README.md

python-httpproblem

Utility library to work with RFC7807 Problem Details for HTTP APIs.

This library is very light-weight, with no external dependencies, fully-tested and works with both Python 2 and Python 3. It has special support for AWS lambda proxy integration output format but it should be easy to map to any other format or framework. Currently only JSON serialization is supported.

Installation

pip install httpproblem

Usage

Build a Problem dict

The problem() function that can be used to build a dict with the problem fields.

>>> pprint(problem(httplib.BAD_REQUEST, 'You do not have enough credit.', 'Your current balance is 30, but that costs 50.', '/account/12345/msgs/abc'))
{'detail': 'Your current balance is 30, but that costs 50.',
 'status': 400,
 'title': 'You do not have enough credit.',
 'type': '/account/12345/msgs/abc'}

You can also use problem extensions

>>> pprint(problem(httplib.BAD_REQUEST, 'You do not have enough credit.', 'Your current balance is 30, but that costs 50.', '/account/12345/msgs/abc', balance=30, accounts=['/account/12345','/account/67890']))
{'accounts': ['/account/12345', '/account/67890'],
 'balance': 30,
 'detail': 'Your current balance is 30, but that costs 50.',
 'status': 400,
 'title': 'You do not have enough credit.',
 'type': '/account/12345/msgs/abc'}

As specified by Predefined Problem Types:

The "about:blank" URI, when used as a problem type, indicates that the problem has no additional semantics beyond that of the HTTP status code.

When "about:blank" is used, the title SHOULD be the same as the recommended HTTP status phrase for that code (e.g., "Not Found" for 404, and so on), although it MAY be localized to suit client preferences (expressed with the Accept-Language request header).

So if this library will automatically fill the title field if the type is not present or about:blank.

>>> problem(404)
{'status': 404, 'title': 'Not Found'}
>>> problem(httplib.BAD_REQUEST, type='about:blank')
{'status': 400, 'type': 'about:blank', 'title': 'Bad Request'}

Build a Problem HTTP response

The problem_http_response() function helps to build HTTP responses using the format used by the AWS lambda proxy integration. The method will automatically fill the Content-Type header with application/problem+json and the HTTP response code with the status.

>>> pprint(problem_http_response(httplib.BAD_REQUEST))
{'body': '{"status": 400, "type": "about:blank", "title": "Bad Request"}',
 'headers': {'Content-Type': 'application/problem+json'},
 'statusCode': 400}

You can map this to other frameworks. For instance for Flask (or Werkzeug):

>>> problem = problem_http_response(400)
>>> print(flask.Response(problem['body'], status=problem['statusCode'], headers=problem['headers']))
<Response 39 bytes [400 BAD REQUEST]>

By default, json.dumps is used to serialize into JSON. This can be changed by using the set_serialize_function

>>> httpproblem.set_serialize_method(lambda data: json.dumps(data, indent=4))
>>> print(problem_http_response(400)['body'])
{
    "status": 400,
    "title": "Bad Request"
}

Raise a Problem exception

The Problem exception class can be used to simplify the error management with try/except. The class has methods to convert it to a Problem dict or HTTP response.

>>> try:
...     raise Problem(httplib.BAD_REQUEST)
... except Problem as e:
...     print(e.to_dict())
...
{'status': 400, 'title': 'Bad Request'}

The to_dict and to_http_response take a with_traceback argument that can be used to include the traceback. You can also set it globally with the activate_traceback() function. For security reasons, the default is to not include the traceback and it is recommended to not activate it in production.

>>> # Add traceback by call argument
>>> try:
...     raise Problem(httplib.BAD_REQUEST)
... except Problem as e:
...     pprint(e.to_dict(with_traceback=True))
...
{'status': 400,
 'title': 'Bad Request',
 'traceback': 'Traceback (most recent call last):\n  File "<stdin>", line 2, in <module>\nProblem: {\'status\': 400, \'title\': \'Bad Request\'}\n'}
>>>
>>> # Add traceback globally
>>> httpproblem.activate_traceback()
>>> try:
...     raise Problem(httplib.BAD_REQUEST)
... except Problem as e:
...     pprint(e.to_dict())
...
{'status': 400,
 'title': 'Bad Request',
 'traceback': 'Traceback (most recent call last):\n  File "<stdin>", line 2, in <module>\nProblem: {\'status\': 400, \'title\': \'Bad Request\'}\n'}