring into request.url
+    if (!request.url && request.uri) {
+        let host = request.headers.get('host');
+        request.url = `http://${host}${request.uri}`;
+        if (request.querystring) {
+            request.url += `?${request.querystring}`;
+        }
+    }
+
+    return request;
+}
 
 async function cloudfront(config) {
     const paywallsConfig = {
-        paywallsAPIHost: config.get('PAYWALLS_CLOUD_API_HOST'),
-        paywallsAPIKey: config.get('PAYWALLS_API_KEY'),
-        paywallsPublisherId: config.get('PAYWALLS_PUBLISHER_ID')
+        paywallsAPIHost: config.PAYWALLS_CLOUD_API_HOST || PAYWALLS_CLOUD_API_HOST,
+        paywallsAPIKey: config.PAYWALLS_API_KEY,
+        paywallsPublisherId: config.PAYWALLS_PUBLISHER_ID
     };
     await loadAgentPatterns(paywallsConfig);
 
-    return async function handle(request) {
+    return async function handle(event, ctx) {
+        let request = event.Records[0].cf.request;
+        request = requestShim(request);
         if (await isRecognizedBot(paywallsConfig, request)) {
             const authz = await checkAgentStatus(paywallsConfig, request);
 
-            await logAccess(paywallsConfig, request, authz);
+            // log the result asynchronously
+            ctx.callbackWaitsForEmptyEventLoop = false;
+            logAccess(paywallsConfig, request, authz);
 
             if (authz.access === 'deny') {
                 return setCloudFrontHeaders(authz);
             }
         }
-
     };
 }
 
 
-
 /**
  * Initializes the appropriate handler based on the CDN.
  * @param {string} cdn - The name of the CDN (e.g., 'cloudflare', 'fastly', 'cloudfront').
+ * @param {Object} [config={}] - Optional configuration object for the CDN handler.
  * @returns {Function} - The handler function for the specified CDN.
  */
 export async function init(cdn, config = {}) {