add DotEnv dumps/dump

[mattwwarren]

fixes #502

[mattwwarren]

Maybe this should be a classmethod

[AMDphreak]

How does this code snippet write anything to the .env file? I see no file write operation.

People just want to read and write. Don't need to 'dump' just write. Write one, two, or some dict of variables. The library should have an function for users to add to its dict, and a way to write new items to the file.

During the process of writing to the file, add the items to the dict. Do not require programmers to first add them to the lib's dict and then explicitly write them to the file. The intuition of the programmer is that this lib has a direct connection to the .env file, so they shouldn't even know that you have a dict in memory.

main API calls:
load_dotenv() (since you chose load over read)
write_dotenv(dict[]) <- equivalent to add_var() but also writes the file. Assumption is the programmer wants to append to .env file.

special use functions:
add_var(dict) <- puts var in dict but doesn't write the file.
add_var(dict[]) <- Just overload it. I know it's not plural but programmers like consistency. this would be for someone who for some reason needs to write lots of vars without
rem_var(dict) <- removes item from array with arrayname.remove() function
rem_var(dict[]) <- not plural but overload it.

I would like to use the .env variable to store settings from a user, since the programmer can provide an API key, or the user can provide an API key. I suppose this isn't the usual way to do things, but this is simpler for me than creating files in The user folder or AppData.