18.1.7. email.encoders: Encoders

When creating Message objects from scratch, you often need to encode the payloads for transport through compliant mail servers. This is especially true for image/* and text/* type messages containing binary data.

The email package provides some convenient encodings in its encoders module. These encoders are actually used by the MIMEAudio and MIMEImage class constructors to provide default encodings. All encoder functions take exactly one argument, the message object to encode. They usually extract the payload, encode it, and reset the payload to this newly encoded value. They should also set the Content-Transfer-Encoding header as appropriate.

Note that these functions are not meaningful for a multipart message. They must be applied to individual subparts instead, and will raise a TypeError if passed a message whose type is multipart.

Here are the encoding functions provided:


Encodes the payload into quoted-printable form and sets the Content-Transfer-Encoding header to quoted-printable [1]. This is a good encoding to use when most of your payload is normal printable data, but contains a few unprintable characters.


Encodes the payload into base64 form and sets the Content-Transfer-Encoding header to base64. This is a good encoding to use when most of your payload is unprintable data since it is a more compact form than quoted-printable. The drawback of base64 encoding is that it renders the text non-human readable.


This doesn’t actually modify the message’s payload, but it does set the Content-Transfer-Encoding header to either 7bit or 8bit as appropriate, based on the payload data.


This does nothing; it doesn’t even set the Content-Transfer-Encoding header.


[1]Note that encoding with encode_quopri() also encodes all tabs and space characters in the data.

Previous topic

18.1.6. email.charset: Representing character sets

Next topic

18.1.8. email.errors: Exception and Defect classes

This Page