python storing requests headers as a variable?

Question

I have

headers = {'content-type': 'application/soap+xml'}

which works fine. however I want this to be able to be specified as an argument so in my args lets say I have

--wsheaders {'content-type':\s'application/soap+xml'}

This comes through fine as

{'content-type': 'application/soap+xml'}

... however when I do

headers = args.wsheaders

I get a ton of errors. So obviously headers cant be a string. But there must be a way to store this where it can be read back in a variable? Any ideas?

EDIT: I cant use the single quotes around the argument. I'm still learning this but it appears I can't use single or double quotes around an argument when passing the arguments from a file, like I CAN but it doesn't work to group the chars together, it instead becomes a literal part of the argument which isn't helpful since I lose everything before or after a space. see the attachment for what it looks like in the file.

--wsheaders '{'content-type': 'application/soap+xml'}'

error when using single quotes or double quotes

Sync03.py: error: unrecognized arguments: 'application/soap+xml'}'
Sync03.py: error: unrecognized arguments: 'application/soap+xml'}"

so instead I use

--wsheaders {'content-type':\s'application/soap+xml'}

but then the \s needs to be replaced after the arg is entered which I do, but then the variable is a string and I'm back to the origial problem.

type=json.loads

doesn't work in the argument definition when there's a \s since it's an unrecognized json. If there was a way to replace the \s with argparse it might work to do that first...but I dont think thats possible. With a space then with a \s

Sync03.py: error: argument --wsheaders: invalid loads value: "{'content-type':"
Sync03.py: error: argument --wsheaders: invalid loads value: "{'content-type':\\s'application/soap+xml'}"

EDIT

parser.add_argument('--wsheaders', type=lambda x: json.loads(x.replace('\s', '').replace('\'', '"')))

This worked as per bschlueter comment directly below.

Answer 1

Assuming you're consuming --wsheaders in your python script as an argument somehow, all that you need to do is quote the headers when you pass them to the script:

the_script --wsheaders '{"content-type": "application/soap+xml"}'

Then parse them as json to get the dict.

Using pyyaml:

>>> import yaml
>>> yaml.load(args.wsheaders)
{'content-type': 'application/soap+xml'}

EDIT:

If you are using argparse (as you should be), you can easily do the parsing when the args are parsed by declaring the argument with:

parser = argparse.ArgumentParser()
parser.add_argument('--wsheaders', type=json.loads)
args = parser.parse_args()

Answer 2

I assumed you used argparse.

In fact argparse uses sys.argv which captures passed arguments as string. So in this case you are passing dictionary as string and for this reason you need to parse it as dictionary because requests needs header as dictionary.

For this reason headers = args.wsheaders is useless since args.wsheaders is string. You need to use json or ast to parse it as dictionary as below-

headers = ast.literal_eval(args.wsheaders)

More details of passing dictionary as command-line argument and parsing them correctly at here, here and here.