HKUDS/Vibe-Trading · error · ValueError

malformed expression: {exc}

Error message

malformed expression: {exc}

What it means

_safe_arith parses the expression with ast.parse(mode='eval'); a SyntaxError is wrapped into this ValueError so callers get a uniform validation exception. It fires before any evaluation, purely on unparseable input such as unbalanced parentheses or empty/garbage strings.

Source

Thrown at agent/src/tools/financial_rigor_tool.py:135

    The expression is parsed and evaluated recursively with Decimal arithmetic,
    so ``0.1 + 0.2`` is exactly ``0.3`` — no IEEE-754 drift, and no ``eval``.
    Only numbers and the operators ``+ - * /`` (with optional unary sign) are
    permitted; any other AST node raises ``ValueError``.

    Args:
        expr: Arithmetic expression string, e.g. ``"510 * 9.11e9"``.

    Returns:
        The exact Decimal result.

    Raises:
        ValueError: If the expression is malformed or contains a disallowed
            element.
    """
    try:
        tree = ast.parse(expr, mode="eval")
    except SyntaxError as exc:
        raise ValueError(f"malformed expression: {exc}") from exc
    return _eval_arith_node(tree.body)


# ---------------------------------------------------------------------------
# Core verification routines (pure, return structured dicts, no I/O)
# ---------------------------------------------------------------------------

def verify_market_cap(
    price: Any, shares: Any, reported_cap: Any, currency: str = "",
) -> dict[str, Any]:
    """Verify ``market cap = price × shares`` against a reported value.

    Args:
        price: Current share price.
        shares: Total share count.
        reported_cap: The market-cap figure being checked.
        currency: Optional currency label for display only.

View on GitHub (pinned to 80ffdda44c)

Solutions

1. Sanitize the input (strip commas/currency symbols, normalize quotes) before calling
2. Validate the expression with ast.parse in the caller for a cleaner upstream error
3. Retry with a corrected expression when the tool reports a parse failure

Example fix

# before
exact_calc("1,000 * 0.05")
# after
exact_calc("1000 * 0.05")

import ast
expr = user_input.replace(",", "").replace("$", "")
try:
    ast.parse(expr, mode="eval")
except SyntaxError as e:
    raise ValueError(f"bad formula at col {e.offset}: {e.msg}") from e

try:
    value = _safe_arith(expr)
except ValueError as e:
    if str(e).startswith("malformed expression"):
        return None / re-ask user for a corrected formula
    raise

When it happens

Trigger: Calling exact_calc or _safe_arith with '', '2 +', '(1+2', or non-Python text like 'two plus two'.

Common situations: User-typed formulas, truncated LLM output, or copy-paste artifacts (smart quotes, thousand separators like '1,000')

Understand the failure class

• Parsing and encoding errors: unexpected token, malformed input — why parsers reject input and how to find the real culprit.

Related errors

• unsupported operator: {type(node.op).__name__}
• unsupported unary operator: {type(node.op).__name__}
• invalid alpha_id
• alpha_id not found
• invalid period: {exc}

AI-assisted analysis of HKUDS/Vibe-Trading@80ffdda44c (2026-08-28). Data as JSON: /api/errors/effa8584b5c13f5f. Report an issue: GitHub.