Spaces:
Running
Running
File size: 9,076 Bytes
ee7d7b9 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 | """
🖥️ SANDBOXED CODE EXECUTOR — AI Chat v2
=========================================
Executes Python code in a sandboxed subprocess with:
- Timeout protection (max 30 seconds)
- Output capture (stdout, stderr)
- Matplotlib figure capture
- DataFrame result capture
- Memory limits
- No file system access outside workspace
This enables the AI chat to generate and execute code inline,
showing results directly in the conversation.
"""
import os
import sys
import json
import subprocess
import tempfile
import logging
import base64
import time
import uuid
from typing import Dict, Any, Optional, List, Tuple
from pathlib import Path
logger = logging.getLogger(__name__)
# Maximum execution time in seconds
MAX_EXECUTION_TIME = 30
# Maximum output size in characters
MAX_OUTPUT_SIZE = 50000
class CodeExecutor:
"""
Execute Python code in a sandboxed subprocess.
Features:
- Captures stdout, stderr
- Captures matplotlib figures as base64 PNG
- Captures pandas DataFrame results
- Enforces timeout
- Returns structured results
"""
def __init__(self, user_id: str, workspace_dir: str = None):
self.user_id = user_id
self.workspace_dir = workspace_dir or self._get_workspace_dir()
os.makedirs(self.workspace_dir, exist_ok=True)
def _get_workspace_dir(self) -> str:
"""Get or create workspace directory for this user."""
base = os.path.join(os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__)))), ".workspace")
user_dir = os.path.join(base, self.user_id, "code_exec")
os.makedirs(user_dir, exist_ok=True)
return user_dir
def execute(self, code: str, context: Dict[str, Any] = None) -> Dict[str, Any]:
"""
Execute Python code and return structured results.
Args:
code: Python code to execute
context: Optional context dict (e.g., user's dataframe info)
Returns:
{
"success": bool,
"stdout": str,
"stderr": str,
"figures": [{"base64": str, "format": "png"}],
"tables": [{"columns": [...], "data": [[...]]}],
"result": str, # repr of last expression
"execution_time_ms": int,
"error": str or None
}
"""
# Create a wrapper script that captures everything
exec_id = str(uuid.uuid4())[:8]
script_path = os.path.join(self.workspace_dir, f"exec_{exec_id}.py")
output_path = os.path.join(self.workspace_dir, f"output_{exec_id}.json")
figures_dir = os.path.join(self.workspace_dir, f"figures_{exec_id}")
os.makedirs(figures_dir, exist_ok=True)
# Build the wrapper script
wrapper_code = self._build_wrapper(code, output_path, figures_dir, context)
try:
with open(script_path, 'w', encoding='utf-8') as f:
f.write(wrapper_code)
# Execute in subprocess
start_time = time.time()
env = os.environ.copy()
env["PYTHONIOENCODING"] = "utf-8"
env["MPLBACKEND"] = "Agg" # Non-interactive matplotlib backend
result = subprocess.run(
[sys.executable, script_path],
capture_output=True,
text=True,
timeout=MAX_EXECUTION_TIME,
cwd=self.workspace_dir,
env=env,
encoding='utf-8',
errors='replace'
)
elapsed_ms = int((time.time() - start_time) * 1000)
# Read structured output
output_data = {}
if os.path.exists(output_path):
try:
with open(output_path, 'r', encoding='utf-8') as f:
output_data = json.load(f)
except:
pass
# Collect figures
figures = []
for fig_file in sorted(os.listdir(figures_dir)):
if fig_file.endswith('.png'):
fig_path = os.path.join(figures_dir, fig_file)
with open(fig_path, 'rb') as f:
b64 = base64.b64encode(f.read()).decode('utf-8')
figures.append({"base64": b64, "format": "png"})
stdout = (result.stdout or "")[:MAX_OUTPUT_SIZE]
stderr = (result.stderr or "")[:MAX_OUTPUT_SIZE]
# Filter common warnings from stderr
stderr_lines = [l for l in stderr.split('\n') if l.strip() and not any(w in l for w in [
'UserWarning', 'FutureWarning', 'DeprecationWarning', 'RuntimeWarning'
])]
stderr_clean = '\n'.join(stderr_lines)
return {
"success": result.returncode == 0,
"stdout": stdout,
"stderr": stderr_clean,
"figures": figures,
"tables": output_data.get("tables", []),
"result": output_data.get("result", ""),
"execution_time_ms": elapsed_ms,
"error": stderr_clean if result.returncode != 0 else None
}
except subprocess.TimeoutExpired:
return {
"success": False,
"stdout": "",
"stderr": f"Execution timed out after {MAX_EXECUTION_TIME} seconds",
"figures": [],
"tables": [],
"result": "",
"execution_time_ms": MAX_EXECUTION_TIME * 1000,
"error": f"Code execution timed out after {MAX_EXECUTION_TIME}s"
}
except Exception as e:
return {
"success": False,
"stdout": "",
"stderr": str(e),
"figures": [],
"tables": [],
"result": "",
"execution_time_ms": 0,
"error": str(e)
}
finally:
# Cleanup temp files
for path in [script_path, output_path]:
try:
if os.path.exists(path):
os.remove(path)
except:
pass
try:
import shutil
if os.path.exists(figures_dir):
shutil.rmtree(figures_dir, ignore_errors=True)
except:
pass
def _build_wrapper(self, code: str, output_path: str, figures_dir: str,
context: Dict = None) -> str:
"""Build the execution wrapper script."""
# Escape the paths for Windows
output_path_escaped = output_path.replace('\\', '\\\\')
figures_dir_escaped = figures_dir.replace('\\', '\\\\')
wrapper = f'''
import sys
import os
import json
import warnings
warnings.filterwarnings("ignore")
# Setup matplotlib for non-interactive use
import matplotlib
matplotlib.use("Agg")
import matplotlib.pyplot as plt
# Standard data science imports
import pandas as pd
import numpy as np
# Track figures
_figure_count = [0]
_tables = []
_result_value = [None]
_output_path = r"{output_path_escaped}"
_figures_dir = r"{figures_dir_escaped}"
# Monkey-patch plt.show to save figures instead
_original_show = plt.show
def _capture_show(*args, **kwargs):
for fig_num in plt.get_fignums():
fig = plt.figure(fig_num)
fig_path = os.path.join(_figures_dir, f"figure_{{_figure_count[0]:03d}}.png")
fig.savefig(fig_path, dpi=120, bbox_inches="tight", facecolor="white")
_figure_count[0] += 1
plt.close("all")
plt.show = _capture_show
# Helper to display DataFrames
def display(obj):
if isinstance(obj, pd.DataFrame):
_tables.append({{
"columns": obj.columns.tolist(),
"data": obj.head(50).values.tolist(),
"shape": list(obj.shape)
}})
print(obj.to_string(max_rows=20))
else:
print(repr(obj))
try:
# Execute user code
exec_globals = {{
"pd": pd, "np": np, "plt": plt,
"display": display, "print": print,
"__builtins__": __builtins__,
}}
exec("""
{code.replace(chr(92), chr(92)+chr(92)).replace('"', chr(92)+'"')}
""", exec_globals)
# Save any unsaved figures
if plt.get_fignums():
_capture_show()
except Exception as e:
import traceback
traceback.print_exc()
# Save structured output
try:
output = {{
"tables": _tables,
"result": "",
"figure_count": _figure_count[0]
}}
with open(_output_path, "w", encoding="utf-8") as f:
json.dump(output, f, default=str)
except:
pass
'''
return wrapper
def execute_code_for_user(user_id: str, code: str, context: Dict = None) -> Dict[str, Any]:
"""Convenience function to execute code for a user."""
executor = CodeExecutor(user_id=user_id)
return executor.execute(code, context)
|