Avoid eager optional comparator imports

.pre-commit-config.yaml

@@ -1,3 +1,5 @@
+exclude: ^(tests/|code_to_optimize/|pie_test_set/|experiments/|\.codeflash/)
+
 repos:
   - repo: https://github.com/astral-sh/ruff-pre-commit
     rev: v0.15.8
@@ -9,7 +11,8 @@ repos:
     hooks:
       - id: mypy
         name: mypy
-        entry: uv run mypy --non-interactive --config-file pyproject.toml
+        entry: uv run mypy --non-interactive --config-file pyproject.toml @mypy_allowlist.txt
         language: system
+        pass_filenames: false
         types: [python]
         require_serial: true

codeflash/verification/comparator.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 import _thread
 import array
 import ast
@@ -15,6 +17,8 @@
 import weakref
 import xml.etree.ElementTree as ET
 from collections import ChainMap, OrderedDict, deque
+from functools import cache
+from importlib import import_module
 from importlib.util import find_spec
 from typing import Any, Optional
 
@@ -35,33 +39,6 @@
 HAS_NUMBA = find_spec("numba") is not None
 HAS_PYARROW = find_spec("pyarrow") is not None
 
-if HAS_NUMPY:
-    import numpy as np
-if HAS_SCIPY:
-    import scipy  # type: ignore  # noqa: PGH003
-if HAS_JAX:
-    import jax  # type: ignore  # noqa: PGH003
-    import jax.numpy as jnp  # type: ignore  # noqa: PGH003
-if HAS_XARRAY:
-    import xarray  # type: ignore  # noqa: PGH003
-if HAS_TENSORFLOW:
-    import tensorflow as tf  # type: ignore  # noqa: PGH003
-if HAS_SQLALCHEMY:
-    import sqlalchemy  # type: ignore  # noqa: PGH003
-if HAS_PYARROW:
-    import pyarrow as pa  # type: ignore  # noqa: PGH003
-if HAS_PANDAS:
-    import pandas  # noqa: ICN001
-if HAS_TORCH:
-    import torch  # type: ignore  # noqa: PGH003
-if HAS_NUMBA:
-    import numba  # type: ignore  # noqa: PGH003
-    from numba.core.dispatcher import Dispatcher  # type: ignore  # noqa: PGH003
-    from numba.typed import Dict as NumbaDict  # type: ignore  # noqa: PGH003
-    from numba.typed import List as NumbaList  # type: ignore  # noqa: PGH003
-if HAS_PYRSISTENT:
-    import pyrsistent  # type: ignore  # noqa: PGH003
-
 # Pattern to match pytest temp directories: /tmp/pytest-of-<user>/pytest-<N>/
 # These paths vary between test runs but are logically equivalent
 PYTEST_TEMP_PATH_PATTERN = re.compile(r"/tmp/pytest-of-[^/]+/pytest-\d+/")  # noqa: S108
@@ -117,6 +94,18 @@
 )
 
 
+@cache
+def _optional_module(module_name: str) -> Optional[Any]:
+    if find_spec(module_name) is None:
+        return None
+    return import_module(module_name)
+
+
+def _object_module_matches(obj: Any, module_prefix: str) -> bool:
+    module_name = type(obj).__module__
+    return module_name == module_prefix or module_name.startswith(f"{module_prefix}.")
+
+
 def _normalize_temp_path(path: str) -> str:
     """Normalize temporary file paths by replacing session-specific components.
 
@@ -133,7 +122,7 @@ def _is_temp_path(s: str) -> bool:
     return PYTEST_TEMP_PATH_PATTERN.search(s) is not None or PYTHON_TEMPFILE_PATTERN.search(s) is not None
 
 
-def _extract_exception_from_message(msg: str) -> Optional[BaseException]:  # noqa: FA100
+def _extract_exception_from_message(msg: str) -> Optional[BaseException]:
     """Try to extract a wrapped exception type from an error message.
 
     Looks for patterns like "got ExceptionType('..." that indicate a wrapped exception.
@@ -153,7 +142,7 @@ def _extract_exception_from_message(msg: str) -> Optional[BaseException]:  # noq
     return None
 
 
-def _get_wrapped_exception(exc: BaseException) -> Optional[BaseException]:  # noqa: FA100
+def _get_wrapped_exception(exc: BaseException) -> Optional[BaseException]:
     """Get the wrapped exception if this is a simple wrapper.
 
     Returns the inner exception if:
@@ -272,22 +261,26 @@ def comparator(orig: Any, new: Any, superset_obj: bool = False) -> bool:
                 return False
             return comparator(orig_referent, new_referent, superset_obj)
 
-        if HAS_JAX:
+        if HAS_JAX and (_object_module_matches(orig, "jax") or _object_module_matches(orig, "jaxlib")):
+            jax = _optional_module("jax")
+            jnp = _optional_module("jax.numpy")
             # Handle JAX arrays first to avoid boolean context errors in other conditions
-            if isinstance(orig, jax.Array):
+            if jax is not None and jnp is not None and isinstance(orig, jax.Array):
                 if orig.dtype != new.dtype:
                     return False
                 if orig.shape != new.shape:
                     return False
                 return bool(jnp.allclose(orig, new, equal_nan=True))
 
         # Handle xarray objects before numpy to avoid boolean context errors
-        if HAS_XARRAY:
-            if isinstance(orig, (xarray.Dataset, xarray.DataArray)):
+        if HAS_XARRAY and _object_module_matches(orig, "xarray"):
+            xarray = _optional_module("xarray")
+            if xarray is not None and isinstance(orig, (xarray.Dataset, xarray.DataArray)):
                 return orig.identical(new)
 
         # Handle TensorFlow objects early to avoid boolean context errors
-        if HAS_TENSORFLOW:
+        if HAS_TENSORFLOW and _object_module_matches(orig, "tensorflow"):
+            tf = _optional_module("tensorflow")
             if isinstance(orig, tf.Tensor):
                 if orig.dtype != new.dtype:
                     return False
@@ -313,9 +306,7 @@ def comparator(orig: Any, new: Any, superset_obj: bool = False) -> bool:
                 if not comparator(orig.dense_shape.numpy(), new.dense_shape.numpy(), superset_obj):
                     return False
                 return comparator(orig.indices.numpy(), new.indices.numpy(), superset_obj) and comparator(
-                    orig.values.numpy(),  # noqa: PD011
-                    new.values.numpy(),  # noqa: PD011
-                    superset_obj,
+                    orig.values.numpy(), new.values.numpy(), superset_obj
                 )
 
             if isinstance(orig, tf.RaggedTensor):
@@ -325,7 +316,8 @@ def comparator(orig: Any, new: Any, superset_obj: bool = False) -> bool:
                     return False
                 return comparator(orig.to_list(), new.to_list(), superset_obj)
 
-        if HAS_SQLALCHEMY:
+        if HAS_SQLALCHEMY and (hasattr(orig, "_sa_instance_state") or _object_module_matches(orig, "sqlalchemy")):
+            sqlalchemy = _optional_module("sqlalchemy")
             try:
                 insp = sqlalchemy.inspection.inspect(orig)
                 insp = sqlalchemy.inspection.inspect(new)
@@ -342,7 +334,8 @@ def comparator(orig: Any, new: Any, superset_obj: bool = False) -> bool:
                 pass
 
         # scipy condition because dok_matrix type is also a instance of dict, but dict comparison doesn't work for it
-        if isinstance(orig, dict) and not (HAS_SCIPY and isinstance(orig, scipy.sparse.spmatrix)):
+        scipy_sparse = _optional_module("scipy.sparse") if HAS_SCIPY and _object_module_matches(orig, "scipy") else None
+        if isinstance(orig, dict) and not (scipy_sparse is not None and isinstance(orig, scipy_sparse.spmatrix)):
             if superset_obj:
                 return all(k in new and comparator(v, new[k], superset_obj) for k, v in orig.items())
             if len(orig) != len(new):
@@ -366,7 +359,8 @@ def comparator(orig: Any, new: Any, superset_obj: bool = False) -> bool:
         if isinstance(orig, _DICT_ITEMS_TYPE):
             return comparator(dict(orig), dict(new), superset_obj)
 
-        if HAS_NUMPY:
+        np = _optional_module("numpy") if HAS_NUMPY and _object_module_matches(orig, "numpy") else None
+        if np is not None:
             if isinstance(orig, (np.datetime64, np.timedelta64)):
                 # Handle NaT (Not a Time) - numpy's equivalent of NaN for datetime
                 if np.isnat(orig) and np.isnat(new):
@@ -420,14 +414,15 @@ def comparator(orig: Any, new: Any, superset_obj: bool = False) -> bool:
                 new_state = new.get_state(legacy=False)
                 return comparator(orig_state, new_state, superset_obj)
 
-        if HAS_SCIPY and isinstance(orig, scipy.sparse.spmatrix):
+        if scipy_sparse is not None and isinstance(orig, scipy_sparse.spmatrix):
             if orig.dtype != new.dtype:
                 return False
             if orig.get_shape() != new.get_shape():
                 return False
             return (orig != new).nnz == 0
 
-        if HAS_PYARROW:
+        if HAS_PYARROW and _object_module_matches(orig, "pyarrow"):
+            pa = _optional_module("pyarrow")
             if isinstance(orig, pa.Table):
                 if orig.schema != new.schema:
                     return False
@@ -469,7 +464,8 @@ def comparator(orig: Any, new: Any, superset_obj: bool = False) -> bool:
             if isinstance(orig, (pa.Schema, pa.Field, pa.DataType)):
                 return bool(orig.equals(new))
 
-        if HAS_PANDAS:
+        if HAS_PANDAS and _object_module_matches(orig, "pandas"):
+            pandas = _optional_module("pandas")
             if isinstance(
                 orig, (pandas.DataFrame, pandas.Series, pandas.Index, pandas.Categorical, pandas.arrays.SparseArray)
             ):
@@ -489,17 +485,18 @@ def comparator(orig: Any, new: Any, superset_obj: bool = False) -> bool:
 
         # This should be at the end of all numpy checking
         try:
-            if HAS_NUMPY and np.isnan(orig):
+            if np is not None and np.isnan(orig):
                 return np.isnan(new)
         except Exception:
             pass
         try:
-            if HAS_NUMPY and np.isinf(orig):
+            if np is not None and np.isinf(orig):
                 return np.isinf(new)
         except Exception:
             pass
 
-        if HAS_TORCH:
+        if HAS_TORCH and _object_module_matches(orig, "torch"):
+            torch = _optional_module("torch")
             if isinstance(orig, torch.Tensor):
                 if orig.dtype != new.dtype:
                     return False
@@ -517,15 +514,23 @@ def comparator(orig: Any, new: Any, superset_obj: bool = False) -> bool:
             if isinstance(orig, torch.device):
                 return orig == new
 
-        if HAS_NUMBA:
+        if HAS_NUMBA and _object_module_matches(orig, "numba"):
+            numba = _optional_module("numba")
+            numba_dispatcher = _optional_module("numba.core.dispatcher")
+            numba_typed = _optional_module("numba.typed")
+            if numba is None or numba_dispatcher is None or numba_typed is None:
+                return False
+            dispatcher = numba_dispatcher.Dispatcher
+            numba_dict = numba_typed.Dict
+            numba_list = numba_typed.List
             # Handle numba typed List
-            if isinstance(orig, NumbaList):
+            if isinstance(orig, numba_list):
                 if len(orig) != len(new):
                     return False
                 return all(comparator(elem1, elem2, superset_obj) for elem1, elem2 in zip(orig, new))
 
             # Handle numba typed Dict
-            if isinstance(orig, NumbaDict):
+            if isinstance(orig, numba_dict):
                 if superset_obj:
                     # Allow new dict to have more keys, but all orig keys must exist with equal values
                     return all(key in new and comparator(orig[key], new[key], superset_obj) for key in orig)
@@ -543,12 +548,15 @@ def comparator(orig: Any, new: Any, superset_obj: bool = False) -> bool:
                 return orig == new
 
             # Handle numba JIT-compiled functions (CPUDispatcher, etc.)
-            if isinstance(orig, Dispatcher):
+            if isinstance(orig, dispatcher):
                 # Compare by identity of the underlying Python function
                 # Two JIT functions are equal if they wrap the same Python function
                 return orig.py_func is new.py_func
 
-        if HAS_PYRSISTENT:
+        if HAS_PYRSISTENT and _object_module_matches(orig, "pyrsistent"):
+            pyrsistent = _optional_module("pyrsistent")
+            if pyrsistent is None:
+                return False
             if isinstance(
                 orig,
                 (

pyproject.toml

@@ -218,7 +218,7 @@ warn_unreachable = true
 install_types = true
 plugins = ["pydantic.mypy"]
 
-exclude = ["tests/", "code_to_optimize/", "pie_test_set/", "experiments/"]
+exclude = ["tests/", "code_to_optimize/", "pie_test_set/", "experiments/", "^\\.codeflash/"]
 
 [[tool.mypy.overrides]]
 module = ["jedi", "jedi.api.classes", "inquirer", "inquirer.themes", "numba", "dill"]
@@ -379,4 +379,3 @@ markers = [
 [build-system]
 requires = ["hatchling", "uv-dynamic-versioning"]
 build-backend = "hatchling.build"
-

tests/test_comparator.py

@@ -5,6 +5,7 @@
 import datetime
 import decimal
 import re
+import subprocess
 import sys
 import uuid
 import weakref
@@ -29,6 +30,27 @@
 from codeflash.verification.equivalence import compare_test_results
 
 
+def test_comparator_import_does_not_load_optional_numeric_modules() -> None:
+    result = subprocess.run(
+        [
+            sys.executable,
+            "-c",
+            (
+                "import sys; "
+                "import codeflash.verification.comparator; "
+                "loaded = {'numpy', 'pandas', 'xarray', 'numexpr'} & set(sys.modules); "
+                "print(','.join(sorted(loaded))); "
+                "raise SystemExit(bool(loaded))"
+            ),
+        ],
+        capture_output=True,
+        text=True,
+        encoding="utf-8",
+        check=False,
+    )
+    assert result.returncode == 0, result.stdout + result.stderr
+
+
 def test_basic_python_objects() -> None:
     a = 5
     b = 5