CSE381- Homework 5 Solved

Objective
The objective of this homework is to:

•       Understand the use of HTTP protocol for communication

•       Develop a C++ program that can process HTTP GET requests

•       Develop program to run a given command and process its outputs

•       Continue to gain familiarity with I/O streams & string processing
 

Grading Rubric:

The program submitted for this homework must pass necessary base case test(s) in order to qualify for earning any score at all. Programs that do not meet base case requirements will be assigned zero score!

Program that do not compile, have a method longer than 25 lines, or

                              just some skeleton code will be assigned zero score.
 

 
conciseness
If your methods are not concise points will be
•       See points set for each command and overall structure, organization, , variable-names, etc. in the next page.

deducted.

•       -1 Points: for each warning generated by the compiler (warnings are most likely sources of errors in C++ programs)

 

NOTE: Violating CSE programming style guidelines is a compiler error! Your program should not have any style violations.  

Background
A web-server is an application layer program that processes HTTP requests and generates responses. Recollect that HTTP is a multi-line text protocol that is used by web-browsers to interact with web-servers. In this homework you will be developing a program to respond to HTTP requests. Overall this program is essentially just string processing from I/O streams with most of the required code copy-pasted from different exercises/lecture slides.

Test files
The following files are supplied for this homework:

1.     You are supplied with several image, html, and text files for testing. 

2.     For functional testing, test files are in test_files directory. These are the same input and output files used by the CODE plug-in for checking.

Chunked Response Requirements [Review from lecture slides]
The response/output from the program/server should be in the following format, with each line terminated with a "\r\n":

 

1.     The first line should be HTTP/1.1 200 OK or HTTP/1.1 404 Not Found (if a specified file was not found). See example outputs 

2.     The second line should be Content-Type: MIMEtype, where MIMEtype is determined based on file extension as discussed further below. By default the MIMEtype is text/plain as in: Content-Type: text/plain

Transfer-Encoding: chunked

X-Client-Header-Count: num
3.     The next line must be  (a fixed constant string)

4.     The next line must be , where num is the number

HTTP headers sent by the client.

5.     The next line must be Connection: Close (a fixed constant string)

6.     The header section must be terminated by "\r\n"

 

Rest of the contents of a file or output of a program are printed line-by-line with line size (in hex on a separate line) followed by the data as shown below:

 

// Send chunk size to client.

os << std::hex << line.size() << "\r\n"; // Write the actual data for the line.

os << line << "\r\n";

             

The end of data must be sent to the client via the following trailer:

 

// Send trailer out to denote end of data os << "0\r\n\r\n";

 

 

             

Request Processing Requirements
Your program should process multiple lines of an HTTP GET request (read via std::cin) in the following manner:

 

1.     Base case -- Return contents of a file or error [15 points]: If the first line of the HTTP request does not start with GET /cgi-bin/exec, then assume that the GET request is in the format: GET /file HTTP/1.1, where file corresponds to the path of the file whose contents is to be returned back to the user using chunked transfer as discussed below. The file should be processed as:

 

a.     If the file is simply / the server should respond with contents of given index.html

 

b.     If the file path is valid (that is file is readable, see ifstream's good() method), the server should respond with the file's contents and appropriate HTTP headers & MIME type for the file deduced from the file name extension. The extension and MIME types are determined (and set in Content-Type output/response header) as tabulated in the table below:

Extension
MIME type
.html
text/html
.png
image/png
.jpg
image/jpeg
.txt
text/plain
Default (i.e., in all other cases)
text/plain
 

The contents of the file should be sent to the client line-by-line using chunked encoding as discussed earlier, along with necessary headers. See following sample input and output

files in test_files directory for details: file_test_inputs.txt, file_test_expected_outputs.txt, img_test_inputs.txt, img_test_expected_outputs.txt, html_test_inputs.txt, and html_test_expected_outputs.txt. 

 

 

!
The base case is relatively straightforward. Consequently, if the base case does not operate as expected, as per the course policy, the program will be assigned zero score.
 

 

2.     Unreadable file [7 points]: If the file is not readable (i.e.

std::ifstream::good() method returns false) the server should respond with a suitable HTTP 404 response/output. For example, see example output shown in test_files/err_test_expected_outputs.txt.

 

 

 

 

3.     Run program & return output [23 points]: If input line starts with "GET /cgibin/exec" then this line corresponds to the case where a program is to be run. In this case, this line will be in the format: GET /cgi-bin/exec?cmd=ls&args=1.txt "hello.cpp" HTTP/1.1 (one line), where cmd contains the command to run and args has list of space separated command-line arguments to the program. Use std::quoted to correctly handle double quoted words in args. Note that you need to

url_decode cmd and args prior to using them as they will be encoded (for an example see request in test_files/ cgi_bin_test_inputs1.txt). URL decoding method is copy-paste from lecture slides.

 

Appropriately process the input, run the specified command, obtain its outputs using pipes (see slide #19, #20 in 07_IPC.pdf, you will need to suitably use simple pipes example but run 1 process and read its outputs using a pipe) and route the outputs back to the client using chunked encoding as discussed earlier.

 

Functional testing
Several test inputs and expected output files are also supplied to streamline your testing as shown below:

 

$ ./hw5 < test_files/base_case1_inputs.txt my_base_case1_output.txt

$ diff my_base_case1_output.txt test_files/base_case1_expected_outputs.txt

 

Note: If your solution is correct then the above diff command will not print any output.

 

Ensure you test with all input files and corresponding expected outputs to check all features of your program.

 

Submit to Canvas
This homework assignment must be turned-in electronically via Canvas CODE plug-in. Ensure your C++ source files are named appropriately. Ensure your program compiles (without any warnings or style errors) successfully. Ensure you have tested operations of your program as indicated. Once you have tested your implementation, upload the following onto Canvas:

 

Ø The 1 C++ source file you developed for this homework

 

Upload the C++ source files to onto Canvas. Do not submit zip/7zip/tar/gzip files. Upload each file independently.