emayili

emayili is a package for sending emails from R. The design goals are:

  • works on all manner of SMTP servers and
  • has minimal dependencies (or dependencies which are easily satisfied).

The package name is an adaption of the Zulu word for email, imeyili.

The documentation for emayili can be found here.

Installation

Get the stable version from CRAN.

install.packages("emayili")

Or grab it directly from GitHub.

# Install from the master branch.
remotes::install_github("datawookie/emayili")
# Install from the development branch.
remotes::install_github("datawookie/emayili", ref = "dev")

Usage

Load the library.

[1] '0.7.17'

Create a message object.

email <- envelope()

Creating a Message

The message has class envelope.

class(email)
[1] "envelope"

Add addresses for the sender and recipient.

email <- email %>%
  from("alice@yahoo.com") %>%
  to("bob@google.com") %>%
  cc("craig@google.com")

There are also bcc() and reply() functions for setting the Bcc and Reply-To fields.

You can supply multiple addresses in a variety of formats:

  • as a single comma-separated string
  • as separate strings; or
  • as a vector of strings.
envelope() %>% to("bob@google.com, craig@google.com, erin@gmail.com")
envelope() %>% to("bob@google.com", "craig@google.com", "erin@gmail.com")
envelope() %>% to(c("bob@google.com", "craig@google.com", "erin@gmail.com"))

Add a subject.

email <- email %>% subject("This is a plain text message!")

Add a text body.

email <- email %>% text("Hello!")

You can use html() to add an HTML body. It accepts either a vector of characters or a tagList() from htmltools.

library(htmltools)

email <- email %>% html(
  tagList(
    h2("Hello"),
    p("World!")
  )
)

Add an attachment.

email <- email %>% attachment("image.jpg")

You can also create the message in a single command:

email <- envelope(
  to = "bob@google.com",
  from = "alice@yahoo.com",
  subject = "This is a plain text message!",
  text = "Hello!"
)

Simply printing a message displays the header information.

email
Date:                         Sat, 19 Aug 2023 04:44:05 GMT
X-Mailer:                     {emayili}-0.7.17
MIME-Version:                 1.0
From:                         alice@yahoo.com
To:                           bob@google.com
Cc:                           craig@google.com
Subject:                      This is a plain text message!

You can identify emails which have been sent using emayili by the presence of an X-Mailer header which includes both the package name and version.

If you want to see the complete MIME object, just convert to a string.

You can also call the print() method and specify details = TRUE.

Options

You can set the envelope.details option to assert that the details should always be printed.

# Always print envelope details.
#
options(envelope.details = TRUE)

By default the results returned by most of the methods are invisible. You can make them visible via the envelope.invisible (default: TRUE).

# Always show envelope.
#
options(envelope.invisible = FALSE)

Interpolating Text

You can use glue syntax to interpolate content into the body of a message.

name <- "Alice"

envelope() %>%
  text("Hello {{name}}!")
Date:                         Sat, 19 Aug 2023 04:44:05 GMT
X-Mailer:                     {emayili}-0.7.17
MIME-Version:                 1.0
Content-Type:                 text/plain; 
                              charset=utf-8; 
                              format=flowed
Content-Transfer-Encoding:    7bit
Content-MD5:                  nhjeY5ZYMzru+kSCGUzNKg==

Hello Alice!

Rendering Markdown

You can render Markdown straight into a message.

Use either plain Markdown.

envelope() %>%
  # Render plain Markdown from a character vector.
  render(
    "Check out [`{emayili}`](https://cran.r-project.org/package=emayili)."
  )
Date:                         Sat, 19 Aug 2023 04:44:05 GMT
X-Mailer:                     {emayili}-0.7.17
MIME-Version:                 1.0
Content-Type:                 text/html; 
                              charset=utf-8

<html><body><p>Check out <a href="https://cran.r-project.org/package=emayili"><code>{emayili}</code></a>.</p></body></html>

Or R Markdown.

envelope() %>%
  # Render R Markdown from a file.
  render("message.Rmd")

In both cases the function will accept either a file path or a character vector containing Markdown text.

Interpolation also works with render().

Rendered CSS

When you render an R Markdown document the resulting HTML includes CSS from three sources:

You can control which of these propagate to the message using the include_css parameter which, by default, is set to c("rmd", "bootstrap", "highlight").

🚨 Note: Gmail doesn’t like the Bootstrap CSS. If you want your styling to work on Gmail you should set include_css = c("rmd", "highlight").

Extra CSS

You can insert extra CSS into your rendered messages.

envelope() %>%
  render("message.Rmd", css_files = "extra.css")

If you are having trouble getting this to work with Gmail then it might be worthwhile taking a look at their CSS support.

Adding an Inline Image

Adding an inline image to an HTML message is possible. There are two ways to achieve this.

1. Base64 Encoding

First you’ll need to Base64 encode the image.

img_base64 <- base64enc::base64encode("image.jpg")

Then create the HTML message body.

html_body <- sprintf('<html><body><img src="data:image/jpeg;base64,%s"></body></html>', img_base64)

And finally add it to the email.

email <- envelope() %>% html(html_body)

Note: It’s important that you specify the appropriate media type (image/jpeg for JPEG and image/png for PNG).

2. Using a CID

Unfortunately some mail clients (like Gmail) will not display Base64 encoded images. In this case using a CID is a working alternative.

First create the message body which references an image by CID.

html_body <- '<html><body><img src="cid:image"></body></html>'

Then attach the image and specify the cid argument.

email <- envelope() %>%
  html(html_body) %>%
  attachment(path = "image.jpg", cid = "image")

Create a Server Object

Create a SMTP server object.

smtp <- server(
  host = "smtp.gmail.com",
  port = 465,
  username = "bob@gmail.com",
  password = "bd40ef6d4a9413de9c1318a65cbae5d7"
)

It’s bad practice to include credentials in a script. A better approach would be to keep the credentials in your .Renviron file.

GMAIL_USERNAME="bob@gmail.com"
GMAIL_PASSWORD="bd40ef6d4a9413de9c1318a65cbae5d7"

You can then pull these variables into R using Sys.getenv() and then create the server object.

smtp <- server(
  host = "smtp.gmail.com",
  port = 465,
  username = Sys.getenv("GMAIL_USERNAME"),
  password = Sys.getenv("GMAIL_PASSWORD")
)

If you’re trying to send email with a host that uses the STARTTLS security protocol (like Gmail, Yahoo! or AOL), then it will most probably be blocked due to insufficient security. In order to circumvent this, you can grant access to less secure apps. See the links below for specifics:

Gmail has recently changed their authentication procedure. If you are having trouble connecting to the Gmail SMTP server then take a look at this.

Sending a Message

Send the message.

smtp(email, verbose = TRUE)

To see the guts of the message as passed to the SMTP server:

print(email, details = TRUE)

Encryption

Both transport-level encryption and end-to-end email encryption are supported. An SMTP connection can be initiated on port 465 (SMTPS, see RFC 8314) or 587 (SMTP-MSA, see RFC 6409), enabling transport-level encryption. Public-key cryptography (via GnuPG) can be used to sign and/or encrypt message contents for end-to-end encryption.

Standards Documents

The following (draft) standards documents relate to emails:

Similar Packages

There is a selection of other R packages which also send emails:

Blog Posts

Support for Gmail, SendGrid & Mailgun
Message Precedence
Message Integrity
Right-to-Left
Styling Figures
Managing CSS
R Markdown Parameters
Rendering R Markdown
Rendering Plain Markdown
Interpolating Message Content
Rudimentary Email Address Validation
Updated Gmail Authentication

Developer Notes

Code Coverage

You can find the test coverage report at Codecov. For development purposes it’s more convenient to use the {covr} package.

Generate a coverage report.

library(covr)

# Tests that are skipped on CRAN should still be included in coverage report.
#
Sys.setenv(NOT_CRAN = "true")

report()

Calculate test coverage.

coverage <- package_coverage()

Coverage statistics as a data frame.

as.data.frame(coverage)

Show lines without coverage.

zero_coverage(coverage)

Checks

Check spelling.

spelling::spell_check_package()

Quick local checks.

devtools::check()

Remote checks (take longer but more thorough).

devtools::check_win_devel()

# Check for a specific platform.
#
rhub::check(platform = "debian-gcc-devel")
rhub::check_on_windows(check_args = "--force-multiarch")
rhub::check_on_solaris()

# Check on a bunch of platforms.
#
rhub::check_for_cran()

# Check on important platforms.
#
rhub::check_for_cran(platforms = c(
  "debian-gcc-release",
  "ubuntu-gcc-release",
  "macos-m1-bigsur-release",
  "windows-x86_64-release",
  NULL
))