{ "cells": [ { "cell_type": "markdown", "id": "38e656cb", "metadata": {}, "source": [ "# Example: Modifying Existing Bytecode\n", "\n", "In this example, we will compile some code, modify the bytecode, and then turn it back into Python code to execute.\n", "\n", "We can make a code object from a string using `compile`:" ] }, { "cell_type": "code", "execution_count": 1, "id": "d6b5b0f9", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "110\n" ] } ], "source": [ "x = True\n", "source_code = \"print(10 + (100 if x else 10))\"\n", "code = compile(source_code, \"\", \"exec\")\n", "exec(code)" ] }, { "cell_type": "markdown", "id": "85231cf0", "metadata": {}, "source": [ "If we look at the code object, we can see that it does have the bytecode, but its represented as byte string, which isn't very helpful:" ] }, { "cell_type": "code", "execution_count": 2, "id": "f2203180", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ " at 0x7f022741c500, file \"\", line 1>\n", "b'e\\x00d\\x00e\\x01r\\x06d\\x01n\\x01d\\x00\\x17\\x00\\x83\\x01\\x01\\x00d\\x02S\\x00'\n" ] } ], "source": [ "print(code)\n", "print(code.co_code)" ] }, { "cell_type": "markdown", "id": "d19a9433", "metadata": {}, "source": [ "We could use Python's built in `dis` module to introspect the code object. This is helpful to look at it, but won't let us change it:" ] }, { "cell_type": "code", "execution_count": 3, "id": "e10855a1", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ " 1 0 LOAD_NAME 0 (print)\n", " 2 LOAD_CONST 0 (10)\n", " 4 LOAD_NAME 1 (x)\n", " 6 POP_JUMP_IF_FALSE 6 (to 12)\n", " 8 LOAD_CONST 1 (100)\n", " 10 JUMP_FORWARD 1 (to 14)\n", " >> 12 LOAD_CONST 0 (10)\n", " >> 14 BINARY_ADD\n", " 16 CALL_FUNCTION 1\n", " 18 POP_TOP\n", " 20 LOAD_CONST 2 (None)\n", " 22 RETURN_VALUE\n" ] } ], "source": [ "import dis\n", "dis.dis(code)" ] }, { "cell_type": "markdown", "id": "0e15bccf", "metadata": {}, "source": [ "So instead, lets turn it into ✨data✨:" ] }, { "cell_type": "code", "execution_count": 4, "id": "30db19a3", "metadata": {}, "outputs": [ { "data": { "text/plain": [ "CodeData(blocks=((Instruction(name='LOAD_NAME', arg=Name(name='print', _index_override=0), _n_args_override=None, line_number=1, _line_offsets_override=()), Instruction(name='LOAD_CONST', arg=Constant(constant=10, _index_override=0), _n_args_override=None, line_number=1, _line_offsets_override=()), Instruction(name='LOAD_NAME', arg=Name(name='x', _index_override=1), _n_args_override=None, line_number=1, _line_offsets_override=()), Instruction(name='POP_JUMP_IF_FALSE', arg=Jump(target=1, relative=False), _n_args_override=None, line_number=1, _line_offsets_override=()), Instruction(name='LOAD_CONST', arg=Constant(constant=100, _index_override=1), _n_args_override=None, line_number=1, _line_offsets_override=()), Instruction(name='JUMP_FORWARD', arg=Jump(target=2, relative=True), _n_args_override=None, line_number=1, _line_offsets_override=())), (Instruction(name='LOAD_CONST', arg=Constant(constant=10, _index_override=0), _n_args_override=None, line_number=1, _line_offsets_override=()),), (Instruction(name='BINARY_ADD', arg=NoArg(_arg=0), _n_args_override=None, line_number=1, _line_offsets_override=()), Instruction(name='CALL_FUNCTION', arg=1, _n_args_override=None, line_number=1, _line_offsets_override=()), Instruction(name='POP_TOP', arg=NoArg(_arg=0), _n_args_override=None, line_number=1, _line_offsets_override=()), Instruction(name='LOAD_CONST', arg=Constant(constant=None, _index_override=2), _n_args_override=None, line_number=1, _line_offsets_override=()), Instruction(name='RETURN_VALUE', arg=NoArg(_arg=0), _n_args_override=None, line_number=1, _line_offsets_override=()))), filename='', first_line_number=1, name='', stacksize=3, type=None, freevars=(), future_annotations=False, _nested=False, _additional_line=None, _additional_args=())" ] }, "execution_count": 4, "metadata": {}, "output_type": "execute_result" } ], "source": [ "from code_data import CodeData\n", "\n", "code_data = CodeData.from_code(code)\n", "code_data" ] }, { "cell_type": "markdown", "id": "13b2e66b", "metadata": {}, "source": [ "This is still a bit hard to see, so let's install Rich's pretty print helper:" ] }, { "cell_type": "code", "execution_count": 5, "id": "b72ea861", "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n",
       "CodeData(\n",
       "    (\n",
       "        (\n",
       "            Instruction(\n",
       "                'LOAD_NAME',\n",
       "                Name('print', _index_override=0),\n",
       "                line_number=1\n",
       "            ),\n",
       "            Instruction(\n",
       "                'LOAD_CONST',\n",
       "                Constant(10, _index_override=0),\n",
       "                line_number=1\n",
       "            ),\n",
       "            Instruction(\n",
       "                'LOAD_NAME',\n",
       "                Name('x', _index_override=1),\n",
       "                line_number=1\n",
       "            ),\n",
       "            Instruction('POP_JUMP_IF_FALSE', Jump(1), line_number=1),\n",
       "            Instruction(\n",
       "                'LOAD_CONST',\n",
       "                Constant(100, _index_override=1),\n",
       "                line_number=1\n",
       "            ),\n",
       "            Instruction('JUMP_FORWARD', Jump(2, relative=True), line_number=1)\n",
       "        ),\n",
       "        (\n",
       "            Instruction(\n",
       "                'LOAD_CONST',\n",
       "                Constant(10, _index_override=0),\n",
       "                line_number=1\n",
       "            ),\n",
       "        ),\n",
       "        (\n",
       "            Instruction('BINARY_ADD', line_number=1),\n",
       "            Instruction('CALL_FUNCTION', 1, line_number=1),\n",
       "            Instruction('POP_TOP', line_number=1),\n",
       "            Instruction(\n",
       "                'LOAD_CONST',\n",
       "                Constant(None, _index_override=2),\n",
       "                line_number=1\n",
       "            ),\n",
       "            Instruction('RETURN_VALUE', line_number=1)\n",
       "        )\n",
       "    ),\n",
       "    filename='',\n",
       "    first_line_number=1,\n",
       "    name='<module>',\n",
       "    stacksize=3\n",
       ")\n",
       "
\n" ], "text/plain": [ "\n", "\u001b[1mCodeData\u001b[0m\u001b[1m(\u001b[0m\n", " \u001b[1m(\u001b[0m\n", " \u001b[1m(\u001b[0m\n", " \u001b[1mInstruction\u001b[0m\u001b[1m(\u001b[0m\n", " 'LOAD_NAME',\n", " \u001b[1mName\u001b[0m\u001b[1m(\u001b[0m'print', _index_override=\u001b[1m0\u001b[0m\u001b[1m)\u001b[0m,\n", " line_number=\u001b[1m1\u001b[0m\n", " \u001b[1m)\u001b[0m,\n", " \u001b[1mInstruction\u001b[0m\u001b[1m(\u001b[0m\n", " 'LOAD_CONST',\n", " \u001b[1mConstant\u001b[0m\u001b[1m(\u001b[0m\u001b[1m10\u001b[0m, _index_override=\u001b[1m0\u001b[0m\u001b[1m)\u001b[0m,\n", " line_number=\u001b[1m1\u001b[0m\n", " \u001b[1m)\u001b[0m,\n", " \u001b[1mInstruction\u001b[0m\u001b[1m(\u001b[0m\n", " 'LOAD_NAME',\n", " \u001b[1mName\u001b[0m\u001b[1m(\u001b[0m'x', _index_override=\u001b[1m1\u001b[0m\u001b[1m)\u001b[0m,\n", " line_number=\u001b[1m1\u001b[0m\n", " \u001b[1m)\u001b[0m,\n", " \u001b[1mInstruction\u001b[0m\u001b[1m(\u001b[0m'POP_JUMP_IF_FALSE', \u001b[1mJump\u001b[0m\u001b[1m(\u001b[0m\u001b[1m1\u001b[0m\u001b[1m)\u001b[0m, line_number=\u001b[1m1\u001b[0m\u001b[1m)\u001b[0m,\n", " \u001b[1mInstruction\u001b[0m\u001b[1m(\u001b[0m\n", " 'LOAD_CONST',\n", " \u001b[1mConstant\u001b[0m\u001b[1m(\u001b[0m\u001b[1m100\u001b[0m, _index_override=\u001b[1m1\u001b[0m\u001b[1m)\u001b[0m,\n", " line_number=\u001b[1m1\u001b[0m\n", " \u001b[1m)\u001b[0m,\n", " \u001b[1mInstruction\u001b[0m\u001b[1m(\u001b[0m'JUMP_FORWARD', \u001b[1mJump\u001b[0m\u001b[1m(\u001b[0m\u001b[1m2\u001b[0m, relative=\u001b[3mTrue\u001b[0m\u001b[1m)\u001b[0m, line_number=\u001b[1m1\u001b[0m\u001b[1m)\u001b[0m\n", " \u001b[1m)\u001b[0m,\n", " \u001b[1m(\u001b[0m\n", " \u001b[1mInstruction\u001b[0m\u001b[1m(\u001b[0m\n", " 'LOAD_CONST',\n", " \u001b[1mConstant\u001b[0m\u001b[1m(\u001b[0m\u001b[1m10\u001b[0m, _index_override=\u001b[1m0\u001b[0m\u001b[1m)\u001b[0m,\n", " line_number=\u001b[1m1\u001b[0m\n", " \u001b[1m)\u001b[0m,\n", " \u001b[1m)\u001b[0m,\n", " \u001b[1m(\u001b[0m\n", " \u001b[1mInstruction\u001b[0m\u001b[1m(\u001b[0m'BINARY_ADD', line_number=\u001b[1m1\u001b[0m\u001b[1m)\u001b[0m,\n", " \u001b[1mInstruction\u001b[0m\u001b[1m(\u001b[0m'CALL_FUNCTION', \u001b[1m1\u001b[0m, line_number=\u001b[1m1\u001b[0m\u001b[1m)\u001b[0m,\n", " \u001b[1mInstruction\u001b[0m\u001b[1m(\u001b[0m'POP_TOP', line_number=\u001b[1m1\u001b[0m\u001b[1m)\u001b[0m,\n", " \u001b[1mInstruction\u001b[0m\u001b[1m(\u001b[0m\n", " 'LOAD_CONST',\n", " \u001b[1mConstant\u001b[0m\u001b[1m(\u001b[0m\u001b[3mNone\u001b[0m, _index_override=\u001b[1m2\u001b[0m\u001b[1m)\u001b[0m,\n", " line_number=\u001b[1m1\u001b[0m\n", " \u001b[1m)\u001b[0m,\n", " \u001b[1mInstruction\u001b[0m\u001b[1m(\u001b[0m'RETURN_VALUE', line_number=\u001b[1m1\u001b[0m\u001b[1m)\u001b[0m\n", " \u001b[1m)\u001b[0m\n", " \u001b[1m)\u001b[0m,\n", " filename='',\n", " first_line_number=\u001b[1m1\u001b[0m,\n", " name='',\n", " stacksize=\u001b[1m3\u001b[0m\n", "\u001b[1m)\u001b[0m\n" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "from rich import pretty\n", "pretty.install()\n", "code_data" ] }, { "cell_type": "markdown", "id": "5d607824", "metadata": {}, "source": [ "That's better!\n", "\n", "We can see now that we have two blocks, each with a list of instructions.\n", "\n", "Let's try to change the additions to subtractions!" ] }, { "cell_type": "code", "execution_count": 6, "id": "61834257", "metadata": {}, "outputs": [], "source": [ "from dataclasses import replace\n", "\n", "new_code_data = replace(\n", " code_data,\n", " blocks=tuple(tuple(\n", " replace(instruction, name=\"BINARY_SUBTRACT\") if instruction.name == \"BINARY_ADD\" else instruction\n", " for instruction in block\n", " ) for block in code_data.blocks)\n", ")" ] }, { "cell_type": "markdown", "id": "6b291cb3", "metadata": {}, "source": [ "Now we can turn this back into code and exec it!" ] }, { "cell_type": "code", "execution_count": 7, "id": "c4270732", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "-90\n" ] } ], "source": [ "new_code = new_code_data.to_code()\n", "exec(new_code)" ] } ], "metadata": { "jupytext": { "cell_metadata_filter": "-all", "formats": "md:myst", "text_representation": { "extension": ".md", "format_name": "myst", "format_version": 0.13, "jupytext_version": "1.11.5" } }, "kernelspec": { "display_name": "Python 3 (ipykernel)", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.10.4" }, "source_map": [ 14, 22, 27, 31, 34, 38, 41, 45, 50, 54, 58, 66, 76, 80 ] }, "nbformat": 4, "nbformat_minor": 5 }