email-bounce-parser 1.0.18 → 1.0.20

package/lib/parser.js

@@ -110,12 +110,13 @@ var REGEXES = {
 
   recipient : [
     /the following addresses had permanent fatal errors (?:.*?)\s\<(.+?)\>/i,
+    /there was a temporary problem delivering your message to (.+?). Gmail will retry/i,
     /the following message to \<(.+?)\> was undeliverable/i,
     /to the following addresses:\s*(.+)/i,
-    /^\<(.+?)\>:/m,
-    /^(.+?)\s?(?:\s\(.*?\))?\n?\s*[\[|<]mailto\:(.+?)[\]|>]/mi, // "john@acme.com<mailto:john@acme.com>" or "john@acme.com[mailto:john@acme.com]" or "john@acme.com (john@acme.com)<mailto:john@acme.com>"
     /^your message wasn't delivered to (.+?)\s/mi, // "Your message wasn't delivered to john@acme.com"
     /^your message to (.+?) couldn\'t/mi, // "Your message to john@acme.com couldn't"
+    /^(.+?)\s?(?:\s\(.*?\))?\n?\s*[\[|<]mailto\:(.+?)[\]|>]/mi, // "john@acme.com<mailto:john@acme.com>" or "john@acme.com[mailto:john@acme.com]" or "john@acme.com (john@acme.com)<mailto:john@acme.com>"
+    /^\<(.+?)\>:/m,
     /\<(.+?)\> \(expanded from \<.+?\>\)\:/, // "<john@acme.com> (expanded from <john@digiflex.com>): "
     /\<(.+?)\>:/ // "<john@acme.com>:"
   ],
@@ -124,28 +125,34 @@ var REGEXES = {
   blocked : /((?:is listed)|blocked|banned|denied|dnsbl)/i,
   spam : /(spam)/i,
 
-  enhanced_types_no_such_recipient : [
-    /(recipient address rejected: access denied)/i, // Outlook DBEB for invalid addresses
-    /(no such (?:account|user|mailbox|recipient))/i,
-    /((?:terminated|disabled|inactive|not found|unavailable|invalid) (?:account|user|mailbox|recipient))/i,
-    /(do(e)?sn\'t exist)/i, // Some Chinese SMTP return "dosn't" instead of "doesn't"
-    /(user (?:.+?) does not exist)/i,
-    /((?:account|user|mailbox|recipient) (?:.+@.+ )?(?:is )?(?:terminated|disabled|inactive|not found|unavailable|invalid))/i
-  ],
-
-  enhanced_types_server_unknown : [
-    /(does not accept mail)/i,
-    /(domain (?:name\s+)?(?:\S+\s)?(?:not found|does not exist))/i, // "domain does not exist" or "domain name does not exist" or "domain acme.com does not exist" or "domain name acme.com does not exist"
-    /(no route to host)/i
-  ],
-
-  enhanced_types_server_timeout : [
-    /(timeout|timed[\s\-]out)/i
-  ],
-
-  enhanced_types_mailbox_full : [
-    /(out of storage|(?:mailbox (?:is )?full)|(?:user (?:is )?over quota))/i // "user is over quota" (iCloud)
-  ],
+  enhanced_types : {
+    no_such_recipient : [
+      /(recipient address rejected: access denied)/i, // Outlook DBEB for invalid addresses
+      /(no such (?:account|user|mailbox|recipient))/i,
+      /((?:terminated|disabled|inactive|not found|unavailable|invalid) (?:account|user|mailbox|recipient))/i,
+      /(do(e)?sn\'t exist)/i, // Some Chinese SMTP return "dosn't" instead of "doesn't"
+      /(user (?:.+?) does not exist)/i,
+      /((?:account|user|mailbox|recipient) (?:.+@.+ )?(?:is )?(?:terminated|disabled|inactive|not found|unavailable|invalid|rejected))/i
+    ],
+
+    server_unknown : [
+      /(does not accept mail)/i,
+      /(domain (?:name\s+)?(?:\S+\s)?(?:not found|does not exist))/i, // "domain does not exist" or "domain name does not exist" or "domain acme.com does not exist" or "domain name acme.com does not exist"
+      /(no route to host)/i
+    ],
+
+    server_timeout : [
+      /(timeout|timed[\s\-]out)/i
+    ],
+
+    mailbox_full : [
+      /(out of storage|(?:mailbox (?:is )?full)|(?:user (?:is )?over quota))/i // "user is over quota" (iCloud)
+    ],
+
+    reputation : [
+      /(rate limited due to IP reputation)/i
+    ]
+  },
 
   error_code : [
     /(\d{3})[\s\-](\d\.\d\.\d{1,3})/,  // "550 5.4.1" or "550-5.4.1" or "550-5.4.11" or "550-5.4.110"
@@ -164,9 +171,7 @@ class Parser {
    * Constructor
    */
   constructor() {
-    this.__regexes = {};
-
-    this.__initRegexes();
+    this.__regexes = this.__initRegexes(REGEXES);
   }
 
 
@@ -243,30 +248,31 @@ class Parser {
   /**
    * Initializes regexes
    * @private
-   * @return {undefined}
+   * @param  {object} obj
+   * @return {object} Regexes
    */
-  __initRegexes() {
-    for (var _key in REGEXES) {
-      var _key_line = `${_key}_line`;
-      var _entry    = REGEXES[_key];
+  __initRegexes(obj) {
+    var _regexes = {};
+
+    for (var _key in obj) {
+      var _entry = obj[_key];
 
       if (Array.isArray(_entry)) {
-        this.__regexes[_key]      = [];
-        this.__regexes[_key_line] = [];
+        _regexes[_key] = [];
 
         for (var _i = 0; _i < _entry.length; _i++) {
-          var _regex = _entry[_i];
-
-          this.__regexes[_key].push(
-            new RE2(_regex)
+          _regexes[_key].push(
+            new RE2(_entry[_i])
           );
         }
+      } else if (typeof _entry === "object" && Object.keys(_entry).length > 0) {
+        _regexes[_key] = this.__initRegexes(_entry);
       } else {
-        var _regex = _entry;
-
-        this.__regexes[_key] = new RE2(_regex);
+        _regexes[_key] = new RE2(_entry);
       }
     }
+
+    return _regexes;
   }
 
 
@@ -299,21 +305,22 @@ class Parser {
         _permanent = true;
       }
 
-      // Make sure the bounce is actually permanent
-      if (_permanent === true) {
-        var _match = this.__regexes.temporary_safe.match(str);
-
-        if (_match && _match.length > 1) {
-          _temporary = true;
-          _permanent = false;
-        }
-      }
-
       if (_code.basic) {
         _type = ERROR_CODES_TYPES[_code.basic] || null
       }
     }
 
+    // In cases where the bounce is detected as permanent, or the \
+    //   temporary / permanent status is still unclear, run additional detection
+    if (_permanent === true || (_temporary === false && _permanent === false)) {
+      var _match = this.__regexes.temporary_safe.match(str);
+
+      if (_match && _match.length > 1) {
+        _temporary = true;
+        _permanent = false;
+      }
+    }
+
     // No label detected? Probably because there's no error code
     if (!_label) {
       // Attempt a lax check
@@ -385,34 +392,22 @@ class Parser {
       _spam = false;
 
     // Detect enhanced-type from bounce content
-    var _match = loopRegexes(this.__regexes.enhanced_types_no_such_recipient, str);
-
-    if (_match && _match.length > 1) {
-      _type = "no_such_recipient";
-    }
-
-    _match = loopRegexes(this.__regexes.enhanced_types_server_unknown, str);
-
-    if (!_type && _match && _match.length > 1) {
-      _type = "server_unknown";
-    }
+    for (var _key in this.__regexes.enhanced_types) {
+      var _match = loopRegexes(
+        this.__regexes.enhanced_types[_key],
 
-    _match = loopRegexes(this.__regexes.enhanced_types_server_timeout, str);
+        str
+      );
 
-    if (!_type && _match && _match.length > 1) {
-      _type = "server_timeout";
-    }
-
-    _match = loopRegexes(this.__regexes.enhanced_types_mailbox_full, str);
-
-    if (!_type && _match && _match.length > 1) {
-      _type = "mailbox_full";
+      if (!_type && _match && _match.length > 1) {
+        _type = _key;
+      }
     }
 
     // Detect enhanced-type from enhanced error code
     // Notice: this is done as a last resort, as sometimes enhanced code \
     //   doesn't clearly convey the actual error
-    if (!_type && _match && code.enhanced) {
+    if (!_type && code.enhanced) {
       _type = ERROR_CODES_TYPES[`${code.basic} ${code.enhanced}`] || null;
     }
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "email-bounce-parser",
-  "version": "1.0.18",
+  "version": "1.0.20",
   "description": "Parses bounce emails and extract errors",
   "author": "Eliott Vincent <eliott@crisp.chat>",
   "main": "lib/index.js",
@@ -23,7 +23,7 @@
     "nodeunit-x": "0.15.0"
   },
   "dependencies": {
-    "re2": "1.20.1"
+    "re2": "1.20.7"
   },
   "keywords": [
     "parser",