Pigeon is a command line tool for automating your email workflow in a cheap and efficient way. Utilize your most efficient dev tools you are already familiar with.
For example, query the subscribers of your newsletter, create a plaintext and html email from a template file, and send it to all of them:
pigeon send-bulk \
sender@your-domain.com \
--receiver-query "select email from user where newsletter_confirmed = true" \
--message-file "message.yaml" \
--display \
--assume-yes> Display query result: shape: (4, 1)
+------------------------------+
| email |
| --- |
| str |
+==============================+
| "marie@curie.com" |
+------------------------------+
| "alexandre@grothendieck.com" |
+------------------------------+
| "emmy@noether.com" |
+------------------------------+
| "elie@cartan.com" |
+------------------------------+
> Sending email to 4 receivers ...
marie@curie.com ... ok
alexandre@grothendieck.com ... ok
emmy@noether.com ... ok
elie@cartan.com ... ok- Install Pigeon
- Getting help
- Usage
- How to connect
- Integrations
- Comparison with Mailchimp, Sendgrid, and ConvertKit
- Testing
Install Pigeon from crates.io
# Build and install pigeon binary to ~/.cargo/bin
cargo install pigeon-rsNote: Run cargo install pigeon-rs again to update to the latest version. Uninstall the pigeon binary with cargo uninstall pigeon-rs.
Install Pigeon from github.com
# Clone repository
git clone git@github.com:quambene/pigeon-rs.git
cd pigeon-rs
# Build and install pigeon binary to ~/.cargo/bin
cargo install --path .Note: Add $HOME/.cargo/bin to your PATH if it is missing:
export PATH="$HOME/.cargo/bin:$PATH"For getting help, try one of the following:
# Check version
pigeon --version
# Print help
pigeon --help
# Print help for subcommand
pigeon help send
pigeon help send-bulk
pigeon help connect
pigeon help init
pigeon help query
pigeon help simple-query
pigeon help readCheck connection to your SMTP server with pigeon connect:
pigeon connectConnecting to SMTP server 'email-smtp.eu-west-1.amazonaws.com' ... ok
See currently supported integrations and how to connect below.
Note: You can also check connection to third-party APIs instead of using the SMTP protocol. For example, using AWS Simple Email Service (SES): pigeon connect aws.
Send a single email with subject and content:
pigeon send \
sender@your-domain.com \
receiver@gmail.com \
--subject "Test subject" \
--content "This is a test email."Send a single email with message defined in separate template file:
pigeon send \
sender@your-domain.com \
receiver@gmail.com \
--message-file "message.yaml"The message template message.yaml is created with subcommand init:
pigeon initNote: One of the advantages of a --message-file is that you can also draft the html version of your email. In contrast, with the options --subject and --content the email will only be sent in plaintext format.
If you prefer a dedicated HTML file for drafting your email, use the following command:
pigeon send \
sender@your-domain.com \
receiver@gmail.com \
--subject "Test subject" \
--text-file "./message.txt" \
--html-file "./message.html"where --text-file defines the plaintext and --html-file the HTML version of your email.
For example, query relevant users which confirmed to receive your newsletter, and send an email to all of them.
Let's check the query first via pigeon query:
pigeon query --display "select email from user where newsletter_confirmed = true"> Display query result: shape: (4, 1)
+------------------------------+
| email |
| --- |
| str |
+==============================+
| "marie@curie.com" |
+------------------------------+
| "alexandre@grothendieck.com" |
+------------------------------+
| "emmy@noether.com" |
+------------------------------+
| "elie@cartan.com" |
+------------------------------+See how to connect below to connect your database.
Note: You can also --save your query as a csv file: pigeon query --save <my-query>.
Now send your newsletter to the queried receivers. If the table column name is different to "email" use --receiver-column to define a different column name. Let's try a --dry-run without confirmation --assume-yes first:
pigeon send-bulk \
albert@einstein.com \
--receiver-query "select email from user where newsletter_confirmed = true" \
--message-file "message.yaml" \
--assume-yes \
--dry-run> Sending email to 4 receivers ...
marie@curie.com ... dry run
alexandre@grothendieck.com ... dry run
emmy@noether.com ... dry run
elie@cartan.com ... dry runAfter double checking, you can submit the same command without --dry-run. Remove --assume-yes as well for explicit confirmation.
Note: You can also send a bulk email to email adresses defined in a csv file instead of a query result. In this case, use option --receiver-file instead of --receiver-query. You can check the contents of a csv file via subcommand read, e.g. pigeon read recipients.csv.
If you need more individual emails, you can personalize your emails with option --personalize. Again, let's start by checking the relevant query:
pigeon query --display "select first_name, last_name, email from user where newsletter_confirmed = true"> Display query result: shape: (4, 3)
+-------------+----------------+------------------------------+
| first_name | last_name | email |
| --- | --- | --- |
| str | str | str |
+=============+================+==============================+
| "Marie" | "Curie" | "marie@curie.com" |
+-------------+----------------+------------------------------+
| "Alexandre" | "Grothendieck" | "alexandre@grothendieck.com" |
+-------------+----------------+------------------------------+
| "Emmy" | "Noether" | "emmy@noether.com" |
+-------------+----------------+------------------------------+
| "Elie" | "Cartan" | "elie@cartan.com" |
+-------------+----------------+------------------------------+In your message template message.yaml use variables in curly brackets, like {first_name} and {last_name}. Then define personalized colums as parameters for option --personalize. Finally, let's display everything with --display:
pigeon send-bulk \
albert@einstein.com \
--receiver-query "select first_name, last_name, email from user where newsletter_confirmed = true" \
--message-file "message.yaml" \
--personalize "first_name" "last_name" \
--display> Display message file: MessageTemplate {
message: Message {
subject: "Issue No. 1",
text: "Dear {first_name} {last_name},
Welcome to my newsletter. We are doing hard sciences here.
Sincerely, Albert Einstein",
html: "Dear {first_name} {last_name},
Welcome to my newsletter. We are doing hard sciences here.
Sincerely, Albert Einstein",
},
}
> Display emails: BulkEmail {
emails: [
Email {
sender: "albert@einstein.com",
receiver: "marie@curie.com",
message: Message {
subject: "Issue No. 1",
text: "Dear Marie Curie,
Welcome to my newsletter. We are doing hard sciences here.
Sincerely, Albert Einstein",
html: "Dear Marie Curie,
Welcome to my newsletter. We are doing hard sciences here.
Sincerely, Albert Einstein",
},
},
Email {
sender: "albert@einstein.com",
receiver: "alexandre@grothendieck.com",
message: Message {
subject: "Issue No. 1",
text: "Dear Alexandre Grothendieck,
Welcome to my newsletter. We are doing hard sciences here.
Sincerely, Albert Einstein",
html: "Dear Alexandre Grothendieck,
Welcome to my newsletter. We are doing hard sciences here.
Sincerely, Albert Einstein",
},
},
Email {
sender: "albert@einstein.com",
receiver: "emmy@noether.com",
message: Message {
subject: "Issue No. 1",
text: "Dear Emmy Noether,
Welcome to my newsletter. We are doing hard sciences here.
Sincerely, Albert Einstein",
html: "Dear Emmy Noether,
Welcome to my newsletter. We are doing hard sciences here.
Sincerely, Albert Einstein",
},
},
Email {
sender: "albert@einstein.com",
receiver: "elie@cartan.com",
message: Message {
subject: "Issue No. 1",
text: "Dear Elie Cartan,
Welcome to my newsletter. We are doing hard sciences here.
Sincerely, Albert Einstein",
html: "Dear Elie Cartan,
Welcome to my newsletter. We are doing hard sciences here.
Sincerely, Albert Einstein",
},
},
],
}
> Should an email be sent to 4 recipients? Yes (y) or no (n)
>Confirm y if you are ready to go.
To connect to a SMTP server, define environment variables SMTP_SERVER, SMTP_USERNAME, and SMTP_PASSWORD. For example, using AWS SES:
SMTP_SERVER=email-smtp.eu-west-1.amazonaws.com
SMTP_USERNAME=...
SMTP_PASSWORD=...Source your environment .env in your current shell:
set -a && source .env && set +aInstead of using SMTP, you can send emails via the API of a specific email provider as well.
Using AWS SES, define the following environment variables:
AWS_ACCESS_KEY_ID=...
AWS_SECRET_ACCESS_KEY=...
AWS_REGION=eu-west-1where AWS_REGION depends on the specified region for your AWS SES account.
Source your environment again:
set -a && source .env && set +aSend an email using --connection:
pigeon send \
sender@your-domain.com \
receiver@gmail.com \
--connection aws \
--message-file "message.yaml"For postgres, the database url is constructed as follows: postgresql://db_user:db_password@db_host:db_port/db_name.
Therefore, set the following environment variables in your environment .env:
DB_HOSTDB_PORTDB_USERDB_PASSWORDDB_NAME
Source your environment again:
set -a && source .env && set +aCAUTION: Connecting via TLS is not supported yet. Forward a local port through a SSH tunnel instead, e.g.:
pigeon query "select email from user where newsletter_confirmed = true" --display --ssh-tunnel 5437In addition to the environment variables above, SERVER_USER and SERVER_HOST have to be set for the SSH connection (ssh user@host).
- MIME
- SMTP
- AWS SES
- PostgreSQL
- CSV
These numbers may be outdated. Do your own research.
The following table compares the price per month for email provider and emails per month.
| 5,000 | 10,000 | 100,000 | |
|---|---|---|---|
| Pigeon+AWS | $4.50 | $5 | $14 |
| Mailchimp Marketing | $9.99 | $20.99 | $78.99 |
| Mailchimp Transactional | - | - | $80 |
| Sendgrid Marketing | $15 | $15 | $120 |
| Sendgrid API | $14.95 | $14.95 | $29.95 |
| ConvertKit | $66 | $100 | $516 |
The following table shows the daily limit for sent emails per provider.
| provider | daily limit |
|---|---|
| Pigeon+AWS | 50,000 |
| Mailchimp | equals monthly limit |
| Sendgrid | equals monthly limit |
Some integration tests require a locally running postgres database:
- Specify the following environment variables:
DB_HOSTDB_PORTDB_USERDB_PASSWORDDB_NAME
- Set up a temporary postgres db:
docker-compose run --rm --service-ports postgres
# Run unit tests and integration tests
cargo test
# Run unit tests
cargo test --lib
# Run integration tests
cargo test --test '*'