[2.0.0Alpha1] encoding warning message [Ubuntu]

[nhasian]

after upgrading to 2.0.0Alpha1 I receive this warning message:

SABnzbd was started with encoding ANSI_X3.4-1968, this should be UTF-8. Expect problems with Unicoded file and directory names in downloads.

However when I check my config file it says:

__version__ = 19
__encoding__ = utf-8

I don't know if it's related but I just noticed on the general tab it says this:

Python Version: 2.7.12 (default, Nov 19 2016, 06:48:10) [GCC 5.4.0 20160609] [ANSI_X3.4-1968]

[sander]

after upgrading to 2.0.0Alpha1 I receive this warning message:

SABnzbd was started with encoding ANSI_X3.4-1968, this should be UTF-8. Expect problems with Unicoded file and directory names in downloads.

Which platform? Linux ...? NAS ....? ... ?


The encoding of your config file is just the encoding of your config file, NOT how SABnzbd is started.

[nhasian]

Oops sorry I am running it on an Ubuntu Linux Server. I use systemd to start it.

after upgrading to 2.0.0Alpha1 I receive this warning message:

SABnzbd was started with encoding ANSI_X3.4-1968, this should be UTF-8. Expect problems with Unicoded file and directory names in downloads.

Which platform? Linux ...? NAS ....? ... ?


The encoding of your config file is just the encoding of your config file, NOT how SABnzbd is started.

[sander]

I am running it on an Ubuntu Linux Server.

OK

I use systemd to start it.

Ah, I know little about systemd etc. Do you use the PPA?

First some introduction:

SABnzbd has introduced Unicode some versions ago. That works. However, you also need to start SABnzbd with the correct encoding / locale. SAB has been reporting for a long time:

Code: Select all

2017-03-04 12:14:49,738::INFO::[sabnzbdplus:1190] Preferred encoding = UTF-8

For 2.0.0, we've decided to warn Linux-users in case of non-UTF encoding because that causes problems in case of certain downloads.
And you're one of the first volunteers.

FWIW: I run 2.0.0Alpha1 on my Ubuntu 14.04 server too, and no problems.

It's about the "locale" that is used. Check with "locale". On my Ubuntu server:

Code: Select all

sander@haring:~$ locale
LANG=en_US.UTF-8
LANGUAGE=
LC_CTYPE="en_US.UTF-8"
LC_NUMERIC=nl_NL.UTF-8
LC_TIME=nl_NL.UTF-8
LC_COLLATE="en_US.UTF-8"
LC_MONETARY=nl_NL.UTF-8
LC_MESSAGES="en_US.UTF-8"
LC_PAPER=nl_NL.UTF-8
LC_NAME=nl_NL.UTF-8
LC_ADDRESS=nl_NL.UTF-8
LC_TELEPHONE=nl_NL.UTF-8
LC_MEASUREMENT=nl_NL.UTF-8
LC_IDENTIFICATION=nl_NL.UTF-8
LC_ALL=

Google "locale ubuntu" to read how to set it.

If you use systemd, maybe the locale or LANG setting is changed in the script.


You can also do this:
Stop sabnzbd as a service,
Start sabnzbd from the commandline:

Code: Select all

env LANG=c sabnzbdplus

... will go wrong

Code: Select all

env LANG=en_US.UTF-8 sabnzbdplus

... will be OK, but only if you have locale en_US.UTF-8 installed, which you can check with "locale -a"

[sander]

For reference / Google: a variant is

Code: Select all

SABnzbd was started with encoding ISO-8859-15, this should be UTF-8. Expect problems with Unicoded file and directory names in downloads.

which I got on my Debian on ARM.

And this how I solved it:

Code: Select all

locale 
sudo dpkg-reconfigure locales ### I removed everything but en_US.UTF-8
exit ### and a new login

locale ### Yes, LANG=en_US.UTF-8
cat /etc/default/locale 
sabnzbdplus ### Preferred encoding = UTF-8

Code: Select all

sander@cubox:~$ locale | grep LANG=
LANG=en_US.UTF-8

sander@cubox:~$ env | grep LANG
LANG=en_US.UTF-8

[nhasian]

To correct this error message on Ubuntu 16.04.2 LTS I had to run the command:

Code: Select all

sudo dpkg-reconfigure locales

I chose en_US.UTF-8 from the list then was presented with two choices:

C.UTF-8
en_US.UTF-8

I don't know what the difference is but I selected en_US.UTF-8. Then I restarted sabnzbd and the warning message disappeared.

[sander]

Then I restarted sabnzbd and the warning message disappeared.

Cool. You can now download posts with Unicode files within the rars.