oasis-data-manager 0.1.3__tar.gz → 0.1.5__tar.gz

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (38) hide show
  1. {oasis-data-manager-0.1.3/oasis_data_manager.egg-info → oasis-data-manager-0.1.5}/PKG-INFO +2 -4
  2. oasis-data-manager-0.1.5/oasis_data_manager/__init__.py +1 -0
  3. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/oasis_data_manager/df_reader/backends/base.py +14 -3
  4. oasis-data-manager-0.1.5/oasis_data_manager/df_reader/backends/pyarrow.py +73 -0
  5. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/oasis_data_manager/df_reader/reader.py +6 -0
  6. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/oasis_data_manager/filestore/backends/aws_s3.py +0 -1
  7. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5/oasis_data_manager.egg-info}/PKG-INFO +2 -4
  8. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/oasis_data_manager.egg-info/SOURCES.txt +1 -1
  9. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/oasis_data_manager.egg-info/requires.txt +2 -2
  10. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/setup.cfg +2 -0
  11. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/setup.py +1 -5
  12. oasis-data-manager-0.1.3/LICENSE +0 -28
  13. oasis-data-manager-0.1.3/oasis_data_manager/__init__.py +0 -1
  14. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/CHANGELOG.rst +0 -0
  15. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/MANIFEST.in +0 -0
  16. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/README.md +0 -0
  17. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/oasis_data_manager/complex/__init__.py +0 -0
  18. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/oasis_data_manager/complex/complex.py +0 -0
  19. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/oasis_data_manager/complex/examples.py +0 -0
  20. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/oasis_data_manager/config.py +0 -0
  21. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/oasis_data_manager/df_reader/__init__.py +0 -0
  22. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/oasis_data_manager/df_reader/backends/__init__.py +0 -0
  23. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/oasis_data_manager/df_reader/backends/dask.py +0 -0
  24. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/oasis_data_manager/df_reader/backends/pandas.py +0 -0
  25. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/oasis_data_manager/df_reader/config.py +0 -0
  26. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/oasis_data_manager/df_reader/exceptions.py +0 -0
  27. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/oasis_data_manager/errors/__init__.py +0 -0
  28. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/oasis_data_manager/filestore/__init__.py +0 -0
  29. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/oasis_data_manager/filestore/backends/__init__.py +0 -0
  30. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/oasis_data_manager/filestore/backends/azure_abfs.py +0 -0
  31. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/oasis_data_manager/filestore/backends/base.py +0 -0
  32. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/oasis_data_manager/filestore/backends/local.py +0 -0
  33. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/oasis_data_manager/filestore/config.py +0 -0
  34. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/oasis_data_manager/filestore/filestore.py +0 -0
  35. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/oasis_data_manager/filestore/log.py +0 -0
  36. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/oasis_data_manager.egg-info/dependency_links.txt +0 -0
  37. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/oasis_data_manager.egg-info/top_level.txt +0 -0
  38. {oasis-data-manager-0.1.3 → oasis-data-manager-0.1.5}/requirements-package.in +0 -0
@@ -1,21 +1,19 @@
1
1
  Metadata-Version: 2.1
2
2
  Name: oasis-data-manager
3
- Version: 0.1.3
3
+ Version: 0.1.5
4
4
  Summary: UNKNOWN
5
5
  Home-page: https://github.com/OasisLMF/OasisDataManager
6
6
  Author: Oasis LMF
7
7
  Author-email: support@oasislmf.org
8
- License: BSD 3-Clause
8
+ License: UNKNOWN
9
9
  Platform: UNKNOWN
10
10
  Classifier: Development Status :: 4 - Beta
11
- Classifier: License :: OSI Approved :: BSD License
12
11
  Classifier: Operating System :: OS Independent
13
12
  Classifier: Programming Language :: Python
14
13
  Classifier: Programming Language :: Python :: 3.6
15
14
  Requires-Python: >=3.6
16
15
  Description-Content-Type: text/markdown
17
16
  Provides-Extra: extra
18
- License-File: LICENSE
19
17
 
20
18
  UNKNOWN
21
19
 
@@ -0,0 +1 @@
1
+ __version__ = '0.1.5'
@@ -64,11 +64,22 @@ class OasisReader:
64
64
  def _read(self):
65
65
  if not self.has_read:
66
66
  if hasattr(self.filename_or_buffer, "name"):
67
- extension = pathlib.Path(self.filename_or_buffer.name).suffix
67
+ parts = pathlib.Path(self.filename_or_buffer.name).parts
68
68
  else:
69
- extension = pathlib.Path(self.filename_or_buffer).suffix
69
+ parts = pathlib.Path(self.filename_or_buffer).parts
70
+
71
+ for part in parts:
72
+ for extension in [".parquet", ".pq"]:
73
+ if part.endswith(extension):
74
+ is_parquet = True
75
+ break
76
+ else:
77
+ continue # if parquet extension is found the outer break will be called exiting the for with is_parquet = True
78
+ break
79
+ else:
80
+ is_parquet = False
70
81
 
71
- if extension in [".parquet", ".pq"]:
82
+ if is_parquet:
72
83
  self.has_read = True
73
84
  self.read_parquet(*self.reader_args, **self.reader_kwargs)
74
85
  else:
@@ -0,0 +1,73 @@
1
+ import pyarrow.dataset as ds
2
+ from functools import reduce
3
+
4
+ from .base import OasisReader
5
+
6
+
7
+ class OasisPyarrowReader(OasisReader):
8
+
9
+ def read_parquet(self, *args, **kwargs):
10
+ def pyarrow_dataset_filter(pandas_filter):
11
+ def pyarrow_dataset_single_filter(pandas_single_filter):
12
+ op_map = {
13
+ "==": lambda x, y: ds.field(x) == y,
14
+ "!=": lambda x, y: ds.field(x) != y,
15
+ ">=": lambda x, y: ds.field(x) >= y,
16
+ "<=": lambda x, y: ds.field(x) <= y,
17
+ ">": lambda x, y: ds.field(x) > y,
18
+ "<": lambda x, y: ds.field(x) < y,
19
+ "in": lambda x, y: ds.field(x).isin(y),
20
+ "not in": lambda x, y: ~ds.field(x).isin(y)
21
+ }
22
+
23
+ field_name, operator, value = pandas_single_filter
24
+ df_filter = op_map.get(operator, lambda x, y: None)(field_name, value)
25
+
26
+ return df_filter
27
+
28
+ def list_of_tuples(lot):
29
+ # Apply pyarrow_dataset_single_filter to each tuple in the list and combine using &
30
+ return reduce(lambda x, y: x & y, (pyarrow_dataset_single_filter(f) for f in lot))
31
+
32
+ def list_of_lists(lol):
33
+ # Convert each list in lol to a tuple of filters using list_of_tuples, then combine using |
34
+ return reduce(lambda x, y: x | y, (list_of_tuples(f) for f in lol))
35
+
36
+ # Check if all elements in pandas_filter are of the same type (either tuple or list)
37
+ element_types = {type(f) for f in pandas_filter}
38
+ if len(element_types) > 1:
39
+ raise ValueError("Mixing and matching tuples and lists")
40
+
41
+ # Process based on the single type found
42
+ if isinstance(pandas_filter[0], tuple):
43
+ new_filter = list_of_tuples(pandas_filter)
44
+ elif isinstance(pandas_filter[0], list):
45
+ new_filter = list_of_lists(pandas_filter)
46
+ else:
47
+ raise TypeError("Unsupported type in filter")
48
+
49
+ return new_filter
50
+
51
+ if 'filters' in kwargs:
52
+ ds_filter = pyarrow_dataset_filter(kwargs['filters'])
53
+ else:
54
+ ds_filter = None
55
+
56
+ if isinstance(self.filename_or_buffer, str):
57
+ if self.filename_or_buffer.startswith(
58
+ "http://"
59
+ ) or self.filename_or_buffer.startswith("https://"):
60
+ dataset = ds.dataset(self.filename_or_buffer, partitioning='hive')
61
+ self.df = dataset.to_table(filter=ds_filter).to_pandas()
62
+ else:
63
+ _, uri = self.storage.get_storage_url(
64
+ self.filename_or_buffer, encode_params=False
65
+ )
66
+
67
+ uri = uri.replace('file://', '')
68
+ dataset = ds.dataset(uri, partitioning='hive')
69
+ self.df = dataset.to_table(filter=ds_filter).to_pandas()
70
+
71
+ else:
72
+ dataset = ds.dataset(self.filename_or_buffer, partitioning='hive')
73
+ self.df = dataset.to_table(filter=ds_filter).to_pandas()
@@ -6,6 +6,7 @@ __all__ = [
6
6
  'OasisDaskReader',
7
7
  'OasisDaskReaderCSV',
8
8
  'OasisDaskReaderParquet',
9
+ 'OasisPyarrowReader',
9
10
  ]
10
11
 
11
12
  """
@@ -16,6 +17,11 @@ __all__ = [
16
17
  from .backends.base import OasisReader
17
18
  from .backends.pandas import OasisPandasReader, OasisPandasReaderCSV, OasisPandasReaderParquet
18
19
 
20
+ try:
21
+ from .backends.pyarrow import OasisPyarrowReader
22
+ except ModuleNotFoundError as e:
23
+ pass
24
+
19
25
  try:
20
26
  from .backends.dask import OasisDaskReader, OasisDaskReaderCSV, OasisDaskReaderParquet
21
27
  except ModuleNotFoundError as e:
@@ -170,7 +170,6 @@ class AwsS3Storage(BaseStorage):
170
170
  s3_additional_kwargs["StorageClass"] = "REDUCED_REDUNDANCY"
171
171
 
172
172
  return {
173
- "anon": not self.access_key and not self.security_token,
174
173
  "key": self.access_key,
175
174
  "secret": self.secret_key,
176
175
  "token": self.security_token,
@@ -1,21 +1,19 @@
1
1
  Metadata-Version: 2.1
2
2
  Name: oasis-data-manager
3
- Version: 0.1.3
3
+ Version: 0.1.5
4
4
  Summary: UNKNOWN
5
5
  Home-page: https://github.com/OasisLMF/OasisDataManager
6
6
  Author: Oasis LMF
7
7
  Author-email: support@oasislmf.org
8
- License: BSD 3-Clause
8
+ License: UNKNOWN
9
9
  Platform: UNKNOWN
10
10
  Classifier: Development Status :: 4 - Beta
11
- Classifier: License :: OSI Approved :: BSD License
12
11
  Classifier: Operating System :: OS Independent
13
12
  Classifier: Programming Language :: Python
14
13
  Classifier: Programming Language :: Python :: 3.6
15
14
  Requires-Python: >=3.6
16
15
  Description-Content-Type: text/markdown
17
16
  Provides-Extra: extra
18
- License-File: LICENSE
19
17
 
20
18
  UNKNOWN
21
19
 
@@ -1,5 +1,4 @@
1
1
  CHANGELOG.rst
2
- LICENSE
3
2
  MANIFEST.in
4
3
  README.md
5
4
  requirements-package.in
@@ -23,6 +22,7 @@ oasis_data_manager/df_reader/backends/__init__.py
23
22
  oasis_data_manager/df_reader/backends/base.py
24
23
  oasis_data_manager/df_reader/backends/dask.py
25
24
  oasis_data_manager/df_reader/backends/pandas.py
25
+ oasis_data_manager/df_reader/backends/pyarrow.py
26
26
  oasis_data_manager/errors/__init__.py
27
27
  oasis_data_manager/filestore/__init__.py
28
28
  oasis_data_manager/filestore/config.py
@@ -6,10 +6,10 @@ typing_extensions
6
6
  [extra]
7
7
  adlfs
8
8
  boto3
9
- dask<=2024.1.1
9
+ dask>2024.1.1
10
10
  dask-geopandas
11
11
  dask-sql
12
12
  distributed
13
- geopandas
13
+ geopandas==0.14.4
14
14
  pyogrio
15
15
  s3fs>=2023.12.2
@@ -2,6 +2,8 @@
2
2
  universal = 1
3
3
 
4
4
  [metadata]
5
+ name = oasis-data-manager
6
+ version = attr: oasis_data_manager.__version__
5
7
  description-file = README.md
6
8
 
7
9
  [tool:pytest]
@@ -39,14 +39,11 @@ setup(
39
39
  packages=find_packages(exclude=("tests", "tests.*", "tests.*.*")),
40
40
  include_package_data=True,
41
41
  package_data={
42
- "": [
43
- "LICENSE",
44
- ],
42
+ "": [],
45
43
  },
46
44
  exclude_package_data={
47
45
  "": ["__pycache__", "*.py[co]"],
48
46
  },
49
- license="BSD 3-Clause",
50
47
  description="",
51
48
  long_description="",
52
49
  long_description_content_type="text/markdown",
@@ -59,7 +56,6 @@ setup(
59
56
  extras_require=get_optional_requirements(),
60
57
  classifiers=[
61
58
  "Development Status :: 4 - Beta",
62
- "License :: OSI Approved :: BSD License",
63
59
  "Operating System :: OS Independent",
64
60
  "Programming Language :: Python",
65
61
  "Programming Language :: Python :: 3.6",
@@ -1,28 +0,0 @@
1
- BSD 3-Clause License
2
-
3
- Copyright (c) 2023, Oasis Loss Modelling Framework
4
-
5
- Redistribution and use in source and binary forms, with or without
6
- modification, are permitted provided that the following conditions are met:
7
-
8
- 1. Redistributions of source code must retain the above copyright notice, this
9
- list of conditions and the following disclaimer.
10
-
11
- 2. Redistributions in binary form must reproduce the above copyright notice,
12
- this list of conditions and the following disclaimer in the documentation
13
- and/or other materials provided with the distribution.
14
-
15
- 3. Neither the name of the copyright holder nor the names of its
16
- contributors may be used to endorse or promote products derived from
17
- this software without specific prior written permission.
18
-
19
- THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
20
- AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
21
- IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
22
- DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
23
- FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
24
- DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
25
- SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
26
- CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
27
- OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
28
- OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
@@ -1 +0,0 @@
1
- __version__ = '0.1.3'