orindi

Transform your e-mail newsletters into webpages and an RSS feed. Uses pico as a front end.

Project maintained by uriel1998 Hosted on GitHub Pages — Theme by mattgraham

orindi

Transform your e-mail newsletters into webpages and an RSS feed. Uses pico as a front end. orindi is an anglicization of ørindi, meaning “message”.

orindi logo

Contents

  1. About
  2. License
  3. Prerequisites
  4. Installation
  5. Usage
  6. TODO

1. About

Transform your e-mail newsletters into webpages and an RSS feed. Uses pico as a front end. orindi is an anglicization of ørindi, meaning “message”.

The idea is that this should be able to slip into your existing workflow, regardless of what that is or what you’re using. And if it doesn’t, it’s meant to be as simple as possible to customize the code.

Why Pico and Procmail?

I’d previously used Kill The Newsletter which did a bangup job. But when the author changed it from using a cloud service to exim, I had a problem. I don’t use exim, and the configuration was in the code that I (quite frankly) don’t have the skill to understand.

Therefore, I wanted something that would rely on procmail, which can work with pretty much any mail transfer agent seamlessly, including after delivery if mail is in the maildir format. (See TODO if you’re about to say I should use something different like courier-maildrop.)

Pico provides a good front-end for similar reasons. It’s lightweight, does not require one to disrupt their existing setup, deals with plain text files, and can generate RSS feeds fairly easily.

2. License

This project is licensed under the MIT License. For the full license, see LICENSE.

3. Prerequisites

4. Installation

Ideally, you should create a virtualenv for this project, as there are a number of python dependencies. Instructions on how to do that are beyond the scope of this document. It is assumed that you have created and activated the virtualenv henceforth.

Install MDA / procmail

Install pandoc

Install pico

The instructions and program as written presume that pico is local to orindi. If the files need to be uploaded elsewhere by a script after processing, you will need a local mirror for the files to be written into. The user that runs orindi should be part of the same user group as pico (www-data, probably).

Install python modules and application

orindi_procmailrc setup

There is only one thing to change in the procmail rc file:

:0Wc:
| /path/to/orindi/orindi_parse.py

Type in the path to orindi_parse.py on the second line, making sure you keep the pipe character. This will pass a copy of each email on to orindi, and then allow it to be delivered normally.

You may try letting procmail not wait between each email to process it; if you do, change the first line to :0c:.

There is an alternate way to use procmail to match mails instead of using orindi, but since orindi must also have that list, you’ll end up having to keep multiple lists synchronized…

Integrating procmail with your MTA is an exercise for the user.

INI files setup

orindi can have one or many ini files; it’s all up to how you want to organize things … and how many things you’re filtering for. You must have orindi.ini in the configuration directory. The main orindi.ini file must have these lines (with the appropriate values):

[DEFAULT]  
BaseDir = /var/www/html/pico/content  
BaseThemeDir = /var/www/html/pico/themes/default  
AppDir = /where/orindi/is/installed/to

After that you can configure any number of feeds and the criteria that they match on. You can put them all in one file, or in separate files in $HOME/.config/orindi. For each feed, there are four elements.

Example:

[feed5]
keyword = crowdfunding
from = @indiegogo.com
    bingo@patreon.com
    no-reply@patreon.com
    messages@gofundme.com
    gfm-community@gofundme.com
    hello@gofundme.com
    news@gofundme.com
    no-reply@kickstarter.com    
subject = Thanks for becoming a backer,Project Update,launched

Any mail that orindi gets that matches either the “from” condition or the “subject” condition will be put into the “crowdfunding” subdirectory and “crowdfunding” feed.

Security

There is no default security set up with orindi. This is not ideal, for obvious reasons.

The easiest way is to use htaccess basic authentication. Instructions on setting up htaccess for Apache and Nginx are available on the web (or the links in this sentence).

Once you’ve set up htaccess authentication, the generated RSS feeds are accessible at:

https://USERNAME:PASSWORD@domain.for.your.site/feed-link

5. Usage

orindi takes an email from either stdin or as a filename from the first command line variable. This provides additional flexibility in inserting it into your already existing workflow.

It even works for already existent mail that is in the MAILDIR format. For example, if you were to use offlineimap and getmail to pull down your email to a single directory, you could then use this bit of bash in a script to process the files:

for i in `find $MAILDIR -type f \( -iname "*.*" ! -iname "dovecot*" ! -name "subscriptions" \) `; do cat "$i" | procmail $PROCMAILD/orindi_procmailrc done

or just as validly, use this bit of bash to do the same thing:

for i in `find $MAILDIR -type f \( -iname "*.*" ! -iname "dovecot*" ! -name "subscriptions" \) `; do $PATH/TO/orindi_parse.py "$i" done

(If you wished to remove the files, obviously just add rm "$i" to the loop.)

A warning with the latter; the filenames that dovecot uses (at least) will sometimes completely pooch the process.

Or you could even use the same to process a different mailbox after some other program has had a crack at it.

Cleaning up

When processing, orindi will set the outputted file to the same date/time as the mail header, which will help with maintenance and housekeeping.

The easiest way to clean up old files is through using a script like this, substituting your pico content directory where appropriate:

for i in `find /var/www/html/pico/content/ -type d | grep -v -e "content/$" | xargs realpath`; do 
    find "$i" -mtime +30 | xargs rm -f
    find "$i" -atime +30 | xargs rm -f
done

The reverse grep is needed so you don’t accidentally delete your subject index files in the main content directory. Of course, if you use pico for other sites, you will want to edit the script appropriately.

Viewing the output

If pico is installed in the subdirectory /pico to your domain, then you can view it at yourdomain.com/pico. Each keyword’s output is viewable at yourdomain/pico/?keyword and the RSS feed is available at yourdomain/pico/?keyword-feed.

To use the example above, that means I could see an index of the emails at yourdomain/pico/?crowdfunding and get the RSS feed at yourdomain/pico/?crowdfunding-feed.

MOST emails look pretty much how they’re supposed to. Linkedin emails seem to get mangled to hell. Some extra spaces and linefeeds, but I’m not sure I want to keep trying to parse them in the main program…

6. TODO

Roadmap:

Steven Saus injects people with radioactivity for his day job, but only to serve the forces of good.
Mostly.