[ENH] Add read support for Google Cloud Storage (#20729)

bnaul · TomAugspurger · commit 95427d5983ef · 2018-06-26T09:59:23.000-05:00
* Google Cloud Storage support using gcsfs
diff --git a/ci/appveyor-27.yaml b/ci/appveyor-27.yaml
@@ -6,6 +6,7 @@ dependencies:
   - beautifulsoup4
   - bottleneck
   - dateutil
+  - gcsfs
   - html5lib
   - jinja2=2.8
   - lxml
diff --git a/ci/check_imports.py b/ci/check_imports.py
@@ -5,6 +5,7 @@
 
 blacklist = {
     'bs4',
+    'gcsfs',
     'html5lib',
     'ipython',
     'jinja2'
diff --git a/ci/circle-36-locale_slow.yaml b/ci/circle-36-locale_slow.yaml
@@ -5,6 +5,7 @@ channels:
 dependencies:
   - beautifulsoup4
   - cython
+  - gcsfs
   - html5lib
   - ipython
   - jinja2
diff --git a/ci/requirements-optional-conda.txt b/ci/requirements-optional-conda.txt
@@ -3,6 +3,7 @@ blosc
 bottleneck
 fastparquet
 feather-format
+gcsfs
 html5lib
 ipython>=5.6.0
 ipykernel
diff --git a/ci/requirements-optional-pip.txt b/ci/requirements-optional-pip.txt
@@ -5,6 +5,7 @@ blosc
 bottleneck
 fastparquet
 feather-format
+gcsfs
 html5lib
 ipython>=5.6.0
 ipykernel
diff --git a/ci/travis-27.yaml b/ci/travis-27.yaml
@@ -9,6 +9,7 @@ dependencies:
   - fastparquet
   - feather-format
   - flake8=3.4.1
+  - gcsfs
   - html5lib
   - ipython
   - jemalloc=4.5.0.post
diff --git a/ci/travis-36.yaml b/ci/travis-36.yaml
@@ -8,6 +8,7 @@ dependencies:
   - dask
   - fastparquet
   - feather-format
+  - gcsfs
   - geopandas
   - html5lib
   - ipython
diff --git a/doc/source/install.rst b/doc/source/install.rst
@@ -276,6 +276,7 @@ Optional Dependencies
 * `Jinja2 <http://jinja.pocoo.org/>`__: Template engine for conditional HTML formatting.
 * `s3fs <http://s3fs.readthedocs.io/>`__: necessary for Amazon S3 access (s3fs >= 0.0.7).
 * `blosc <https://pypi.org/project/blosc>`__: for msgpack compression using ``blosc``
+* `gcsfs <http://gcsfs.readthedocs.io/>`__: necessary for Google Cloud Storage access (gcsfs >= 0.1.0).
 * One of
   `qtpy  <https://github.com/spyder-ide/qtpy>`__ (requires PyQt or PySide),
   `PyQt5 <https://www.riverbankcomputing.com/software/pyqt/download5>`__,
diff --git a/doc/source/whatsnew/v0.24.0.txt b/doc/source/whatsnew/v0.24.0.txt
@@ -18,7 +18,7 @@ Other Enhancements
 - :func:`Series.mode` and :func:`DataFrame.mode` now support the ``dropna`` parameter which can be used to specify whether NaN/NaT values should be considered (:issue:`17534`)
 - :func:`to_csv` now supports ``compression`` keyword when a file handle is passed. (:issue:`21227`)
 - :meth:`Index.droplevel` is now implemented also for flat indexes, for compatibility with :class:`MultiIndex` (:issue:`21115`)
-
+- Added support for reading from Google Cloud Storage via the ``gcsfs`` library (:issue:`19454`)
 
 .. _whatsnew_0240.api_breaking:
 
diff --git a/pandas/conftest.py b/pandas/conftest.py
@@ -1,3 +1,5 @@
+import importlib
+
 import pytest
 
 import numpy as np
@@ -249,3 +251,17 @@ def any_int_dtype(request):
     """
 
     return request.param
+
+
+@pytest.fixture
+def mock():
+    """
+    Fixture providing the 'mock' module.
+
+    Uses 'unittest.mock' for Python 3. Attempts to import the 3rd party 'mock'
+    package for Python 2, skipping if not present.
+    """
+    if PY3:
+        return importlib.import_module("unittest.mock")
+    else:
+        return pytest.importorskip("mock")
diff --git a/pandas/io/common.py b/pandas/io/common.py
@@ -88,7 +88,7 @@ def _is_url(url):
     """
     try:
         return parse_url(url).scheme in _VALID_URLS
-    except:
+    except Exception:
         return False
 
 
@@ -165,7 +165,15 @@ def is_s3_url(url):
     """Check for an s3, s3n, or s3a url"""
     try:
         return parse_url(url).scheme in ['s3', 's3n', 's3a']
-    except:  # noqa
+    except Exception:
+        return False
+
+
+def is_gcs_url(url):
+    """Check for a gcs url"""
+    try:
+        return parse_url(url).scheme in ['gcs', 'gs']
+    except Exception:
         return False
 
 
@@ -208,6 +216,13 @@ def get_filepath_or_buffer(filepath_or_buffer, encoding=None,
                                          compression=compression,
                                          mode=mode)
 
+    if is_gcs_url(filepath_or_buffer):
+        from pandas.io import gcs
+        return gcs.get_filepath_or_buffer(filepath_or_buffer,
+                                          encoding=encoding,
+                                          compression=compression,
+                                          mode=mode)
+
     if isinstance(filepath_or_buffer, (compat.string_types,
                                        compat.binary_type,
                                        mmap.mmap)):
diff --git a/pandas/io/excel.py b/pandas/io/excel.py
@@ -46,7 +46,7 @@
 io : string, path object (pathlib.Path or py._path.local.LocalPath),
     file-like object, pandas ExcelFile, or xlrd workbook.
     The string could be a URL. Valid URL schemes include http, ftp, s3,
-    and file. For file URLs, a host is expected. For instance, a local
+    gcs, and file. For file URLs, a host is expected. For instance, a local
     file could be file://localhost/path/to/workbook.xlsx
 sheet_name : string, int, mixed list of strings/ints, or None, default 0
 
diff --git a/pandas/io/gcs.py b/pandas/io/gcs.py
@@ -0,0 +1,16 @@
+""" GCS support for remote file interactivity """
+try:
+    import gcsfs
+except ImportError:
+    raise ImportError("The gcsfs library is required to handle GCS files")
+
+
+def get_filepath_or_buffer(filepath_or_buffer, encoding=None,
+                           compression=None, mode=None):
+
+    if mode is None:
+        mode = 'rb'
+
+    fs = gcsfs.GCSFileSystem()
+    filepath_or_buffer = fs.open(filepath_or_buffer, mode)
+    return filepath_or_buffer, None, compression, True
diff --git a/pandas/io/json/json.py b/pandas/io/json/json.py
@@ -231,9 +231,9 @@ def read_json(path_or_buf=None, orient=None, typ='frame', dtype=True,
     Parameters
     ----------
     path_or_buf : a valid JSON string or file-like, default: None
-        The string could be a URL. Valid URL schemes include http, ftp, s3, and
-        file. For file URLs, a host is expected. For instance, a local file
-        could be ``file://localhost/path/to/table.json``
+        The string could be a URL. Valid URL schemes include http, ftp, s3,
+        gcs, and file. For file URLs, a host is expected. For instance, a local
+        file could be ``file://localhost/path/to/table.json``
 
     orient : string,
         Indication of expected JSON string format.
diff --git a/pandas/tests/dtypes/test_inference.py b/pandas/tests/dtypes/test_inference.py
@@ -20,7 +20,7 @@
                     DatetimeIndex, TimedeltaIndex, Timestamp,
                     Panel, Period, Categorical, isna, Interval,
                     DateOffset)
-from pandas.compat import u, PY2, PY3, StringIO, lrange
+from pandas.compat import u, PY2, StringIO, lrange
 from pandas.core.dtypes import inference
 from pandas.core.dtypes.common import (
     is_timedelta64_dtype,
@@ -128,7 +128,7 @@ def test_is_dict_like_fails(ll):
     assert not inference.is_dict_like(ll)
 
 
-def test_is_file_like():
+def test_is_file_like(mock):
     class MockFile(object):
         pass
 
@@ -166,10 +166,7 @@ class MockFile(object):
     # Iterator but no read / write attributes
     data = [1, 2, 3]
     assert not is_file(data)
-
-    if PY3:
-        from unittest import mock
-        assert not is_file(mock.Mock())
+    assert not is_file(mock.Mock())
 
 
 @pytest.mark.parametrize(
diff --git a/pandas/tests/io/parser/common.py b/pandas/tests/io/parser/common.py
@@ -1546,7 +1546,7 @@ def test_file_handles(self):
                     assert not m.closed
                 m.close()
 
-    def test_invalid_file_buffer(self):
+    def test_invalid_file_buffer(self, mock):
         # see gh-15337
 
         class InvalidBuffer(object):
@@ -1577,11 +1577,8 @@ def seek(self, pos, whence=0):
 
         tm.assert_frame_equal(result, expected)
 
-        if PY3:
-            from unittest import mock
-
-            with tm.assert_raises_regex(ValueError, msg):
-                self.read_csv(mock.Mock())
+        with tm.assert_raises_regex(ValueError, msg):
+            self.read_csv(mock.Mock())
 
     @tm.capture_stderr
     def test_skip_bad_lines(self):
diff --git a/pandas/tests/io/test_gcs.py b/pandas/tests/io/test_gcs.py
@@ -0,0 +1,47 @@
+import numpy as np
+import pytest
+
+from pandas import DataFrame, date_range, read_csv
+from pandas.compat import StringIO
+from pandas.io.common import is_gcs_url
+from pandas.util import _test_decorators as td
+from pandas.util.testing import assert_frame_equal
+
+
+def test_is_gcs_url():
+    assert is_gcs_url("gcs://pandas/somethingelse.com")
+    assert is_gcs_url("gs://pandas/somethingelse.com")
+    assert not is_gcs_url("s3://pandas/somethingelse.com")
+
+
+@td.skip_if_no('gcsfs')
+def test_read_csv_gcs(mock):
+    df1 = DataFrame({'int': [1, 3], 'float': [2.0, np.nan], 'str': ['t', 's'],
+                     'dt': date_range('2018-06-18', periods=2)})
+    with mock.patch('gcsfs.GCSFileSystem') as MockFileSystem:
+        instance = MockFileSystem.return_value
+        instance.open.return_value = StringIO(df1.to_csv(index=False))
+        df2 = read_csv('gs://test/test.csv', parse_dates=['dt'])
+
+    assert_frame_equal(df1, df2)
+
+
+@td.skip_if_no('gcsfs')
+def test_gcs_get_filepath_or_buffer(mock):
+    df1 = DataFrame({'int': [1, 3], 'float': [2.0, np.nan], 'str': ['t', 's'],
+                     'dt': date_range('2018-06-18', periods=2)})
+    with mock.patch('pandas.io.gcs.get_filepath_or_buffer') as MockGetFilepath:
+        MockGetFilepath.return_value = (StringIO(df1.to_csv(index=False)),
+                                        None, None, False)
+        df2 = read_csv('gs://test/test.csv', parse_dates=['dt'])
+
+    assert_frame_equal(df1, df2)
+    assert MockGetFilepath.called
+
+
+@pytest.mark.skipif(td.safe_import('gcsfs'),
+                    reason='Only check when gcsfs not installed')
+def test_gcs_not_present_exception():
+    with pytest.raises(ImportError) as e:
+        read_csv('gs://test/test.csv')
+        assert 'gcsfs library is required' in str(e.value)
diff --git a/pandas/util/_print_versions.py b/pandas/util/_print_versions.py
@@ -96,6 +96,7 @@ def show_versions(as_json=False):
         ("fastparquet", lambda mod: mod.__version__),
         ("pandas_gbq", lambda mod: mod.__version__),
         ("pandas_datareader", lambda mod: mod.__version__),
+        ("gcsfs", lambda mod: mod.__version__),
     ]
 
     deps_blob = list()
