HOW_TO_COMMANDLINE_BLENDER.txt

How to run a Blender add-on from the command line.
WHY: its easier than re-zipping, and restarting blender everytime you want to 
    test a change to an add-on you're developing

In short: run blender from the command line and 
use an option to get it to run a python script
that registers and invokes your add-on.

Here's the command that will do it in the end (explanation below):

blender E:\assets\arrow.blend --python-use-system-env --python run-addon.py


Explanation / Steps:
--add blender exec to your system path (for convenience)
--make a new directory somehere. say: "E:\blender\dev"
--add this directory to your PYTHONPATH environment variable 
    (for importing a module that you'll put inside the directory you just made)
--probably need to: disable the existing version of the add-on you want to work on in blender.
--copy the source folder for the add-on into your 'E:\blender\dev' dir. 
    (where source folder means the one containing the top level __init__.py)
--optional: rename the copied add-on directory/module. e.g. 'babylon_js' -> 'babylon_dev_js'
--make a new python script file in the 'blender\dev' directory. Call it "run-addon.py" (or whatever you want).
--if your add-on is named 'babylon_dev_js'
        run-addon.py should look like this:

#### RUN-ADDON.PY #####
import babylon_dev_js 
import bpy

print('hi from run-addon.py')

babylon_dev_js.register()

# try bab exporter
print('trying the bab exporter')

# invoke the add-on (call its execute() method)
bpy.ops.export.bjs()

# quit blender for convenience, if you want
bpy.ops.wm.quit_blender()

#### END OF RUN-ADDON.PY ####

Now you can run:

blender E:\assets\arrow.blend --python-use-system-env --python run-addon.py

to test the add-on. ('arrow.blend' is the file you want to
 open in blender. Leave out to run on the default scene.)


SOME NOTES:
In my case, I had to help the addon to work without user interaction.
E.g. set a value for self.filepath in the execute() method before it did anything
that relied on having a value there:

self.filepath = "E:\some\path\arrow.babylon" 

Since my dev copy of the add-on has a different name from the stable version
I can check the add-ons name before performing this set up step. useful (i guess) if I ever want to
actually use a non dev version and don't want to have to remember to undo the dev only stuff:

if __name__ == 'babylon_dev_js':
    self.filepath = "E:\some\useful\path\arrow.babylon'



ALTERNATE METHOD: 
install the add-on in blender in the usual way
and simply edit the copy in blender's add-on folder.
BETTER because: fewer steps. no need to edit your PYTHONPATH.
WORSE because: seems messier. maybe you actually want to use the stable version of the add-on.
    If you make a dev copy of the module, you can change its name (i.e. change the folder name).
    You can then check against the __name__ variable in the module's code to see if you're 
    running the dev version or not.

ALTERNATE STEPS:
--Install the add-on
--Add blender to your PATH
--now run-addon.py looks like:

#### RUN-ADDON.PY #####
import bpy

print('hi from run-addon.py')
# try bab exporter
print('trying the bab exporter')

# invoke the add-on (call its execute() method)
bpy.ops.export.bjs()

# quit blender for convenience, if you want
bpy.ops.wm.quit_blender()

#### END OF RUN-ADDON.PY ####

--and I think you can omit the --python-use-system-env flag in the command since you don't
    need to import the add-on module

blender E:\assets\arrow.blend --python run-addon.py