dataset/dataset.json

[
  {
    "task_type": "easy",
    "pr": {
      "id": "2",
      "title": "Missing import",
      "description": "Forgot to import module",
      "language": "python",
      "diffs": [
        {
          "file_name": "main.py",
          "diff": "print(datetime.now())"
        }
      ]
    },
    "ground_truth": {
      "issues": ["missing import datetime"],
      "decision": "reject",
      "fix": "from datetime import datetime\nprint(datetime.now())"
    }
  },
  {
    "task_type": "easy",
    "pr": {
      "id": "9",
      "title": "Missing return statement",
      "description": "Function does not return value",
      "language": "python",
      "diffs": [
        {
          "file_name": "utils.py",
          "diff": "def add(a, b):\n    result = a + b"
        }
      ]
    },
    "ground_truth": {
      "issues": ["missing return statement"],
      "decision": "reject",
      "fix": "def add(a, b):\n    result = a + b\n    return result"
    }
  },
  {
    "task_type": "easy",
    "pr": {
      "id": "10",
      "title": "Wrong comparison operator",
      "description": "Fix equality check",
      "language": "python",
      "diffs": [
        {
          "file_name": "check.py",
          "diff": "if x = 10:\n    print('ten')"
        }
      ]
    },
    "ground_truth": {
      "issues": ["assignment instead of comparison"],
      "decision": "reject",
      "fix": "if x == 10:\n    print('ten')"
    }
  },
  {
    "task_type": "easy",
    "pr": {
      "id": "11",
      "title": "Undefined variable",
      "description": "Variable used before assignment",
      "language": "python",
      "diffs": [
        {
          "file_name": "app.py",
          "diff": "def greet():\n    print(message)\n    message = 'Hello'"
        }
      ]
    },
    "ground_truth": {
      "issues": ["undefined variable", "variable used before assignment"],
      "decision": "reject",
      "fix": "def greet():\n    message = 'Hello'\n    print(message)"
    }
  },
  {
    "task_type": "easy",
    "pr": {
      "id": "12",
      "title": "Clean utility function",
      "description": "Simple string helper",
      "language": "python",
      "diffs": [
        {
          "file_name": "utils.py",
          "diff": "def to_upper(s):\n    return s.upper()"
        }
      ]
    },
    "ground_truth": {
      "issues": [],
      "decision": "approve",
      "fix": ""
    }
  },
  {
    "task_type": "medium",
    "pr": {
      "id": "3",
      "title": "Division function",
      "description": "Handles division",
      "language": "python",
      "diffs": [
        {
          "file_name": "math.py",
          "diff": "def divide(a,b): return a/b"
        }
      ]
    },
    "ground_truth": {
      "issues": ["division by zero"],
      "decision": "reject",
      "fix": "def divide(a, b):\n    if b == 0:\n        return None\n    return a / b"
    }
  },
  {
    "task_type": "medium",
    "pr": {
      "id": "4",
      "title": "Inefficient loop",
      "description": "Optimizing search",
      "language": "python",
      "diffs": [
        {
          "file_name": "search.py",
          "diff": "for i in range(len(arr)):\n    if arr[i] == target:\n        return True"
        }
      ]
    },
    "ground_truth": {
      "issues": ["inefficient loop"],
      "decision": "approve",
      "fix": "return target in arr"
    }
  },
  {
    "task_type": "medium",
    "pr": {
      "id": "13",
      "title": "Mutable default argument",
      "description": "Function with default list argument",
      "language": "python",
      "diffs": [
        {
          "file_name": "helper.py",
          "diff": "def append_item(item, lst=[]):\n    lst.append(item)\n    return lst"
        }
      ]
    },
    "ground_truth": {
      "issues": ["mutable default argument"],
      "decision": "reject",
      "fix": "def append_item(item, lst=None):\n    if lst is None:\n        lst = []\n    lst.append(item)\n    return lst"
    }
  },
  {
    "task_type": "medium",
    "pr": {
      "id": "14",
      "title": "Unhandled exception",
      "description": "File read without error handling",
      "language": "python",
      "diffs": [
        {
          "file_name": "reader.py",
          "diff": "def read_file(path):\n    with open(path) as f:\n        return f.read()"
        }
      ]
    },
    "ground_truth": {
      "issues": ["unhandled exception", "missing error handling"],
      "decision": "reject",
      "fix": "def read_file(path):\n    try:\n        with open(path) as f:\n            return f.read()\n    except FileNotFoundError:\n        return None"
    }
  },
  {
    "task_type": "medium",
    "pr": {
      "id": "15",
      "title": "Integer overflow risk",
      "description": "Large number multiplication",
      "language": "python",
      "diffs": [
        {
          "file_name": "compute.py",
          "diff": "def factorial(n):\n    result = 1\n    for i in range(1, n+1):\n        result *= i\n    return result"
        }
      ]
    },
    "ground_truth": {
      "issues": ["missing input validation"],
      "decision": "reject",
      "fix": "def factorial(n):\n    if n < 0:\n        raise ValueError('n must be non-negative')\n    result = 1\n    for i in range(1, n+1):\n        result *= i\n    return result"
    }
  },
  {
    "task_type": "hard",
    "pr": {
      "id": "6",
      "title": "Authentication logic",
      "description": "Adds login system",
      "language": "python",
      "diffs": [
        {
          "file_name": "auth.py",
          "diff": "def login(password):\n    if password == 'admin123':\n        return True"
        }
      ]
    },
    "ground_truth": {
      "issues": ["hardcoded password", "security vulnerability"],
      "decision": "reject",
      "fix": "import bcrypt\n\ndef login(password, hashed_password):\n    return bcrypt.checkpw(password.encode(), hashed_password)"
    }
  },
  {
    "task_type": "hard",
    "pr": {
      "id": "7",
      "title": "SQL query issue",
      "description": "Fetch user data",
      "language": "python",
      "diffs": [
        {
          "file_name": "db.py",
          "diff": "query = \"SELECT * FROM users WHERE id = \" + user_id"
        }
      ]
    },
    "ground_truth": {
      "issues": ["sql injection"],
      "decision": "reject",
      "fix": "query = \"SELECT * FROM users WHERE id = %s\"\ncursor.execute(query, (user_id,))"
    }
  },
  {
    "task_type": "hard",
    "pr": {
      "id": "8",
      "title": "Cross-file null bug",
      "description": "User fetch logic",
      "language": "python",
      "diffs": [
        {
          "file_name": "service.py",
          "diff": "def get_user(id):\n    return db[id]"
        },
        {
          "file_name": "controller.py",
          "diff": "user = get_user(None)"
        }
      ]
    },
    "ground_truth": {
      "issues": ["invalid input", "null handling"],
      "decision": "reject",
      "fix": "def get_user(id):\n    if id is None:\n        raise ValueError('id must not be None')\n    return db[id]\n\nuser = get_user(user_id)"
    }
  },
  {
    "task_type": "hard",
    "pr": {
      "id": "16",
      "title": "Race condition in counter",
      "description": "Shared counter increment",
      "language": "python",
      "diffs": [
        {
          "file_name": "counter.py",
          "diff": "counter = 0\n\ndef increment():\n    global counter\n    counter += 1"
        }
      ]
    },
    "ground_truth": {
      "issues": ["race condition", "thread safety"],
      "decision": "reject",
      "fix": "import threading\n\ncounter = 0\nlock = threading.Lock()\n\ndef increment():\n    global counter\n    with lock:\n        counter += 1"
    }
  },
  {
    "task_type": "hard",
    "pr": {
      "id": "17",
      "title": "Insecure deserialization",
      "description": "Load user data from request",
      "language": "python",
      "diffs": [
        {
          "file_name": "api.py",
          "diff": "import pickle\n\ndef load_user(data):\n    return pickle.loads(data)"
        }
      ]
    },
    "ground_truth": {
      "issues": ["insecure deserialization", "security vulnerability"],
      "decision": "reject",
      "fix": "import json\n\ndef load_user(data):\n    return json.loads(data)"
    }
  },
  {
    "task_type": "hard",
    "pr": {
      "id": "18",
      "title": "Path traversal vulnerability",
      "description": "Serve user requested files",
      "language": "python",
      "diffs": [
        {
          "file_name": "files.py",
          "diff": "def read_file(filename):\n    with open('/var/data/' + filename) as f:\n        return f.read()"
        }
      ]
    },
    "ground_truth": {
      "issues": ["path traversal", "security vulnerability"],
      "decision": "reject",
      "fix": "import os\n\ndef read_file(filename):\n    base = '/var/data/'\n    full_path = os.path.realpath(os.path.join(base, filename))\n    if not full_path.startswith(base):\n        raise ValueError('Invalid file path')\n    with open(full_path) as f:\n        return f.read()"
    }
  }
]