Now uses a customized bootstrap that looks something like Jellyfin. Some
small ui changes were needed. This be overridden by downloading bs5's css and using the custom_css
option if you don't like it. sass file is included for your own modification. Changes made to master have been added also.
live reloading was intended for previous release, but needed some
tweaking. Settings that still require a restart are marked with an R.
Fixed issue where default values weren't being filled in on reload of
config that broke emails if settings were changed at all.
Currently all setting changes require a restart to apply, so there's a
bit of commented out code that i implemented before i realized.
Still needs tooltips for each setting.
Added option email/no_username to disable username input on form, and
instead use the provided email address as the username. Also added
missing 'packaging' dep from pevious update.
Automatically fixes '*ProviderId' in user templates for versions of
Jellyfin >= 10.6.0. Removed unnecessary configparser dependency, the one
actually used is part of python. jf_api now has an info attribute.
The dependency on 'secrets' has been removed, as I realized that the
cryptographic 'secrets' package is included in python, and the one on
PyPI was related to LDAP. This the issues with python-ldap on some
systems no longer exist.
setup.py has been removed, and Poetry is now used to package for
install. This fixed an issue i had with uploading to PyPI, so one can
now run 'pip install jellyfin-accounts' to install.
user_template and other files are now accessed via JSONStorage, which
has dictionary like attributes for each file which can be used like a
dictionary, without the need to manually read and write the file. This
was done so that other storage types (e.g a database) can be added in
future.
--get_policy is now --get_defaults, as it now allows you to store a
default user configuration and displayPreferences, which define the
layout of the home screen. It can also now display non publicly visible
accounts.
Originally i loaded the requirements through calling the web api in
javascript, but decided to switch it to jinja as i only recently found
out its capable of stuff more complicated than replacing strings.
