Skip to content

Disassembler API

The AppleCommander disassembler API is a reusable component that can be utilized by other Java projects. The easiest mechanism is to use a dependency manager such as Gradle or Maven.

<dependency>
    <groupId>org.applecommander</groupId>
    <artifactId>acdasm</artifactId>
    <version>6.0</version>
</dependency>

dependencies {
    implementation 'org.applecommander:acdasm:6.0'
}

API

Invoking the disassembler is done through a "builder" type interface. For example:

List<Instruction> assembly = Disassembler.with(code)
        .startingAddress(startAddress)
        .bytesToSkip(offset)
        .bytesToDecode(length)
        .use(instructionSet)
        .section(libraries)
        .decode(labels);

The only required components are code and the InstructionSet.

The following instruction sets are available:

Name Creation
6502 InstructionSet6502.for6502()
6502 + Illegal Instructions InstructionSet6502.for6502withIllegalInstructions
65C02 InstructionSet6502.for65C02()
SWEET-16 InstructionSetSWEET16.forSWEET16()
6502 + SWEET-16 InstructionSet6502Switching.withSwitching()
Z80 InstructionSetZ80.forZ80()
p-code InstructionSetPCode.forApplePascal()

As a minimal example, this would setup to disassemble 6502 code:

byte[] code = File.readAllBytes(file);
List<Instruction> assembly = Disassembler.with(code)
    .use(InstructionSet6502.for6502())
    .decode();

If you want the labels, create a Map<String,Integer> to capture them and pass it in to the decode(Map) method.

Important Interfaces/Classes/Records

The Dissassembler decode interface returns a list of Instruction components. Each component contains the address, bytes involved, the opcode mnemonic, and a list of operands. The operands have two format options: format() will format the operand as expected while format(String) will replace an address with the label instead. To determine if an operand contains an address, the address() method returns a Java Optional class, so use it as applicable as the return value is never null.

Instructionint address();byte[] code();String mnemonic();List<Operand> operands();Optional<String> description();Optional<Operand> addressRef();OperandString opFmt();String value();Optional<Integer> address();String format();String format(String label);InstructionSetDefaults defaults();List<Instruction> decode(Program program);List<OpcodeTable> opcodeTables();OpcodeTableString name();String opcodeExample(int opcode);Defaultsint startAddress();List<String> libraryLabels();int bytesPerInstruction();boolean includeDescription();decodes
Instructionint address();byte[] code();String mnemonic();List<Operand> operands();Optional<String> description();Optional<Operand> addressRef();OperandString opFmt();String value();Optional<Integer> address();String format();String format(String label);InstructionSetDefaults defaults();List<Instruction> decode(Program program);List<OpcodeTable> opcodeTables();OpcodeTableString name();String opcodeExample(int opcode);Defaultsint startAddress();List<String> libraryLabels();int bytesPerInstruction();boolean includeDescription();decodes

If there is a new instruction set to add, implement the InstructionSet. The instruction set also supplies Defaults that have various attributes that can be passed into the Disassembler.

Note that an instruction set not only decodes a program but also supplies opcode examples for the GenerateMarkdown program. Since the instruction sets are manually added into this application, the opcodeTables() method can likely be ignored (at least at first).