Adding a message sender as a contact

[pigmonkey]

This is the last issue that is preventing me from fully migrating from abook to ppl!

abook has an --add-email argument. According to the man page:

--add-email
          Read an e-mail message from stdin and add the sender to the addressbook.

In my ~/.muttrc I use it like this:

# Add sender to address book
macro index,pager a "<pipe-message>abook --add-email<return>"

When I'm on a message in Mutt I hit the a key. I'm then prompted if I want to add the sender to abook.

Add "John Doe <john@doe.com>" to /home/user/.addressbook? (y/n)

I'm prompted like this for every sender found in the message. This includes the sender set in the header, but if it is a forwarded message it also includes the sender of the original message.

If I hit 'y', a new contact is added to my address book with a name of "John Doe" and an email of "john@doe.com".

If the message did not have a proper From: header set with the sender's name, the prompt would look like this:

Add "john@doe.com <john@doe.com>" to /home/user/.addressbook? (y/n)

If I selected 'y' the new contact would then have a name of "john@doe.com".

Adding this functionality would expand the scope of ppl from just managing contacts to also scraping messages. I understand if you do not want to go down that road. It may make more sense just to write a small, separate script that does the scraping and passes the results to ppl.

[henrycatalinismith]

The way I'd probably achieve this is by adding something like the mail gem as a dependency and writing something along these lines:

email = Mail.new($stdin)
contact = Ppl::Entity::Contact.new
contact.email_addresses << email.sender.address

So while ppl would nominally be scraping messages, this is functionality we can have without any real scope creep in terms of adding lots of "clever" code to ppl.

I like this idea. I think I'd probably use mutt more often if I had more helpful stuff like this set up, so this has caught my imagination a little. I think the way I'll go with this is to implement it in the form of some sort of emailscrape (definitely not its final name) command, with flags like --sender and --recipients for indicating which types of email addresses are to be scraped into the address book. That way it has room to grow into other use-cases, such as scraping all email addresses from the headers and body of an email which I think is what GMail does.

And since this is partly about abook feature parity, this will probably eventually need a --quiet flag as a counterpart to abook's --add-email-quiet. But that can wait.

[henrycatalinismith]

This is in 1.19.0 as the scrape command