In this exercise, you have to build upon the client that you implemented previously, i.e. the code that sends a request to a particular server. Now that you have two versions of the server, one that output integers and one that output strings, you might want to combine them into a single application using a simple protocol to request particular output format. Then, the client must be able to request the operation to perform, i.e. the sum or the multiplication of the integers.
Let us first define this protocol in details. The client sends its request using the following message format.
0 1 2 3 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ |Unused |Opcod|F| Payload (max 256 bytes) ... +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
The first 4 bits are unused, i.e. they are simply not considered and can contain any value. The next 3 bits represent the code of the operation (Opcode) that the server should execute. The currently defined opcodes are:
- 0: The sum operation.
- 1: The multiplication operation.
The final, least-significant, bit represent the output format (F) that should be used for the result. The formats defined are:
- 0: The network-ordered unsigned integer format.
- 1: The ASCII string format. ! There is no
\0
to indicate the end of the string ! It is not needed as the length of the packet can be used to find it !
This header byte is followed by the payload, that can be up to 256 bytes long. The payload is a sequence of network-ordered unsigned integers. Up to 64 integers can thus be sent in a request. Each request contains at least one integer within the payload.
Upon receiving a request, the server performs the required operation and outputs the result in the following message format.
0 1 2 3 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ |S| Unused | Payload (max 256 bytes) ... +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
The 1st and most-significant bit contains the status code (S). 0 indicates an error and 1 indicates a success. The remaining 7 bits are unused, i.e. they are simply not considered and can contain any value. The error code is used by the server when receiving requests with unknown opcodes. Messages with the status code set to error can contains a payload, but it is not meaningful to the client.
Implement a client function that takes a series of integers as parameter and sends them in the correct format to the given server address. Set the correct operation in the request based on the function parameters. After sending the request, wait for the receipt of the response and return the result of the operation.
For each of the calls that you use, if one of them were to encounter an error, immediately return -1. All require headers are already included.
Manpages of interest: