Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/ConradIrwin/em-imap
An event machine based IMAP client
https://github.com/ConradIrwin/em-imap
Last synced: 3 months ago
JSON representation
An event machine based IMAP client
- Host: GitHub
- URL: https://github.com/ConradIrwin/em-imap
- Owner: ConradIrwin
- License: mit
- Created: 2011-05-02T11:03:57.000Z (over 13 years ago)
- Default Branch: master
- Last Pushed: 2020-09-18T05:12:10.000Z (about 4 years ago)
- Last Synced: 2024-05-31T20:01:10.488Z (5 months ago)
- Language: Ruby
- Homepage:
- Size: 130 KB
- Stars: 69
- Watchers: 11
- Forks: 31
- Open Issues: 10
-
Metadata Files:
- Readme: README.md
- License: LICENSE.MIT
Awesome Lists containing this project
README
## ⚠️⚠️⚠️ Insecure and unmaintained! ⚠️⚠️⚠️
This gem should not be used, it is vulnerable to SSL man in the middle attacks as documented here: https://github.com/ConradIrwin/em-imap/issues/25. Pull requests for fixes are most welcome!
----
An [EventMachine](http://rubyeventmachine.com/) based [IMAP](http://tools.ietf.org/html/rfc3501) client.
## Installation
gem install em-imap
## Usage
This document tries to introduce concepts of IMAP alongside the facilities of the library that handle them, to give you an idea of how to perform basic IMAP operations. IMAP is more fully explained in [RFC3501](http://tools.ietf.org/html/rfc3501), and the details of the library are of course in the source code.
### Connecting
Before you can communicate with an IMAP server, you must first connect to it. There are three connection parameters, the hostname, the port number, and whether to use SSL/TLS. As with every method in EM::IMAP, `EM::IMAP::Client#connect` returns a [deferrable](http://eventmachine.rubyforge.org/docs/DEFERRABLES.html) enhanced by the [deferrable\_gratification](https://github.com/samstokes/deferrable_gratification) library.
For example, to connect to Gmail's IMAP server, you can use the following snippet:
```ruby
require 'rubygems'
require 'em-imap'EM::run do
client = EM::IMAP.new('imap.gmail.com', 993, true)
client.connect.errback do |error|
puts "Connecting failed: #{error}"
end.callback do |hello_response|
puts "Connecting succeeded!"
end.bothback do
EM::stop
end
end
```### Authenticating
There are two authentication mechanisms in IMAP, `LOGIN` and `AUTHENTICATE`, exposed as two methods on the EM::IMAP client, `.login(username, password)` and `.authenticate(mechanism, *args)`. Again these methods both return deferrables, and the cleanest way to tie deferrables together is to use the [`.bind!`](http://samstokes.github.com/deferrable_gratification/doc/DeferrableGratification/Combinators.html#bind!-instance_method) method from deferrable\_gratification.
Extending our previous example to also log in to Gmail:
```ruby
client = EM::IMAP.new('imap.gmail.com', 993, true)
client.connect.bind! do
client.login("[email protected]", ENV["GMAIL_PASSWORD"])
end.callback do
puts "Connected and logged in!"
end.errback do |error|
puts "Connecting or logging in failed: #{error}"
end
```The `.authenticate` method is more advanced and uses the same extensible mechanism as [Net::IMAP](http://www.ruby-doc.org/stdlib/libdoc/net/imap/rdoc/classes/Net/IMAP.html). The two mechanisms supported by default are `'LOGIN'` and [`'CRAM-MD5'`](http://www.ietf.org/rfc/rfc2195.txt), other mechanisms are provided by gems like [gmail\_xoauth](https://github.com/nfo/gmail_xoauth).
### Mailbox-level IMAP
Once the authentication has completed successfully, you can perform IMAP commands that don't require a currently selected mailbox. For example to get a list of the names of all Gmail mailboxes (including labels):
```ruby
client = EM::IMAP.new('imap.gmail.com', 993, true)
client.connect.bind! do
client.login("[email protected]", ENV["GMAIL_PASSWORD"])
end.bind! do
client.list
end.callback do |list|
puts list.map(&:name)
end.errback do |error|
puts "Connecting, logging in or listing failed: #{error}"
end
```The useful commands available to you at this point are `.list`, `.create(mailbox)`, `.delete(mailbox)`, `.rename(old_mailbox, new_mailbox)`, `.status(mailbox)`. `.select(mailbox)` and `.examine(mailbox)` are discussed in the next section, and `.subscribe(mailbox)`, `.unsubscribe(mailbox)`, `.lsub` and `.append(mailbox, message, flags?, date_time)` are unlikely to be useful to you immediately. For a full list of IMAP commands, and detailed considerations, please refer to [RFC3501](http://tools.ietf.org/html/rfc3501).
### Message-level IMAP
In order to do useful things which actual messages, you need to first select a mailbox to interact with. There are two commands for doing this, `.select(mailbox)`, and `.examine(mailbox)`. They are the same except that `.examine` opens a mailbox in read-only mode; so that no changes are made (i.e. performing commands doesn't mark emails as read).
For example to search for all emails relevant to em-imap in Gmail:
```ruby
client = EM::IMAP.new('imap.gmail.com', 993, true)
client.connect.bind! do
client.login("[email protected]", ENV["GMAIL_PASSWORD"])
end.bind! do
client.select('[Google Mail]/All Mail')
end.bind! do
client.search('ALL', 'SUBJECT', 'em-imap')
end.callback do |results|
puts results
end.errback do |error|
puts "Something failed: #{error}"
end
```Once you have a list of message sequence numbers, as returned by search, you can actually read the emails with `.fetch`:
```ruby
client = EM::IMAP.new('imap.gmail.com', 993, true)
client.connect.bind! do
client.login("[email protected]", ENV["GMAIL_PASSWORD"])
end.bind! do
client.select('[Google Mail]/All Mail')
end.bind! do
client.search('ALL', 'SUBJECT', 'em-imap')
end.bind! do |results|
client.fetch(results, 'BODY[TEXT]')
end.callback do |emails|
puts emails.map{|email| email.attr['BODY[TEXT]'] }
end.errback do |error|
puts "Something failed: #{error}"
end
```The useful commands available to you at this point are `.search(*args)`, `.expunge`, `.fetch(messages, attributes)`, `.store(messages, name, values)` and `.copy(messages, mailbox)`. If you'd like to work with UIDs instead of sequence numbers, there are UID based alternatives: `.uid_search`, `.uid_fetch`, `.uid_store` and `.uid_copy`. The `.close` command and `.check` command are unlikely to be useful to you immediately.
### Untagged responses
IMAP has the notion of untagged responses (aka. unsolicited responses). The idea is that sometimes when you run a command you'd like to be updated on the state of the mailbox with which you are interacting, even though notification isn't always required. To listen for these responses, the deferrables returned by each client method have a `.listen(&block)` method. All responses received by the server, up to and including the response that completes the current command will be passed to your block.
For example, we could insert a listener into the above example to find out some interesting numbers:
```ruby
end.bind! do
client.select('[Google Mail]/All Mail').listen do |response|
case response.name
when "EXISTS"
puts "There are #{response.data} total emails in All Mail"
when "RECENT"
puts "There are #{response.data} new emails in All Mail"
end
end
end.bind! do
```### Concurrency
IMAP is an explicitly concurrent protocol: clients MAY send commands without waiting for the previous command to complete, and servers MAY send any untagged response at any time.
If you want to receive server responses at any time, you can call `.add_response_handler(&block)` on the client. This returns a deferrable like the IDLE command, on which you can call `stop` to stop receiving responses (which will cause the deferrable to succeed). You should also listen on the `errback` of this deferrable so that you know when the connection is closed:
```ruby
handler = client.add_response_handler do |response|
puts "Server says: #{response}"
end.errback do |e|
puts "Connection closed?: #{e}"
end
EM::Timer.new(600){ handler.stop }
```If you want to send commands without waiting for previous replies, you can also do so. em-imap handles the few cases where this is not permitted (for example, during an IDLE command) by queueing the command until the connection becomes available again. If you do this, bear in mind that any blocks that are listening on the connection may receive responses from multiple commands interleaved.
```ruby
client = EM::Imap.new('imap.gmail.com', 993, true)
client.connect.callback do
logger_in = client.login('[email protected]', ENV["GMAIL_PASSWORD"])
selecter = client.select('[Google Mail]/All Mail')
searcher = client.search('from:[email protected]').callback do |results|
puts results
endlogger_in.errback{ |e| selecter.fail e }
selecter.errback{ |e| searcher.fail e }
searcher.errback{ |e| "Something failed: #{e}" }
end
```### IDLE
IMAP has an IDLE command (aka push-email) that lets the server notify the client when there are new emails to be read. This command is exposed at a low-level, but it's quite hard to use directly. Instead you can simply ask the client to `wait_for_new_emails(&block)`. This takes care of re-issuing the IDLE command every 29 minutes, in addition to ensuring that the connection isn't IDLEing while you're trying to process the results.
```ruby
client = EM::IMAP.new('imap.gmail.com', 993, true)
client.connect.bind! do
client.login("[email protected]", ENV["GMAIL_PASSWORD"])
end.bind! do
client.select('INBOX')
end.bind! doclient.wait_for_new_emails do |response|
client.fetch(response.data).callback{ |fetched| puts fetched.inspect }
endend.errback do |error|
puts "Something failed: #{error}"
end
```The block you pass to `wait_for_new_emails` should return a deferrable. If that deferrable succeeds then the IDLE loop will continue, if that deferrable fails then the IDLE loop will also fail. If you don't return a deferrable, it will be assumed that you didn't want to handle the incoming email, and IDLEing will be immediately resumed.
## TODO
em-imap is still very much a work-in-progress, and the API will change as time goes by.
Before version 1, at least the following changes should be made:
1. Stop using Net::IMAP in quite so many bizarre ways, probably clearer to copy-paste the code and rename relevant classes (particular NoResponseError..)
2. Find a nicer API for some commands (maybe some objects to represent mailboxes, and/or messages?)
3. Document argument serialization.
4. Support SORT and THREAD.
5. Put the in-line documentation into a real format.### Breaking Changes
Between Version 0.1(.x) and 0.2, the connection setup API changed. Previously you would call `EM::IMAP.connect`, now that is broken into two steps: `EM::IMAP.new` and `EM::IMAP::Client#connect` as documented above. This makes it less likely people will write `client = connect.bind!` by accident, and allows you to bind to the `errback` of the connection as a whole should you wish to.
## Meta-foo
Em-imap is made available under the MIT license, see LICENSE.MIT for details
Patches and pull-requests are welcome.