main.py

import editdistance
from hexdump2 import hexdump
import gradio as gr
import shlex
import subprocess
import tempfile

description = """This is a space testing a method for evaluating the quality of decompilation.

Currently unhandled features:
* PIC stuff
  * Global references
  * Function calls
  * Wildcards in target function?
"""


def compile(compiler, flags, source):
    # Create a temporary file for the C source code
    with tempfile.NamedTemporaryFile(suffix=".c", delete=False) as temp_c_file:
        temp_c_file.write(source.encode())
        temp_c_file_name = temp_c_file.name

    # Create a temporary file for the object file
    with tempfile.NamedTemporaryFile(suffix=".o", delete=False) as temp_o_file:
        temp_o_file_name = temp_o_file.name

    # Compile the C file to an object file
    subprocess.run(
        [compiler, "-c", temp_c_file_name]
        + shlex.split(flags)
        + ["-o", temp_o_file_name],
        check=True,
    )

    # Create a temporary file for the raw bytes
    with tempfile.NamedTemporaryFile(suffix=".raw", delete=True) as raw_bytes_file:
        subprocess.run(
            [
                "objcopy",
                "--only-section",
                ".text",
                "-O",
                "binary",
                temp_o_file_name,
                raw_bytes_file.name,
            ]
        )
        compiled_bytes = raw_bytes_file.read()
        print(compiled_bytes)

    return compiled_bytes


def predict(target_bytes, source, compiler, flags):
    target_bytes = bytes.fromhex(target_bytes)
    compiled_bytes = compile(compiler, flags, source)
    return (
        hexdump(compiled_bytes, result="return"),
        hexdump(target_bytes, result="return"),
        editdistance.eval(compiled_bytes, target_bytes)
    )


def run():
    demo = gr.Interface(
        fn=predict,
        description=description,
        inputs=[
            gr.Textbox(
                lines=10,
                label="Bytes of Target Function (in hex)",
                value="b8 2a 00 00 00 c3",
            ),
            gr.Textbox(
                lines=10,
                label="Decompiled C Source Code",
                value="int foo() { return 42; }",
            ),
            gr.Textbox(label="Compiler", value="g++"),
            gr.Textbox(label="Compiler Flags", value="-O2"),
        ],
        outputs=[
            gr.Textbox(label="Compiled bytes"),
            gr.Textbox(label="Target bytes"),
            gr.Number(label="Edit distance (lower is better)"),
        ],
    )

    demo.launch(server_name="0.0.0.0", server_port=7860)


run()