pysmt

library0.9.6pypypi✓ verified 87d ago

PySMT is a Python library designed for solver-agnostic manipulation and solving of Satisfiability Modulo Theories (SMT) formulae. It provides a unified API to define, manipulate, and solve SMT problems using various SMT-LIB compliant solvers. The library is actively maintained, with regular releases, and its current stable version is 0.9.6.

ai-ml data

pip install pysmt

package ↗homepage ↗

INSTALL

IMPORT

SIG · PYSMT

pysmt

ai-mlpythonv0.9.6

Install

1.8s avg

Import

101ms

Disk

19MB

Pass rate

10/ 10

Env Coverage10 / 10

glibc

3.9–3.13

musl

3.9–3.13

Install & Compatibility

Where this runs

tested against v0.9.6 · pip install

no network on importno background threads

Install × environment matrix

Each cell = how many times install + import succeeded across repeated harness runs. Partial = flaky.
glibc = Debian/Ubuntu slim · musl = Alpine Linux

musl

py 3.10–3.920 runs

installs and imports cleanly · install 0.0s · import 0.107s · 20.6MB

glibc

py 3.10–3.920 runs

installs and imports cleanly · install 1.8s · import 0.095s · 21MB

19MB installed

● package 19MB

Code

Verified usage

Verified import paths — ran on the pinned version, not inferred.

Symbol, And, Not, is_sat

✓ from pysmt.shortcuts import Symbol, And, Not, is_sat

The `pysmt.shortcuts` module provides the most commonly used functions and operators for formula construction and solving.

INT, BOOL

✓ from pysmt.typing import INT, BOOL

Data types for SMT variables (symbols) are defined in `pysmt.typing`.

Solver

✓ from pysmt.shortcuts import Solver

For incremental solving or explicit solver control, use the `Solver` class from `shortcuts`.

This quickstart demonstrates defining Boolean SMT variables, constructing a logical formula, checking its satisfiability using a shortcut, performing variable substitution, and using an explicit solver to extract a model.

from pysmt.shortcuts import Symbol, And, Not, is_sat, Solver from pysmt.typing import BOOL # Define Boolean symbols varA = Symbol("A", BOOL) varB = Symbol("B", BOOL) # Construct a formula: (A AND (NOT B)) f = And([varA, Not(varB)]) # Check satisfiability using a shortcut res_f = is_sat(f) print(f"Formula '{f}' is SAT? {res_f}") # Perform a substitution: replace B with A in f, resulting in (A AND (NOT A)) g = f.substitute({varB: varA}) res_g = is_sat(g) print(f"Formula '{g}' is SAT? {res_g}") # Example with explicit solver and model extraction with Solver(name="z3", suppress_warnings=True) as solver: solver.add_assertion(f) if solver.solve(): print(f"Model for '{f}':") for s in [varA, varB]: print(f" {s} = {solver.get_value(s)}") else: print(f"'{f}' is UNSAT")

pysmt --version

Debug

Known issues

breakingPython 2.7 support was officially dropped in PySMT version 0.9.0. Versions 0.9.0 and later only support Python 3+. The last release to support Python 2.7 was 0.8.0.

fix

Upgrade to Python 3.5+ (recommended) or higher. For Python 2.7, use PySMT <0.9.0.

affects: >=0.9.0

gotchaWhen `pysmt.shortcuts` is imported, infix notation (e.g., `a & b`) is enabled by default. While convenient for quick experimentation, it can make complex code less clear by blurring the line between Python and SMT operators. For clarity, prefer explicit textual operators (e.g., `And(a, b)`).

fix

Import specific textual operators from `pysmt.shortcuts` (e.g., `from pysmt.shortcuts import And, Or, Not`) and use them directly for better code readability and to avoid unexpected operator precedence issues.

affects: All versions

gotchaSMT solvers are not direct Python dependencies of PySMT. While `pip install pysmt` installs the library, you need to explicitly install at least one solver (e.g., Z3, MathSAT) using `python -m pysmt install --<solver-name>` to perform actual solving. Solvers need to have their Python bindings accessible, potentially requiring `PYTHONPATH` adjustments for custom installations.

fix

Run `python -m pysmt install --<solver-name>` (e.g., `python -m pysmt install --z3`) to install a solver. If issues persist, check `pysmt-install --check` and ensure your `PYTHONPATH` includes the solver bindings directory, especially for non-standard installations.

affects: All versions

breakingThe classes `Integer`, `Fraction`, and `Numerals` were moved to `pysmt.constants` in version 0.6.0. Methods from `pysmt.utils` also moved to `pysmt.constants`. The `pysmt.numeral` module was removed.

fix

Update imports and references to use `pysmt.constants.Fraction` (or `Integer`, `Numerals`) instead of previous locations like `pysmt.numeral` or `pysmt.utils`.

affects: >=0.6.0

deprecatedThe methods `Solver.declare_variable` and `Solver.set_options` were removed in PySMT version 0.9.5.

fix

Remove calls to `Solver.declare_variable` as variables are typically declared via `Symbol()` and added to the solver with `add_assertion()`. For solver options, consult the documentation for solver-specific configuration or use the `Solver` constructor's `options` argument if available.

affects: >=0.9.5

Errors

Common errors & fixes

pysmt.exceptions.NoSolverAvailableError: No Solver is available

This error occurs when PySMT attempts to perform a solving operation (e.g., `is_sat()`, `Solver().solve()`) but cannot find an installed and accessible SMT solver. This could be because no solver was installed, or its Python bindings are not correctly configured.

fix

Install a solver using `python -m pysmt install --<solver-name>` (e.g., `python -m pysmt install --z3`). Verify installation with `pysmt-install --check`. If the solver was installed manually or in a non-standard location, ensure its Python bindings directory is added to your `PYTHONPATH`.

TypeError: unsupported operand type(s) for &: 'FNode' and 'bool'

This typically happens when mixing PySMT's infix operators (which expect `FNode` objects on both sides) with standard Python boolean literals or expressions without explicit conversion. For example, `f & True` might raise this if `True` isn't wrapped as a PySMT Boolean constant.

fix

Ensure all operands in SMT expressions are `FNode` objects. For Python boolean literals, use `Bool(True)` or `Bool(False)` from `pysmt.shortcuts`. Better yet, use explicit logical operators like `And()`, `Or()`, `Not()` to avoid ambiguity and improve clarity.

Solver installed, but `pysmt-install --check` or `is_sat()` still reports it as unavailable.

Even if `pysmt-install --<solver-name>` completes without errors, the Python bindings for the solver might not be correctly located by PySMT, particularly if they are installed in a non-default location or if the `PYTHONPATH` environment variable is not properly configured.

fix

After running `python -m pysmt install`, check the output for the location of the installed bindings. If they are not in a standard `site-packages` directory, you might need to manually add the path to your `PYTHONPATH` environment variable. For example, `export PYTHONPATH=$HOME/.smt_solvers/python_bindings:$PYTHONPATH`.

Upgrade

Version history

0.9.6latest on PyPI · released Jun 24, 2024

Audit

Dependencies

gmpy2optionalOptional, for efficient multi-precision number handling. Defaults to Python's fractions module if not installed.

CythonoptionalOptional, used to improve performance of some internal components, e.g., SMT-LIB parser.

Agent activity

16 hits · last 30 days

node

OpenAI (training)

Resources

packagehttps://pypi.org/project/pysmt/ ↗

homepagehttp://www.pysmt.org ↗