1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349
|
"""
lazy_loader
===========
Makes it easy to load subpackages and functions on demand.
"""
import ast
import importlib
import importlib.util
import os
import sys
import threading
import types
import warnings
__version__ = "0.4"
__all__ = ["attach", "load", "attach_stub"]
threadlock = threading.Lock()
def attach(package_name, submodules=None, submod_attrs=None):
"""Attach lazily loaded submodules, functions, or other attributes.
Typically, modules import submodules and attributes as follows::
import mysubmodule
import anothersubmodule
from .foo import someattr
The idea is to replace a package's `__getattr__`, `__dir__`, and
`__all__`, such that all imports work exactly the way they would
with normal imports, except that the import occurs upon first use.
The typical way to call this function, replacing the above imports, is::
__getattr__, __dir__, __all__ = lazy.attach(
__name__,
['mysubmodule', 'anothersubmodule'],
{'foo': ['someattr']}
)
This functionality requires Python 3.7 or higher.
Parameters
----------
package_name : str
Typically use ``__name__``.
submodules : set
List of submodules to attach.
submod_attrs : dict
Dictionary of submodule -> list of attributes / functions.
These attributes are imported as they are used.
Returns
-------
__getattr__, __dir__, __all__
"""
if submod_attrs is None:
submod_attrs = {}
if submodules is None:
submodules = set()
else:
submodules = set(submodules)
attr_to_modules = {
attr: mod for mod, attrs in submod_attrs.items() for attr in attrs
}
__all__ = sorted(submodules | attr_to_modules.keys())
def __getattr__(name):
if name in submodules:
return importlib.import_module(f"{package_name}.{name}")
elif name in attr_to_modules:
submod_path = f"{package_name}.{attr_to_modules[name]}"
submod = importlib.import_module(submod_path)
attr = getattr(submod, name)
# If the attribute lives in a file (module) with the same
# name as the attribute, ensure that the attribute and *not*
# the module is accessible on the package.
if name == attr_to_modules[name]:
pkg = sys.modules[package_name]
pkg.__dict__[name] = attr
return attr
else:
raise AttributeError(f"No {package_name} attribute {name}")
def __dir__():
return __all__
if os.environ.get("EAGER_IMPORT", ""):
for attr in set(attr_to_modules.keys()) | submodules:
__getattr__(attr)
return __getattr__, __dir__, list(__all__)
class DelayedImportErrorModule(types.ModuleType):
def __init__(self, frame_data, *args, message, **kwargs):
self.__frame_data = frame_data
self.__message = message
super().__init__(*args, **kwargs)
def __getattr__(self, x):
if x in ("__class__", "__file__", "__frame_data", "__message"):
super().__getattr__(x)
else:
fd = self.__frame_data
raise ModuleNotFoundError(
f"{self.__message}\n\n"
"This error is lazily reported, having originally occured in\n"
f' File {fd["filename"]}, line {fd["lineno"]}, in {fd["function"]}\n\n'
f'----> {"".join(fd["code_context"] or "").strip()}'
)
def load(fullname, *, require=None, error_on_import=False):
"""Return a lazily imported proxy for a module.
We often see the following pattern::
def myfunc():
import numpy as np
np.norm(...)
....
Putting the import inside the function prevents, in this case,
`numpy`, from being imported at function definition time.
That saves time if `myfunc` ends up not being called.
This `load` function returns a proxy module that, upon access, imports
the actual module. So the idiom equivalent to the above example is::
np = lazy.load("numpy")
def myfunc():
np.norm(...)
....
The initial import time is fast because the actual import is delayed
until the first attribute is requested. The overall import time may
decrease as well for users that don't make use of large portions
of your library.
Warning
-------
While lazily loading *sub*packages technically works, it causes the
package (that contains the subpackage) to be eagerly loaded even
if the package is already lazily loaded.
So, you probably shouldn't use subpackages with this `load` feature.
Instead you should encourage the package maintainers to use the
`lazy_loader.attach` to make their subpackages load lazily.
Parameters
----------
fullname : str
The full name of the module or submodule to import. For example::
sp = lazy.load('scipy') # import scipy as sp
require : str
A dependency requirement as defined in PEP-508. For example::
"numpy >=1.24"
If defined, the proxy module will raise an error if the installed
version does not satisfy the requirement.
error_on_import : bool
Whether to postpone raising import errors until the module is accessed.
If set to `True`, import errors are raised as soon as `load` is called.
Returns
-------
pm : importlib.util._LazyModule
Proxy module. Can be used like any regularly imported module.
Actual loading of the module occurs upon first attribute request.
"""
with threadlock:
module = sys.modules.get(fullname)
have_module = module is not None
# Most common, short-circuit
if have_module and require is None:
return module
if "." in fullname:
msg = (
"subpackages can technically be lazily loaded, but it causes the "
"package to be eagerly loaded even if it is already lazily loaded."
"So, you probably shouldn't use subpackages with this lazy feature."
)
warnings.warn(msg, RuntimeWarning)
spec = None
if not have_module:
spec = importlib.util.find_spec(fullname)
have_module = spec is not None
if not have_module:
not_found_message = f"No module named '{fullname}'"
elif require is not None:
try:
have_module = _check_requirement(require)
except ModuleNotFoundError as e:
raise ValueError(
f"Found module '{fullname}' but cannot test "
"requirement '{require}'. "
"Requirements must match distribution name, not module name."
) from e
not_found_message = f"No distribution can be found matching '{require}'"
if not have_module:
if error_on_import:
raise ModuleNotFoundError(not_found_message)
import inspect
try:
parent = inspect.stack()[1]
frame_data = {
"filename": parent.filename,
"lineno": parent.lineno,
"function": parent.function,
"code_context": parent.code_context,
}
return DelayedImportErrorModule(
frame_data,
"DelayedImportErrorModule",
message=not_found_message,
)
finally:
del parent
if spec is not None:
module = importlib.util.module_from_spec(spec)
sys.modules[fullname] = module
loader = importlib.util.LazyLoader(spec.loader)
loader.exec_module(module)
return module
def _check_requirement(require: str) -> bool:
"""Verify that a package requirement is satisfied
If the package is required, a ``ModuleNotFoundError`` is raised
by ``importlib.metadata``.
Parameters
----------
require : str
A dependency requirement as defined in PEP-508
Returns
-------
satisfied : bool
True if the installed version of the dependency matches
the specified version, False otherwise.
"""
import packaging.requirements
try:
import importlib.metadata as importlib_metadata
except ImportError: # PY37
import importlib_metadata
req = packaging.requirements.Requirement(require)
return req.specifier.contains(
importlib_metadata.version(req.name),
prereleases=True,
)
class _StubVisitor(ast.NodeVisitor):
"""AST visitor to parse a stub file for submodules and submod_attrs."""
def __init__(self):
self._submodules = set()
self._submod_attrs = {}
def visit_ImportFrom(self, node: ast.ImportFrom):
if node.level != 1:
raise ValueError(
"Only within-module imports are supported (`from .* import`)"
)
if node.module:
attrs: list = self._submod_attrs.setdefault(node.module, [])
aliases = [alias.name for alias in node.names]
if "*" in aliases:
raise ValueError(
"lazy stub loader does not support star import "
f"`from {node.module} import *`"
)
attrs.extend(aliases)
else:
self._submodules.update(alias.name for alias in node.names)
def attach_stub(package_name: str, filename: str):
"""Attach lazily loaded submodules, functions from a type stub.
This is a variant on ``attach`` that will parse a `.pyi` stub file to
infer ``submodules`` and ``submod_attrs``. This allows static type checkers
to find imports, while still providing lazy loading at runtime.
Parameters
----------
package_name : str
Typically use ``__name__``.
filename : str
Path to `.py` file which has an adjacent `.pyi` file.
Typically use ``__file__``.
Returns
-------
__getattr__, __dir__, __all__
The same output as ``attach``.
Raises
------
ValueError
If a stub file is not found for `filename`, or if the stubfile is formmated
incorrectly (e.g. if it contains an relative import from outside of the module)
"""
stubfile = (
filename if filename.endswith("i") else f"{os.path.splitext(filename)[0]}.pyi"
)
if not os.path.exists(stubfile):
raise ValueError(f"Cannot load imports from non-existent stub {stubfile!r}")
with open(stubfile) as f:
stub_node = ast.parse(f.read())
visitor = _StubVisitor()
visitor.visit(stub_node)
return attach(package_name, visitor._submodules, visitor._submod_attrs)
|