Your mission: implement a web service that takes in a set of bytes as input and outputs a MOS 6502 microprocessor family disassembly of those bytes as output. For example, assuming the following input bytes:
a9 bd a0 bd 20 28 ba
The following output (or similar) should be produced:
0x0000 a9 bd LDA #$bd
0x0002 a0 bd LDY #$bd
0x0004 20 28 ba JSR $ba28
You can, based on your own preferences, change the format of the output, but the origin address and bytes used in
disassembly should always be visible. You can deal with unknown opcodes however you deem appropriate
(outputting just the hex bytes is sufficient, for example). You should assume the input begins at address 0 (or
you can make this configurable in some fashion). You should always try to consume all of the input data (or you
can make this configurable too!). There are some example binaries in the test-bin/ directory of
the source tree that contain valid 6502 code that can be used for testing.
Information on the MOS 6502 instruction set is available all over the internet, but a handy instruction set guide that should provide sufficient information is available at https://www.masswerk.at/6502/6502_instruction_set.html
A simple testing scaffold is provided for testing and development convenience that should serve as a starting point.
You can cargo run the scaffold in one terminal to start up the testing web server and then run cargo test in
another to send a test request to it. You can modify the existing test and add new ones as desired.
You can spend as much time as you feel is required for a baseline implementation. The idea is not to find out if you can read an opcode listing and implement instruction decoding based on that - we can assume this is the case. Instead, try to sketch out an architecture and abstractions for your program that are easy to understand and possibly extend. Be sure to consider things like "what if we wanted to implement support for additional architectures?"
If you can find a way to produce proper disassembly from the binary data without implementing a full disassembler, feel free to do so as well.
Incomplete solutions are also better than nothing! Don't worry if you don't have time to implement decoding for all instructions or are missing some undocumented ones! If you find the task to be unexpectedly difficult, don't worry! If you can provide some initial code and are prepared to explain your implementation idea, that is fine as well.
Don't hesitate to use pseudocode in places where you feel a lot of boilerplate or obvious implementations are required.