Anonymization in/out

lib/anonymizer.js

@@ -0,0 +1,109 @@
+const multiparty = require('multiparty');
+
+const emailRegex = /([a-zA-Z0-9._%+-]+@[a-zA-Z0-9.-]+\.[a-zA-Z]{2,6})/g;
+const fullNameRegex = /\b([A-Z][a-z]+ [A-Z][a-z]+)\b/g; 
+const singleNameRegex = /\b([A-Z][a-z]+)\b/g; 
+const addressRegex = /\b(\d{1,5}\s\w+(?:\s\w+)*(?:\s(Street|St|Avenue|Ave|Boulevard|Blvd|Road|Rd|Lane|Ln|Drive|Dr|Court|Ct|Circle|Cir))?)\b/g;
+const phoneRegex = /(\+?\d{1,4}[\s-]?)?(\(?\d{3}\)?[\s-]?)?[\d\s-]{7,10}/g;
+const urlRegex = /https?:\/\/[^\s/$.?#].[^\s]*/g;
+const stateRegex = /\b[A-Z]{2}\b/g;
+const zipRegex = /\b\d{5}(?:-\d{4})?\b/g;
+const passwordRegex = /(?<=name="password"[\s\S]*?\r\n\r\n)([A-Za-z0-9]+[A-Za-z0-9]*|[A-Za-z0-9]*[A-Za-z0-9]+)/g;
+
+//default
+let anonymizationEnabled = true;
+
+//Text
+function anonymizeText(text) {
+    if (typeof text !== 'string') {
+        return text;
+    }
+    text = text.replace(emailRegex, '[email protected]');
+    text = text.replace(phoneRegex, '000-000-0000');
+    text = text.replace(urlRegex, 'https://anonymized-url.com');
+    text = text.replace(fullNameRegex, 'Anonymized Name');
+    text = text.replace(singleNameRegex, (match, p1, offset, string) => {
+        if (fullNameRegex.test(string.substring(offset - p1.length, offset + p1.length))) {
+            return match;
+        }
+        return 'Anonymized Name';
+    });
+    text = text.replace(passwordRegex, 'anonymized_password');
+    text = text.replace(/\bAnonymized Name Anonymized Name\b/g, 'Anonymized Name');
+    return text;
+}
+
+//Object
+function anonymizeObject(obj) {
+    if (!anonymizationEnabled) {
+        return obj;
+    }
+
+    for (let key in obj) {
+        if (obj.hasOwnProperty(key)) {
+            if (typeof obj[key] === 'object' && !Array.isArray(obj[key])) {
+                anonymizeObject(obj[key]);
+            } else if (Array.isArray(obj[key])) {
+                obj[key] = obj[key].map(item => anonymizeText(item));
+            } else {
+                if (key === 'street' || key === 'city' || key === 'country') {
+                    obj[key] = 'Anonymized Address';
+                } else if (key === 'state') {
+                    obj[key] = 'XX';
+                } else if (key === 'zip') {
+                    obj[key] = '00000';
+                } else {
+                    obj[key] = anonymizeText(obj[key]);
+                }
+            }
+        }
+    }
+    return obj;
+}
+
+//Form
+function anonymizeFormData(req, callback) {
+    if (!anonymizationEnabled) {
+        return callback(null, req);
+    }
+
+    const form = new multiparty.Form();
+    const fields = {};
+    const files = {};
+
+    form.on('field', (name, value) => {
+        if (name === 'password') {
+            fields[name] = 'anonymized_password';
+        } else {
+            fields[name] = anonymizeText(value);
+        }
+    });
+
+    form.on('file', (name, file) => {
+        files[name] = file;
+    });
+
+    form.on('close', () => {
+        callback(null, { fields, files });
+    });
+
+    form.on('error', (err) => {
+        callback(err, null);
+    });
+
+    form.parse(req);
+}
+
+//flag setter
+function setAnonymization(enabled) {
+    anonymizationEnabled = enabled;
+}
+
+
+module.exports = {
+    anonymizeObject,
+    anonymizeFormData,
+    anonymizeText,
+    setAnonymization
+};
+

lib/main.js

@@ -9,6 +9,7 @@ var http = require('http'),
     gh = require('./lib-helper.js'),
     proxy = require('./ThrottleProxy/throttleServer.js'),
     config = require('./config.js');
+    anonymizer = require('../lib/anonymizer.js');
 
 
 
@@ -56,6 +57,9 @@ var timers=[];
 var minDelay;//ms
 var maxDelay;//ms
 var isRandomDelay=false;
+
+var anonymizationEnabled = gh.args(true);
+
 /**
  * END of Properties section
  * *************************
@@ -76,6 +80,11 @@ var EVENT={
     error:'error'
 }
 
+function anonymizeMode(flag) {
+    anonymizationEnabled = flag;
+    anonymizer.setAnonymization(flag);
+}
+
 function incrCounter(id){
     if(countObj[id]!=null){
         countObj[id]++;
@@ -408,6 +417,11 @@ function request(options,cb){
     }else{
         _options=options;
     }
+
+    if (_options.body && anonymizationEnabled) {
+        _options.body = anonymizer.anonymizeText(_options.body);
+    }
+
     let protocol;
     if (options.port === "443" && !isThrottled()) {
         _options['agent'] = gh.keepAliveAgentHttps;
@@ -417,19 +431,30 @@ function request(options,cb){
         _options['agent'] = gh.keepAliveAgentHttp;
         protocol = http;
     }
-    var req = protocol.request(_options, function (response) {
-        response.on('end', function () {
+    var req = protocol.request(_options, function(response) {
+        let responseBody = ''; 
+
+        response.on('data', function(chunk) {
+            responseBody += chunk;
+        });
+
+        response.on('end', function() {
             incrCounter(response.statusCode);
+            if (anonymizationEnabled) {
+                responseBody = anonymizer.anonymizeText(responseBody);
+            }
+            if (cb) {
+                cb(response, responseBody);
+            }
         });
 
-        if(cb){
-            cb(response);
-        }
         response.resume();
     });
-    if(options.body){
-        req.write(options.body)
+
+    if (options.body) {
+        req.write(options.body);
     }
+
     req.end();
     return req;
 }
@@ -549,6 +574,7 @@ exports.randomDelayBetweenRequests = function(value){randomDelayBetweenRequests=
 exports.throttleRequests_bps = function(value){throttleRequests_bps=value};
 exports.isPipelinedRequest = function(value){isPipelinedRequest=value};
 exports.debugMode = function(value){debugMode=value; gh.consoleDebug(value); gh.shouldDebug(value)};
+exports.anonymizeMode = anonymizeMode;
 exports.request=request;
 exports.multiRequest=multiRequest;
 exports.isThrottled=isThrottled;