Base guard logic on module availability (#692)

* Replace `backed_guard` with `requires_module`

* Remove backend_guard

test/test_librosa_compatibility.py

@@ -7,9 +7,9 @@ import torch
 from torch.testing._internal.common_utils import TestCase
 import torchaudio
 import torchaudio.functional as F
-from torchaudio.common_utils import _check_module_exists
+from torchaudio._internal.module_utils import is_module_available
 
-LIBROSA_AVAILABLE = _check_module_exists('librosa')
+LIBROSA_AVAILABLE = is_module_available('librosa')
 
 if LIBROSA_AVAILABLE:
     import numpy as np

torchaudio/__init__.py

@@ -14,12 +14,12 @@ from torchaudio import (
 )
 from torchaudio._backend import (
     check_input,
-    _audio_backend_guard,
     _get_audio_backend_module,
     get_audio_backend,
     set_audio_backend,
 )
 from torchaudio._soundfile_backend import SignalInfo, EncodingInfo
+from torchaudio._internal import module_utils as _mod_utils
 
 try:
     from .version import __version__, git_version  # noqa: F401
@@ -27,6 +27,10 @@ except ImportError:
     pass
 
 
+if _mod_utils.is_module_available('torchaudio._torchaudio'):
+    from . import _torchaudio
+
+
 def load(filepath: Union[str, Path],
          out: Optional[Tensor] = None,
          normalization: Union[bool, float, Callable] = True,
@@ -75,7 +79,6 @@ def load(filepath: Union[str, Path],
         1.
 
     """
-
     return _get_audio_backend_module().load(
         filepath,
         out=out,
@@ -124,7 +127,7 @@ def save(filepath: str, src: Tensor, sample_rate: int, precision: int = 16, chan
     )
 
 
-@_audio_backend_guard("sox")
+@_mod_utils.requires_module('torchaudio._torchaudio')
 def save_encinfo(filepath: str,
                  src: Tensor,
                  channels_first: bool = True,
@@ -192,8 +195,6 @@ def save_encinfo(filepath: str,
         src = src.transpose(1, 0)
     # save data to file
     src = src.contiguous()
-
-    from . import _torchaudio
     _torchaudio.write_audio_file(filepath, src, signalinfo, encodinginfo, filetype)
 
 
@@ -211,11 +212,10 @@ def info(filepath: str) -> Tuple[SignalInfo, EncodingInfo]:
          >>> si, ei = torchaudio.info('foo.wav')
          >>> rate, channels, encoding = si.rate, si.channels, ei.encoding
     """
-
     return _get_audio_backend_module().info(filepath)
 
 
-@_audio_backend_guard("sox")
+@_mod_utils.requires_module('torchaudio._torchaudio')
 def sox_signalinfo_t() -> SignalInfo:
     r"""Create a sox_signalinfo_t object. This object can be used to set the sample
     rate, number of channels, length, bit precision and headroom multiplier
@@ -235,12 +235,10 @@ def sox_signalinfo_t() -> SignalInfo:
         >>> si.precision = 16
         >>> si.length = 0
     """
-
-    from . import _torchaudio
     return _torchaudio.sox_signalinfo_t()
 
 
-@_audio_backend_guard("sox")
+@_mod_utils.requires_module('torchaudio._torchaudio')
 def sox_encodinginfo_t() -> EncodingInfo:
     r"""Create a sox_encodinginfo_t object.  This object can be used to set the encoding
     type, bit precision, compression factor, reverse bytes, reverse nibbles,
@@ -270,8 +268,6 @@ def sox_encodinginfo_t() -> EncodingInfo:
         >>> ei.opposite_endian = torchaudio.get_sox_bool(0)
 
     """
-
-    from . import _torchaudio
     ei = _torchaudio.sox_encodinginfo_t()
     sdo = get_sox_option_t(2)  # sox_default_option
     ei.reverse_bytes = sdo
@@ -280,7 +276,7 @@ def sox_encodinginfo_t() -> EncodingInfo:
     return ei
 
 
-@_audio_backend_guard("sox")
+@_mod_utils.requires_module('torchaudio._torchaudio')
 def get_sox_encoding_t(i: int = None) -> EncodingInfo:
     r"""Get enum of sox_encoding_t for sox encodings.
 
@@ -291,8 +287,6 @@ def get_sox_encoding_t(i: int = None) -> EncodingInfo:
     Returns:
         sox_encoding_t: A sox_encoding_t type for output encoding
     """
-
-    from . import _torchaudio
     if i is None:
         # one can see all possible values using the .__members__ attribute
         return _torchaudio.sox_encoding_t
@@ -300,7 +294,7 @@ def get_sox_encoding_t(i: int = None) -> EncodingInfo:
         return _torchaudio.sox_encoding_t(i)
 
 
-@_audio_backend_guard("sox")
+@_mod_utils.requires_module('torchaudio._torchaudio')
 def get_sox_option_t(i: int = 2) -> Any:
     r"""Get enum of sox_option_t for sox encodinginfo options.
 
@@ -311,15 +305,13 @@ def get_sox_option_t(i: int = 2) -> Any:
     Returns:
         sox_option_t: A sox_option_t type
     """
-
-    from . import _torchaudio
     if i is None:
         return _torchaudio.sox_option_t
     else:
         return _torchaudio.sox_option_t(i)
 
 
-@_audio_backend_guard("sox")
+@_mod_utils.requires_module('torchaudio._torchaudio')
 def get_sox_bool(i: int = 0) -> Any:
     r"""Get enum of sox_bool for sox encodinginfo options.
 
@@ -331,8 +323,6 @@ def get_sox_bool(i: int = 0) -> Any:
     Returns:
         sox_bool: A sox_bool type
     """
-
-    from . import _torchaudio
     if i is None:
         return _torchaudio.sox_bool
     else:
@@ -350,7 +340,7 @@ _SOX_SUCCESS_CODE = 0
 # https://fossies.org/dox/sox-14.4.2/sox_8h.html#a8e07e80cebeff3339265d89c387cea93a9ef2b87ec303edfe40751d9a85fadeeb
 
 
-@_audio_backend_guard("sox")
+@_mod_utils.requires_module("torchaudio._torchaudio")
 def initialize_sox() -> int:
     """Initialize sox for use with effects chains.
 
@@ -370,7 +360,6 @@ def initialize_sox() -> int:
     if _SOX_INITIALIZED is None:
         raise RuntimeError('SoX effects chain has been already shut down. Can not initialize again.')
     if not _SOX_INITIALIZED:
-        from . import _torchaudio
         code = _torchaudio.initialize_sox()
         if code == _SOX_SUCCESS_CODE:
             _SOX_INITIALIZED = True
@@ -379,7 +368,7 @@ def initialize_sox() -> int:
     return _SOX_SUCCESS_CODE
 
 
-@_audio_backend_guard("sox")
+@_mod_utils.requires_module("torchaudio._torchaudio")
 def shutdown_sox() -> int:
     """Showdown sox for effects chain.
 
@@ -394,7 +383,6 @@ def shutdown_sox() -> int:
     """
     global _SOX_INITIALIZED
     if _SOX_INITIALIZED:
-        from . import _torchaudio
         code = _torchaudio.shutdown_sox()
         if code == _SOX_INITIALIZED:
             _SOX_INITIALIZED = None

torchaudio/_backend.py

@@ -45,22 +45,6 @@ def _get_audio_backend_module() -> Any:
     return _audio_backends[backend]
 
 
-def _audio_backend_guard(backends: Union[str, List[str]]) -> Any:
-
-    if isinstance(backends, str):
-        backends = [backends]
-
-    def decorator(func):
-        @wraps(func)
-        def wrapper(*args, **kwargs):
-            if get_audio_backend() not in backends:
-                raise RuntimeError("Function {} requires backend to be one of {}.".format(func.__name__, backends))
-            return func(*args, **kwargs)
-        return wrapper
-
-    return decorator
-
-
 def check_input(src: Tensor) -> None:
     if not torch.is_tensor(src):
         raise TypeError('Expected a tensor, got %s' % type(src))

torchaudio/common_utils.py → torchaudio/_internal/module_utils.py

 import importlib.util
+from functools import wraps
 
 
-def _check_module_exists(*modules: str) -> bool:
+def is_module_available(*modules: str) -> bool:
     r"""Returns if a top-level module with :attr:`name` exists *without**
     importing it. This is generally safer than try-catch block around a
     `import X`. It avoids third party libraries breaking assumptions of some of
@@ -9,3 +10,26 @@ def _check_module_exists(*modules: str) -> bool:
     (see librosa/#747, torchvision/#544).
     """
     return all(importlib.util.find_spec(m) is not None for m in modules)
+
+
+def requires_module(*modules: str):
+    """Decorate function to give error message if invoked without required optional modules.
+
+    This decorator is to give better error message to users rather
+    than raising ``NameError:  name 'module' is not defined`` at random places.
+    """
+    missing = [m for m in modules if not is_module_available(m)]
+
+    if not missing:
+        # fall through. If all the modules are available, no need to decorate
+        def decorator(func):
+            return func
+    else:
+        req = f'module: {missing[0]}' if len(missing) == 1 else f'modules: {missing}'
+
+        def decorator(func):
+            @wraps(func)
+            def wrapped(*args, **kwargs):
+                raise RuntimeError(f'{func.__module__}.{func.__name__} requires {req}')
+            return wrapped
+    return decorator

torchaudio/_sox_backend.py

@@ -3,10 +3,16 @@ from typing import Optional, Tuple
 
 import torch
 from torch import Tensor
+
 import torchaudio
+from torchaudio._internal import module_utils as _mod_utils
 from torchaudio._soundfile_backend import SignalInfo, EncodingInfo
 
+if _mod_utils.is_module_available('torchaudio._torchaudio'):
+    from . import _torchaudio
+
 
+@_mod_utils.requires_module('torchaudio._torchaudio')
 def load(filepath: str,
          out: Optional[Tensor] = None,
          normalization: bool = True,
@@ -35,7 +41,6 @@ def load(filepath: str,
     if offset < 0:
         raise ValueError("Expected positive offset value")
 
-    from . import _torchaudio
     sample_rate = _torchaudio.read_audio_file(
         filepath,
         out,
@@ -53,6 +58,7 @@ def load(filepath: str,
     return out, sample_rate
 
 
+@_mod_utils.requires_module('torchaudio._torchaudio')
 def save(filepath: str, src: Tensor, sample_rate: int, precision: int = 16, channels_first: bool = True) -> None:
     r"""See torchaudio.save"""
 
@@ -65,8 +71,7 @@ def save(filepath: str, src: Tensor, sample_rate: int, precision: int = 16, chan
     return torchaudio.save_encinfo(filepath, src, channels_first, si)
 
 
+@_mod_utils.requires_module('torchaudio._torchaudio')
 def info(filepath: str) -> Tuple[SignalInfo, EncodingInfo]:
     r"""See torchaudio.info"""
-
-    from . import _torchaudio
     return _torchaudio.get_info(filepath)

torchaudio/kaldi_io.py

@@ -5,11 +5,9 @@ from typing import Any, Callable, Iterable, Tuple
 
 import torch
 from torch import Tensor
-from torchaudio.common_utils import _check_module_exists
+from torchaudio._internal import module_utils as _mod_utils
 
-_KALDI_IO_AVAILABLE = _check_module_exists('kaldi_io', 'numpy')
-
-if _KALDI_IO_AVAILABLE:
+if _mod_utils.is_module_available('kaldi_io', 'numpy'):
     import numpy as np
     import kaldi_io
 
@@ -38,15 +36,13 @@ def _convert_method_output_to_tensor(file_or_fd: Any,
     Returns:
         Iterable[Tuple[str, Tensor]]: The string is the key and the tensor is vec/mat
     """
-    if not _KALDI_IO_AVAILABLE:
-        raise ImportError('Could not import kaldi_io. Did you install it?')
-
     for key, np_arr in fn(file_or_fd):
         if convert_contiguous:
             np_arr = np.ascontiguousarray(np_arr)
         yield key, torch.from_numpy(np_arr)
 
 
+@_mod_utils.requires_module('kaldi_io', 'numpy')
 def read_vec_int_ark(file_or_fd: Any) -> Iterable[Tuple[str, Tensor]]:
     r"""Create generator of (key,vector<int>) tuples, which reads from the ark file/stream.
 
@@ -66,6 +62,7 @@ def read_vec_int_ark(file_or_fd: Any) -> Iterable[Tuple[str, Tensor]]:
     return _convert_method_output_to_tensor(file_or_fd, kaldi_io.read_vec_int_ark, convert_contiguous=True)
 
 
+@_mod_utils.requires_module('kaldi_io', 'numpy')
 def read_vec_flt_scp(file_or_fd: Any) -> Iterable[Tuple[str, Tensor]]:
     r"""Create generator of (key,vector<float32/float64>) tuples, read according to Kaldi scp.
 
@@ -82,6 +79,7 @@ def read_vec_flt_scp(file_or_fd: Any) -> Iterable[Tuple[str, Tensor]]:
     return _convert_method_output_to_tensor(file_or_fd, kaldi_io.read_vec_flt_scp)
 
 
+@_mod_utils.requires_module('kaldi_io', 'numpy')
 def read_vec_flt_ark(file_or_fd: Any) -> Iterable[Tuple[str, Tensor]]:
     r"""Create generator of (key,vector<float32/float64>) tuples, which reads from the ark file/stream.
 
@@ -98,6 +96,7 @@ def read_vec_flt_ark(file_or_fd: Any) -> Iterable[Tuple[str, Tensor]]:
     return _convert_method_output_to_tensor(file_or_fd, kaldi_io.read_vec_flt_ark)
 
 
+@_mod_utils.requires_module('kaldi_io', 'numpy')
 def read_mat_scp(file_or_fd: Any) -> Iterable[Tuple[str, Tensor]]:
     r"""Create generator of (key,matrix<float32/float64>) tuples, read according to Kaldi scp.
 
@@ -114,6 +113,7 @@ def read_mat_scp(file_or_fd: Any) -> Iterable[Tuple[str, Tensor]]:
     return _convert_method_output_to_tensor(file_or_fd, kaldi_io.read_mat_scp)
 
 
+@_mod_utils.requires_module('kaldi_io', 'numpy')
 def read_mat_ark(file_or_fd: Any) -> Iterable[Tuple[str, Tensor]]:
     r"""Create generator of (key,matrix<float32/float64>) tuples, which reads from the ark file/stream.
 

torchaudio/sox_effects.py

@@ -3,10 +3,14 @@ from typing import Any, Callable, List, Optional, Tuple, Union
 import torch
 import torchaudio
 from torch import Tensor
-from torchaudio._backend import _audio_backend_guard
 
+from torchaudio._internal import module_utils as _mod_utils
 
-@_audio_backend_guard("sox")
+if _mod_utils.is_module_available('torchaudio._torchaudio'):
+    from . import _torchaudio
+
+
+@_mod_utils.requires_module('torchaudio._torchaudio')
 def effect_names() -> List[str]:
     """Gets list of valid sox effect names
 
@@ -15,12 +19,10 @@ def effect_names() -> List[str]:
     Example
         >>> EFFECT_NAMES = torchaudio.sox_effects.effect_names()
     """
-
-    from . import _torchaudio
     return _torchaudio.get_effect_names()
 
 
-@_audio_backend_guard("sox")
+@_mod_utils.requires_module('torchaudio._torchaudio')
 def SoxEffect():
     r"""Create an object for passing sox effect information between python and c++
 
@@ -28,8 +30,6 @@ def SoxEffect():
         SoxEffect: An object with the following attributes: ename (str) which is the
         name of effect, and eopts (List[str]) which is a list of effect options.
     """
-
-    from . import _torchaudio
     return _torchaudio.SoxEffect()
 
 
@@ -123,7 +123,7 @@ class SoxEffectsChain(object):
         e.eopts = eargs
         self.chain.append(e)
 
-    @_audio_backend_guard("sox")
+    @_mod_utils.requires_module('torchaudio._torchaudio')
     def sox_build_flow_effects(self,
                                out: Optional[Tensor] = None) -> Tuple[Tensor, int]:
         r"""Build effects chain and flow effects from input file to output tensor
@@ -150,7 +150,6 @@ class SoxEffectsChain(object):
         # print("effect options:", [x.eopts for x in self.chain])
 
         torchaudio.initialize_sox()
-        from . import _torchaudio
         sr = _torchaudio.build_flow_effects(self.input_file,
                                             out,
                                             self.channels_first,