Homework 2: A Barebones HTTP/1.1 Client

In this programming exercise, you will create a barebones web client. While python includes a basic http client module http.client, this assignment will serve as a learning experience for translating a protocol into an implementation. Your deliverable will be a client which only implements the GET method and follows the basics of the HTTP/1.1 specification, enough to download files as one would with the command line program curl.

HTTP/1.1 Features

HTTP/1.0 describes the most basic functionality that an HTTP client is required to do. HTTP/1.1 includes several new features that extend the protocol. For this assignment, you will only be required to implement these additional features:

  • Include a Host: header
  • Correctly interpret Transfer-encoding: chunked
  • Include a Connection: close header, or handle persistent connections

These new features are described in James Marshall’s excellent HTTP Made Really Easy under the HTTP/1.1 clients subsection.

Note that the RFCs are your friends: if you’re having trouble with Transfer-encoding, check [the RFC][http] for hints!

Basic HTTP functionality

As seen in class, HTTP is a stateless request-response protocol that consists of an initial line, zero or more headers, and zero or more bytes of content. Your program will implement a function, retrieve_url, which takes a url (as a str) as its only argument, and uses the HTTP protocol to retrieve and return the body’s bytes (do not decode those bytes into a string). Consult the book or your class notes for the basics of the HTTP protocol.

You may assume that the URL will not include a fragment, query string, or authentication credentials. You are not required to follow any redirects - only return bytes when receiving a 200 OK response from the server. If for any reason your program cannot retrieve the resource correctly, retrieve_url should return None.

Testing Script

Testing your code does not have to be a manual affair. This assignment lends itself very well to automated testing.

We have provided a testing script for you. You will need to install the requests library to use it (this can be done using pip install requests, or possibly pip3 install requests, depending on your setup). While you cannot use requests to implement your assignment, you can use it to test your code.

The requests module has far more features implemented from the HTTP spec (and a different interface) than the code that you are implementing.

One you have requests installed, you can call the testing script with python ./hw2_test.py (with an optional --debug flag to provide more information). The testing script will compare your implementation of the retrieve_url function with a correct one, when calling a set of URLs. You should make sure that your function is giving the output that matches the known-correct output fetched by the testing script.

Remember, you only need to implement the features listed above. You should probably implement the Host: header (important) and the Connection: close header (easy) first, and then add chunked transfer encoding later. If you have any doubts about what you can/can’t use, or what you should/shouldn’t implement, ask on Piazza.


A trivial template is provided in this repository, as hw2.py.


Grading will be done automatically using a script. For this assignment, we will be providing a testing harness. This is not guaranteed to be the method we use for grading, but it will likely be very similar. If you wish, you can share test cases you have written with the class. Students who share test cases publicly will very likely receive extra participation credit.

Your program will be tested (at least) on these urls:

    'http://www.example.com',  # most basic example
    'http://accc.uic.edu/contact',  # longer basic example
    'http://i.imgur.com/fyxDric.jpg',  # is an image
    'http://illinois.edu/doesnotexist',  # causes 404
    'http://www.ifyouregisterthisforclassyouareoutofcontrol.com/', # NXDOMAIN
    'http://marvin.cs.uic.edu:8080/', # nonstandard port
    'http://www.httpwatch.com/httpgallery/chunked/chunkedimage.aspx' # chunked encoding

If you’re debugging a problem or simply curious, try firing up Wireshark, and then fetch the URL both with the requests library or curl program as well as with your code. You’ll be able to compare both requests as they were sent, as well as the responses received.

File Submission Requirements

You should include the following files in your repo when submitting the assignment. All other files will be ignored.

  • README.md: this file
  • hw2.py: python3 code that implements the function retrieve_url, matching the requirements discussed in this assignment.
  • netid.txt: a text file that contains only your UIC NetID.

Points and Scoring

There is a total of 16 main points, and 4 bonus points on this assignment, for a grand total of 20 possible points.

  • 1 point for correctly handling each of the 7 URLs mentioned above.
  • 3 points for correctly handling each of 3 additional URLs
  • 1 bonus point for correctly handling each of the additional test cases given in the TEST_CASES_BONUS list in the hw2_test.py testing script.
  • 1 bonus point for submitting a hw2.py that is “fully compliant” with the pep8 and pylint. “Fully compliant” here means that pep8 returns with no error messages, and pylint scores your code 10/10.

Your assignment is only eligible for bonus points if you get at least 5 points. I.e. you cannot submit a well-formatted, but completely broken, hw2.py and get the bonus point.

Not correctly including the netid.txt file will be an automatic 0 for the assignment.

Allowable sources

You may not use any libraries which implement parts or the whole of the HTTP specification - you must perform the basic request and response parsing/generation yourself, as well as the chunked content encoding.

Do not import or use any python libraries, or third party code, beyond what is imported in the skeleton / hw2.py file in your repo.

These resources may be useful:

Using any code from another source, even a single line, even with a citation, is not allowed. This includes using any implementation code from the standard library itself. I highly recommend not even Googling for solutions to portions of this homework - as soon as you’ve seen an alternate implementation, it is very hard to write one’s own.

Due Date

This assignment is due Friday, February 3 at 3pm. See the syllabus for the late turnin policy.

Remember, if you don’t push, we don’t see it! If you push even one second too late, your assignment won’t be graded. We will be counting turnin by push time, rather than by commit time, so please make sure to leave yourself ample time to verify that your code has been submitted successfully.