* Remove "pre-1.0" warnings in docs, readme
* Jump trove classifer from pre-alpha all the way up to stable
(arguably this should have been "beta" for the past several months)
* The default (GitHub) readme should point
to the stable docs version, rather than
the latest development version.
* The frozen links in PyPI should use the full
patch version number (X.Y.Z), not just the minor
X.Y version. (Leftover from Djrill's branch-based
version management; Anymail uses tags
for versions, and old way was creating incorrect
frozen doc links for patch releases.)
* **Future breaking change:**
Rename all Anymail backends to just `EmailBackend`,
matching Django's naming convention.
(E.g., switch to "anymail.backends.mailgun.EmailBackend"
rather than "anymail.backends.mailgun.MailgunBackend".)
The old names still work, but will issue a DeprecationWarning
and will be removed in some future release.
(Apologies for this change; the old naming convention was
a holdover from Djrill, and I wanted consistency with
other Django EmailBackends before hitting 1.0.)
Fixes#49.
* Trailing comma after "anymail" (see #40)
* Note order doesn't matter
* Change tuple to list (match examples to
Django 1.9+ project template)
[ci skip]
Allow custom MAILGUN_SENDER_DOMAIN in Anymail
settings. (Replaces need to use global esp_extra.)
Improve docs to cover cases where this is needed.
(esp_extra sender_domain is still supported for
overriding individual messages.)
Fixes#26.
Also includes:
* Change AnymailTestMixin.assertDoesNotWarn
to filter specific warning classes.
* Look specifically for AnymailInsecureWebhookWarning
in WebhookBasicAuthTestsMixin.test_warns_if_no_auth
(because we don't care *in that test case* about
DeprecatedInDjango10 warnings).
* message.template_id to use ESP stored templates
* message.merge_data and merge_global_data
to supply per-recipient/global merge variables
(with or without an ESP stored template)
* When using per-recipient merge_data, tell ESP to use
batch send: individual message per "to" address.
(Mailgun does this automatically; SendGrid requires
using a different "to" field; Mandrill requires
`preserve_recipients=False`; Postmark doesn't
support *this type* of batch sending with merge data.)
* Allow message.from_email=None (must be set after
init) and message.subject=None to suppress those
fields in API calls (for ESPs that allow "From" and
"Subject" in their template definitions).
Mailgun:
* Emulate merge_global_data by copying to
recipient-variables for each recipient.
SendGrid:
* Add delimiters to merge field names via
esp_extra['merge_field_format'] or
ANYMAIL_SENDGRID_MERGE_FIELD_FORMAT setting.
Mandrill:
* Remove Djrill versions of these features;
update migration notes.
Closes#5.
Simplify install to just `pip install django-anymail`.
(Rather than `... django-anymail[mailgun]`
All of the ESPs so far require requests, so just move
that into the base requirements. (Chances are your
Django app already needs requests for some other
reason, anyway.)
Truly unique ESP dependencies (e.g., boto for
AWS-SES) could still use the setup extra features
mechanism.
* Add filename param to attach_inline_image
* Add attach_inline_image_file function
(parallels EmailMessage.attach and attach_file)
* Use `Content-Disposition: inline` to decide
whether an attachment should be handled inline
(whether or not it's an image, and whether or not
it has a Content-ID)
* Stop conflating filename and Content-ID, for
ESPs that allow both. (Solves problem where
Google Inbox was displaying inline images
as attachments when sent through SendGrid.)
* Don't maintain similar overview in README.rst and docs/index.rst -- instead just include relevant portions of readme in the docs
* Patch up README version numbers and doc links in setup.py long_description to freeze them to the version being setup
* Suppress the Travis build status indicator in the docs and PyPI, since it can't be frozen to the specific version in question
Treat image attachments with a Content-ID header as embedded, rather than ordinary attachments. (Rationale is that you must set the content-id to be able to refer to the image within your html, so that's a reasonable indicator to handle it as embedded.)
