Matthew,
Did you notice this:
pi@raspberrypi:/home/weewx/bin $ ./wee_extension install=stackedwindrose-2.0
.0.tar.gz
Traceback (most recent call last):
File "./wee_extension", line 83, in <module>
main()
File "./wee_extension", line 62, in main
config_path, config_dict = weecfg.read_config(options.config, _args)
File "/home/weewx/bin/weecfg/__init__.py", line 218, in read_config
locations=locations, file_name=file_name)
File "/home/weewx/bin/weecfg/__init__.py", line 193, in find_file
raise IOError("%s is not a file" % file_path)
IOError: install=stackedwindrose-2.0.0.tar.gz is not a file
pi@raspberrypi:/home/weewx/bin $
It came from this post
<https://groups.google.com/d/msg/weewx-user/5GyFXgDu_CE/IbvyQnvUAAAJ> in
the user group where a user was trying to install an extension I had put
together.
When testing on a setup.py install on a VM at home I was executing:
/home/weewx/bin/wee_extension --install=/home/gary/weewx_extensions/
stackedwindrose-2.0.0.tar.gz
from /home/gary and everything worked (install/uninstall/list) many times
over the course of a day (multiple sessions). Never missed a beat. The user
claims that stackedwindrose-2.0.0.tar.gz is in /home/weewx/bin. I tried the
same command/setup as the user claims he used this morning on my VM and I
received the same error the user cited. The only way I could get it to work
was using full paths. I did a bit of debugging in weecfg.__init__ and as
far as I could tell wee_extension was getting hung up looking for weewx.conf
(certainly file_name == 'weewx.conf' when the error was thrown). I would
have expected that if __init__ was looking for weewx.conf and it found that
stackedwindrose-2.0.0.tar.gz was not a file it would have just moved on
rather than halting. I didn't get much further before I had to go out. I
have just come home again and sat down to do some more debugging and under
the same situation the error is not being thrown and the extension is
installing/uninstalling as expected.
I feel this is a bit of a non-question but wanted to run it by you while I
still had (some of) the logs. I will keep on this and try a diferent VM to
see if I can reproduce the issue.
Gary
