Merge branch 'master' of github.com:crisp-oss/email-bounce-parser

Author: valeriansaliou

lib/parser.js

@@ -35,11 +35,13 @@ var ERROR_CODES_TYPES = {
   "550 5.1.1"  : "no_such_recipient",
   "550 5.2.1"  : "recipient_disabled_or_recipient_rate_limited",
   "550 5.7.1"  : "spam_content",
+  "550 5.7.26" : "authentication_failure",
   "551"        : "user_not_local",
   "552"        : "action_aborted_exceeded_storage_allocation",
   "552 5.2.2"  : "mailbox_over_quota",
   "552 5.2.3"  : "message_size_limit",
   "553"        : "action_not_taken_mailbox_name_not_allowed",
+  "553 5.1.2"  : "no_such_recipient",
   "554"        : "transaction_failed",
   "554 5.7.1"  : "spam_content",
   "555"        : "parameters_not_recognized",
@@ -78,19 +80,20 @@ var REGEXES = {
     /Message rejected by\:\s+([\S\.]+?)$/m, // Outlook error
     /from (\S+) \[([\d\.]+)\](?:\:(\d+)\:)?/m,
     /conversation with (\S+)\[([\d\.]+)\](?:\:(\d+)\:)? timed out/m,
+    /lost connection with (\S+)\[([\d\.]+)\](?:\:(\d+)\:)?/m, // "lost connection with mx.mail-data.net[134.244.220.69]"
     /Generating server: (\S+)(?:\[([\d\.]+)\])?(?:\:(\d+)\:)?/m // "Generating server: PH7PR02MB9549495.namprd02.prod.outlook.com"
   ],
 
   error : [
-    /said\: (\d{3}(?:[\s\-]\d\.\d\.\d{1,3})?) (.+) \(in reply to/,
-    /said\: (\d{3}(?:[\s\-]\d\.\d\.\d{1,3})?)-(.+)/,
+    /said\: (\d{3}(?:[\s\-\#\']{1,2}\d\.\d\.\d{1,3})?) (.+) \(in reply to/,
+    /said\: (\d{3}(?:[\s\-\#\']{1,2}\d\.\d\.\d{1,3})?)-(.+)/,
     /(\d{3}\-\'\d\.\d\.\d{1,3}) (.+)/,
-    /LMTP error after .+?\:\<.+?\>\: (\d{3}(?:[\s\-]\d\.\d\.\d{1,3})?) <.+?\>(.+)$/m,
-    /remote server returned \'(\d{3}(?:[\s\-]\d\.\d\.\d{1,3})?) (.+)\'/i,
-    /The response from the remote server was:\s(\d{3}(?:[\s\-]\d\.\d\.\d{1,3})?) (.+)/,
-    /refused to talk to me\: (\d{3}(?:[\s\-]\d\.\d\.\d{1,3})?) (.+)/,
-    /(?:error|reason)\:\s*?(\d{3}(?:[\s\-]\d\.\d\.\d{1,3})?) (.+)/i,
-    /(\d{3}[\s\-]\d\.\d\.\d{1,3})/,
+    /LMTP error after .+?\:\<.+?\>\: (\d{3}(?:[\s\-\#\']{1,2}\d\.\d\.\d{1,3})?) <.+?\>(.+)$/m,
+    /remote server returned \'(\d{3}(?:[\s\-\#\']{1,2}\d\.\d\.\d{1,3})?) (.+)\'/i,
+    /The response from the remote server was:\s(\d{3}(?:[\s\-\#\']{1,2}\d\.\d\.\d{1,3})?) (.+)/,
+    /refused to talk to me\: (\d{3}(?:[\s\-\#\']{1,2}\d\.\d\.\d{1,3})?) (.+)/,
+    /(?:error|reason)\:\s*?(\d{3}(?:[\s\-\#\']{1,2}\d\.\d\.\d{1,3})?) (.+)/i,
+    /(\d{3}[\s\-\#\']{1,2}\d\.\d\.\d{1,3})/,
     /(\d{3}\-\'\d\.\d\.\d{1,3})/
   ],
 
@@ -110,7 +113,7 @@ 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,
+    /there was a temporary problem (?:while )?delivering your message to (.+?). Gmail will retry/i,
     /the following message to \<(.+?)\> was undeliverable/i,
     /to the following addresses:\s*(.+)/i,
     /^your message wasn't delivered to (.+?)\s/mi, // "Your message wasn't delivered to [email protected]"
@@ -122,8 +125,8 @@ var REGEXES = {
   ],
 
   temporary_safe : /(temporary|temporarily)/i,
-  blocked : /((?:is listed)|blocked|banned|denied|dnsbl)/i,
-  spam : /(spam)/i,
+  blocked : /((?:is listed)|blocked|block list|banned|denied|dnsbl)/i,
+  spam : /(spam|unsolicited)/i,
 
   enhanced_types : {
     no_such_recipient : [
@@ -150,13 +153,13 @@ var REGEXES = {
     ],
 
     reputation : [
-      /(rate limited due to IP reputation)/i
+      /(rate limited due to IP reputation)/i,
+      /(new or untrusted IP)/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"
-    /(\d{3})\-\'(\d\.\d\.\d{1,3})/, // "550-'5.4.1" or "550-'5.4.11" or "550-'5.4.110"
+    /(\d{3})[\s\-\#\']{1,2}(\d\.\d\.\d{1,3})/,  // "550 5.4.1" or "550 #5.4.1" or "550-5.4.1" or "550-'5.4.1"
     /(\d{3})/ // "550"
   ]
 };
@@ -552,6 +555,7 @@ class Parser {
     _str = _str.replace(this.__regexes.line_break_double_indent_single, "\n\n");
 
     // Replace single Line-break with regular space
+    // TODO: re2 doesn't support lookbehinds, so for now we use a pure-JS regex
     _str = _str.replace(/(?<!\n)\n(?!\n)/gm, " ");
 
     // Replace long indent with double Line-break

package.json

@@ -1,6 +1,6 @@
 {
   "name": "email-bounce-parser",
-  "version": "1.0.20",
+  "version": "1.0.22",
   "description": "Parses bounce emails and extract errors",
   "author": "Eliott Vincent <[email protected]>",
   "main": "lib/index.js",

test/fixtures/421_4_7_0_deferred.txt

@@ -0,0 +1,16 @@
+This is the mail system at host mailer.acme.email.
+
+I'm sorry to have to inform you that your message could not
+be delivered to one or more recipients. It's attached below.
+
+For further assistance, please send mail to postmaster.
+
+If you do so, please include this problem report. You can
+delete your own text from the attached returned message.
+
+                   The mail system
+
+<[email protected]>: host mta6.am0.yahoodns.net[76.159.240.37] said: 421
+    4.7.0 [TSS04] Messages from 39.177.48.16 temporarily deferred due to
+    unexpected volume or user complaints - 4.16.55.1; see
+    https://postmaster.yahooinc.com/error-codes (in reply to MAIL FROM command)

test/fixtures/450_4_2_1_rate_limited.txt

@@ -0,0 +1,21 @@
+This is the mail system at host mailer.acme.email.
+
+I'm sorry to have to inform you that your message could not
+be delivered to one or more recipients. It's attached below.
+
+For further assistance, please send mail to postmaster.
+
+If you do so, please include this problem report. You can
+delete your own text from the attached returned message.
+
+                   The mail system
+
+<[email protected]>: host alt1.gmail-smtp-in.l.google.com[124.215.9.62] said:
+    450-4.2.1 The user you are trying to contact is receiving mail at a rate
+    that 450-4.2.1 prevents additional messages from being delivered. Please
+    resend your 450-4.2.1 message at a later time. If the user is able to
+    receive mail at that 450-4.2.1 time, your message will be delivered. For
+    more information, please 450-4.2.1 visit 450 4.2.1
+    https://support.google.com/mail/?p=ReceivingRate
+    f15-20020a056402354f00b00548d4febb0bsi965249edd.614 - gsmtp (in reply to
+    RCPT TO command)

test/fixtures/450_rate_limited.txt

@@ -0,0 +1,14 @@
+This is the mail system at host mailer.acme.email.
+
+I'm sorry to have to inform you that your message could not
+be delivered to one or more recipients. It's attached below.
+
+For further assistance, please send mail to postmaster.
+
+If you do so, please include this problem report. You can
+delete your own text from the attached returned message.
+
+                   The mail system
+
+<[email protected]>: host mta6.am0.yahoodns.net[76.159.240.27] said: 450 User is
+    receiving mail too quickly tnmpmscs (in reply to RCPT TO command)

test/fixtures/451_4_7_1_reputation.txt

@@ -0,0 +1,15 @@
+This is the mail system at host mailer.acme.email.
+
+I'm sorry to have to inform you that your message could not
+be delivered to one or more recipients. It's attached below.
+
+For further assistance, please send mail to postmaster.
+
+If you do so, please include this problem report. You can
+delete your own text from the attached returned message.
+
+                   The mail system
+
+<[email protected]>: host mx2.emailsrvr.com[180.166.34.2] said: 451 4.7.1
+    Received too many messages from a new or untrusted IP: 93.177.84.61
+    (Z27/6DC2420) (G28) (in reply to RCPT TO command)

test/fixtures/451_4_7_500_server_busy.txt

@@ -0,0 +1,18 @@
+This is the mail system at host mailer.acme.email.
+
+I'm sorry to have to inform you that your message could not
+be delivered to one or more recipients. It's attached below.
+
+For further assistance, please send mail to postmaster.
+
+If you do so, please include this problem report. You can
+delete your own text from the attached returned message.
+
+                   The mail system
+
+<[email protected]>: host
+    etzzogal-com0i.mail.protection.outlook.com[140.74.42.63] said: 451 4.7.500
+    Server busy. Please try again later from [39.177.48.16]. (S77714)
+    [QE2FRA04FT039.eop-fra04.prod.protection.outlook.com
+    2023-11-23T18:53:22.058Z 10DSKD34941BC6NE] (in reply to end of DATA
+    command)

test/fixtures/550_5_0_1_no_such_recipient.txt

@@ -0,0 +1,14 @@
+This is the mail system at host mailer.acme.email.
+
+I'm sorry to have to inform you that your message could not
+be delivered to one or more recipients. It's attached below.
+
+For further assistance, please send mail to postmaster.
+
+If you do so, please include this problem report. You can
+delete your own text from the attached returned message.
+
+                   The mail system
+
+<[email protected]>: host mxb.mailgun.org[43.106.31.24] said:
+    550 5.0.1 Recipient rejected (in reply to RCPT TO command)

test/fixtures/550_5_1_0_address_rejected.txt

@@ -0,0 +1,14 @@
+This is the mail system at host mailer.acme.email.
+
+I'm sorry to have to inform you that your message could not
+be delivered to one or more recipients. It's attached below.
+
+For further assistance, please send mail to postmaster.
+
+If you do so, please include this problem report. You can
+delete your own text from the attached returned message.
+
+                   The mail system
+
+<[email protected]>: host mx1.hc3459-45.iphmx.com[193.183.64.148] said:
+    550 #5.1.0 Address rejected. (in reply to RCPT TO command)

test/fixtures/550_5_7_1_blocked.txt

@@ -0,0 +1,20 @@
+This is the mail system at host mailer.acme.email.
+
+I'm sorry to have to inform you that your message could not
+be delivered to one or more recipients. It's attached below.
+
+For further assistance, please send mail to postmaster.
+
+If you do so, please include this problem report. You can
+delete your own text from the attached returned message.
+
+                   The mail system
+
+<[email protected]>: host
+    outlook-com.olc.protection.outlook.com[140.74.15.33] said: 550 5.7.1
+    Unfortunately, messages from [31.57.142.204] weren't sent. Please contact
+    your Internet service provider since part of their network is on our block
+    list (S3140). You can also refer your provider to
+    http://mail.live.com/mail/troubleshooting.aspx#errors.
+    [WP1FDK33EO054.eop-nam02.prod.protection.outlook.com
+    2023-01-01T00:00:00.001Z 03SDKS32D1WER234] (in reply to MAIL FROM command)