Merge pull request #436 from menshikh-iv/mokeypath4pathlib

Add pathlib monkeypatch with replacement of `pathlib.Path.open`

README.rst

@@ -382,6 +382,25 @@ If your file object doesn't have one, set the ``.name`` attribute to an appropri
 Furthermore, that value has to end with a **known** file extension (see the ``register_compressor`` function).
 Otherwise, the transparent decompression will not occur.
 
+Drop-in replacement of ``pathlib.Path.open``
+--------------------------------------------
+
+Now you can natively use ``smart_open.open`` with your ``Path`` objects.
+You can't transparently read text from compressed file with original ``Path.open``, but can after ``patch_pathlib``.
+
+.. code-block:: python
+
+  >>> from pathlib import Path
+  >>> from smart_open.smart_open_lib import patch_pathlib
+  >>>
+  >>> _ = patch_pathlib()  # replace `Path.open` with `smart_open.open`
+  >>>
+  >>> path = Path("smart_open/tests/test_data/crime-and-punishment.txt.gz")
+  >>>
+  >>> with path.open("r") as infile:
+  ...     print(infile.readline()[:41])
+  В начале июля, в чрезвычайно жаркое время
+
 Comments, bug reports
 =====================
 

smart_open/smart_open_lib.py

@@ -928,3 +928,28 @@ def _encoding_wrapper(fileobj, mode, encoding=None, errors=None):
     if mode[0] in ('w', 'a') or mode.endswith('+'):
         fileobj = codecs.getwriter(encoding)(fileobj, **kw)
     return fileobj
+
+
+class patch_pathlib(object):
+    """Replace `Path.open` with `smart_open.open`"""
+
+    def __init__(self):
+        self.old_impl = _patch_pathlib(open)
+
+    def __enter__(self):
+        return self
+
+    def __exit__(self, exc_type, exc_val, exc_tb):
+        _patch_pathlib(self.old_impl)
+
+
+def _patch_pathlib(func):
+    """Replace `Path.open` with `func`"""
+    pathlib = sys.modules.get("pathlib", None)
+
+    if not pathlib:
+        raise RuntimeError("Can't patch 'pathlib.Path.open', you should import 'pathlib' first")
+
+    old_impl = pathlib.Path.open
+    pathlib.Path.open = func
+    return old_impl

smart_open/tests/test_smart_open.py

@@ -13,6 +13,7 @@
 import tempfile
 import os
 import hashlib
+import pathlib
 
 import boto3
 import mock
@@ -24,6 +25,7 @@
 import smart_open
 from smart_open import smart_open_lib
 from smart_open import webhdfs
+from smart_open.smart_open_lib import patch_pathlib, _patch_pathlib
 
 logger = logging.getLogger(__name__)
 
@@ -287,6 +289,35 @@ def test_gs_uri_contains_slash(self):
         self.assertEqual(parsed_uri.bucket_id, "mybucket")
         self.assertEqual(parsed_uri.blob_id, "mydir/myblob")
 
+    def test_pathlib_monkeypath(self):
+        assert pathlib.Path.open != smart_open.open
+
+        with patch_pathlib():
+            assert pathlib.Path.open == smart_open.open
+
+        assert pathlib.Path.open != smart_open.open
+
+        obj = patch_pathlib()
+        assert pathlib.Path.open == smart_open.open
+
+        _patch_pathlib(obj.old_impl)
+        assert pathlib.Path.open != smart_open.open
+
+    def test_pathlib_monkeypath_read_gz(self):
+        path = pathlib.Path(CURR_DIR) / 'test_data' / 'crime-and-punishment.txt.gz'
+
+        # Check that standart implementation can't work with gzip
+        with path.open("r") as infile:
+            with self.assertRaises(Exception) as context:
+                lines = infile.readlines()
+
+        # Check that out implementation works with gzip
+        obj = patch_pathlib()
+        with path.open("r") as infile:
+            lines = infile.readlines()
+
+        _patch_pathlib(obj.old_impl)
+
 
 class SmartOpenHttpTest(unittest.TestCase):
     """