Cook Up Powerful Mail-Processing Scripts with Jython Procmail Recipes

Cook Up Powerful Mail-Processing Scripts with Jython Procmail Recipes

esides its email-processing function, the ubiquitous, open-source Procmail utility can be configured to serve as a local delivery agent for any mail transfer agent (MTA) such as Sendmail and Postfix as well. In this capacity, Procmail essentially acts like a filter that can be configured to process selected messages en route to a destination mailbox. This article demonstrates how easy it is to use Procmail and Jython to implement powerful mail-processing scripts that route mail and parse message contents to insert selected values into a database.

Despite Jython’s slight lag behind the C Python and current JDK implementations, it is a robust and stable language that interfaces easily with Java libraries. Jython’s command-line interactive interpreter can save you time when trying to figure out how to use an unfamiliar API. It provides instantaneous feedback to method calls so you don’t have to code and compile a true Java class. (Click here to read another article I wrote on the benefits of using Jython). These advantages, among many others, make Jython a sensible scripting language choice for creating robust Procmail email processing recipes.

What You Need:
* Jython 2.1, which may be freely downloaded from
* A Linux or Unix box with Procmail installed. A free Procmail download is available.
* A beginning to intermediate knowledge of Java and an understanding of how the Java classpath mechanism works. Some rudimentary knowledge of Python is helpful.

Procmail: Content Filter/Message-Processing Tool

As mentioned previously, Procmail can serve as a content filter. However, if you’re unfamiliar with the utility, at first glance it probably will seem more like a services framework than an actual message-processing tool. In truth, it is a little of both. When Procmail is passed an email message, it can process the message in its entirety by passing it to another program such as Spam Assassin or by forwarding it to a different mailbox. Alternatively, Procmail can pass the message header, body, or both as “standard in” to another script or application for processing.

To get Procmail to perform either of these processing actions, you must set up a resource configuration file named .procmailrc, which contains the processing rules or “recipes”, in each mail user’s home directory. For those who aren’t Linux/Unix folk, set the “home directory” for each user listed in the /etc/passwd file. From a functional standpoint, it is the place where the user’s “.profile” file resides and where, if you issue the cd command without any arguments, the shell will take you by default. As mail is delivered to the user, Procmail refers to the .procmailrc file in this directory for the processing action(s) it should take.

You define the .procmailrc file recipe(s) with a syntax that is unusual but not especially difficult to understand, as the following simple sample file shows:

Listing 1. Sample .procmailrc File
1. PATH=/usr/bin:/home/emailer/jython:$PATH2. MAILDIR=/var/spool/mail3. LOGFILE=$HOME/procmail.log4. VERBOSE=15. SHELL=/bin/bash6. DEFAULT=/home/emailer/garbagemail7. :0 HBb8. *^Subject:.*Thanks for Shopping With Us*9. {10. :0 c11. ! kgood12. :0 13. | /home/emailer/jython/driver 14. }

* The numbering isn't part of the .procmailrc. The colon-zero (:0)
beginning lines 7, 10, and 12 aren't typos.

You can configure the environment in which Procmail does its work prior to the execution of any processing rules (lines 1-6 above). Most of the environmental settings are familiar to those in the Unix/Linux landscape. (See the Resources section for links to more information on the details of Procmail environmental configuration). One important variable worth mentioning though, is the default location that is specified at line 6. This location is where messages will be stored when they don’t meet the recipe pattern match condition(s) for processing.

The start of each processing recipe is denoted with a colon followed by a zero, as shown on line 7. The repeat of this pattern within the curly braces are the processing actions to be taken (lines 10 and 12). So let’s break down this syntax to learn what this recipe does.

Line 7 specifies that the pattern defined on line 8 should be matched against the header (H) and body (B) of the email message and that only the body (B) should be passed to any defined script as its standard input. The processing rule on line 10 uses the “c” option to instruct Procmail to make a carbon copy of the message, and the exclamation point starting line 11 is the Procmail symbol for email forwarding. Line 11 then, forwards a copy of the message to the “kgood” user on the localhost.

The second rule, beginning with the :0 at line 12, uses the pipe character, just like the Unix command line pipe, to pass the message body to the standard in of the specified script. The example Jython script available for download with this article demonstrates how easy it is to use regular expressions and JDBC in Jython to parse and store data of interest from the message.

Parse and Store Data of Interest

Jython handles standard in as a file-like object. The main loop of the Jython script is as follows:

RCPT=RE("RCPT To:.*")   #pattern to match in message bodylines=sys.stdin.readlines( )      # this is the body of the email message#Python and Jython have a simple, clean syntax that relies on indentations 
rather than braces to demarcate blocks of code for line in lines:
#start looping through message bodyif(RCPT.match(line)==1): rParts=line.split('<') rcptUser=rParts[1][:-1] f.write("RCPT: " + rcptUser + " ") #debug log statement

Within this main loop, patterns in the message body are identified using the Jakarta Regexp Java API. Again, since Jython is implemented in Java and seamlessly interfaces with Java APIs, you need to ensure only that the requisite .jar file is on the Jython classpath in order to access third-party Java libraries. The initial imports for this script include both the Jakarta Regexp libraries and the required JDBC classes from the standard Oracle classes12.jar file:

from oracle.jdbc.driver import OracleDriverfrom java.sql import DriverManagerfrom java.sql import Connectionfrom java.sql import Statementfrom java.sql import ResultSetfrom java.sql import SQLExceptionfrom java.lang import Classfrom org.apache.regexp import REimport sys

Finally, after parsing out the data of interest using regular expressions, the script assembles an insert statement string. Be careful when concatenating the text of such a statement. Oracle requires single quotes around varchar type data, and if you're going to include variables defined elsewhere in the script, you have to be sure that they aren't enclosed in quotes. Here's a simple example where recipient and bounceDate are variables assigned values earlier in the script:

ddl="insert into table_email_bounce (email_address, bounce_date) 
values(lower('" + str(recipient) + "'), to_date
('" + str(bounceDate) + "', 'Dy, DD Mon YYYY HH24:MI:SS'))"f.write(ddl)st.executeQuery(ddl)con.close()

Customize Your Messaging Environment

Procmail is a versatile, easy-to-use, email message-processing utility. When configured with an MTA, Procmail provides email-processing capabilities such as mail forwarding and also offers a way to pass messages to the standard in of a custom processing script. Jython's ability to access Java libraries coupled with Procmail's local delivery services enables you to tailor your messaging environment to meet your needs.


About Our Editorial Process

At DevX, we’re dedicated to tech entrepreneurship. Our team closely follows industry shifts, new products, AI breakthroughs, technology trends, and funding announcements. Articles undergo thorough editing to ensure accuracy and clarity, reflecting DevX’s style and supporting entrepreneurs in the tech sphere.

See our full editorial policy.

About Our Journalist