human-to-sympy 0.1.0

Parse human-typed equation strings into canonical SymPy expressions

human-to-sympy

Parse human-typed equation strings into canonical SymPy expression strings.

human-to-sympy accepts the kind of math people actually type — ^ for power, familiar function names, bare identifiers — and returns the corresponding SymPy expression as a normalized string. It does not perform equivalence comparison; for that, use eq-equiv.

What it does

• Replaces ^ with **.
• Splits on = when present and returns the right-hand side (for generate_sympy_rhs*) or both sides (for generate_sympy_equation).
• Distinguishes function-call sites from bare identifiers and binds them to sympy.Function and sympy.Symbol respectively, while leaving SymPy builtins (pi, E, log, sin, ...) alone.
• Returns a typed result object preserving the failure reason on parse error.

API

from human_to_sympy import (
    SympyEquationGenerationResult,
    SympyGenerationResult,
    check_symbols,
    generate_sympy_equation,
    generate_sympy_rhs,
    generate_sympy_rhs_with_error,
)

• generate_sympy_rhs(text) -> str | None — convenience accessor for the RHS string.
• generate_sympy_rhs_with_error(text) -> SympyGenerationResult — same, but preserves the parse error.
• generate_sympy_equation(text) -> SympyEquationGenerationResult — returns both sides of an lhs = rhs equation.
• check_symbols(text, variables) -> list[str] — warns when expression free symbols are not declared in the supplied variable list.

Example

from human_to_sympy import generate_sympy_rhs

assert generate_sympy_rhs("Fa = 1 / (2 * pi * Ta)") == "1/(2*pi*Ta)"
assert generate_sympy_rhs("(2*pi*Ta*f)^2") == "(2*pi*Ta*f)**2"