multiple terminators for strz

[kaitai-io]

multiple terminators for strz

[GreyCat]

Try something like that:

meta:
    id: dat
    application: script
    endian: le
seq:
    - id: commands
      repeat: eos
      type: script_cmd
types:
    script_cmd:
        seq:
            - id: cmd_start
              contents: [0x23] 
            - id: num
              type: u2
            - id: params
              type: strz
              encoding: SJIS
              terminator: 0x23
              consume: false
              eos-error: false

That allows to parse the file you've provided without errors. However, it is still probably not a very good idea to parse the file using "parse anything up to 0x23" semantic - real parsing code rarely use something like that, and parsing params as a 0x23-terminated string is probably not what you want anyway.

In 99% of cases I've seen (with notable exception of terminator-based strings encoding), if we're dealing with something like a bytecode (i.e. [opcode] [parameters]), variable length parameters depend on the opcode itself, not the termination symbol. Notably, it's generally a very bad idea to add "termination bytes" like 0x23, given that some opcode might have genuine integer 0x23 as a parameter.

It is possible to implement something like "read bytes until some termination byte / sequence of bytes, and then reinterpret them as some complex subtype" in future, but so far I haven't seen a single example where such a semantic would be helpful. I'd be grateful if you'll prove me wrong :)