aboutsummaryrefslogtreecommitdiff
path: root/tests/incite/collections/test_df_collection_thl_web.py
diff options
context:
space:
mode:
authorstuppie2026-09-07 11:47:43 -0600
committerstuppie2026-09-07 11:47:43 -0600
commit092960233652cce1f4dc7841856034a6635e9cd9 (patch)
tree46e5fcd4d1e1b7ed0b987980c6c67ffa6e6b45c7 /tests/incite/collections/test_df_collection_thl_web.py
parent80fd8aab4c7271ddb619b0de18741d7ac77b490b (diff)
parent242579a44855873d5e054e375440e9d3492cd682 (diff)
downloadgeneralresearch-092960233652cce1f4dc7841856034a6635e9cd9.tar.gz
generalresearch-092960233652cce1f4dc7841856034a6635e9cd9.zip
Merge branch 'master' into dev-greg
Diffstat (limited to 'tests/incite/collections/test_df_collection_thl_web.py')
-rw-r--r--tests/incite/collections/test_df_collection_thl_web.py131
1 files changed, 93 insertions, 38 deletions
diff --git a/tests/incite/collections/test_df_collection_thl_web.py b/tests/incite/collections/test_df_collection_thl_web.py
index b09d44c..7253dd0 100644
--- a/tests/incite/collections/test_df_collection_thl_web.py
+++ b/tests/incite/collections/test_df_collection_thl_web.py
@@ -3,25 +3,20 @@ from __future__ import annotations
from collections.abc import Generator
from datetime import datetime
from itertools import product
-from typing import TYPE_CHECKING
import dask.dataframe as dd
import pandas as pd
import pytest
from pandera.pandas import DataFrameSchema
-from generalresearch.incite.collections import DFCollection, DFCollectionType
-
-if TYPE_CHECKING:
- from generalresearch.incite.base import GRLDatasets
- from generalresearch.incite.collections import (
- DFCollectionItem,
- DFCollectionType,
- )
+from generalresearch.incite.collections.base import (
+ DFCollection,
+ DFCollectionType,
+)
-def combo_object() -> Generator[tuple, None, None]:
- for x in product(
+def combo_object() -> Generator[tuple[DFCollectionType, str]]:
+ yield from product(
[
DFCollectionType.USER,
DFCollectionType.WALL,
@@ -30,9 +25,8 @@ def combo_object() -> Generator[tuple, None, None]:
DFCollectionType.AUDIT_LOG,
DFCollectionType.LEDGER,
],
- ["30min", "1H"],
- ):
- yield from x
+ ["30min", "1h"],
+ )
@pytest.mark.parametrize(
@@ -41,7 +35,10 @@ def combo_object() -> Generator[tuple, None, None]:
class TestDFCollection_thl_web:
def test_init(
- self, df_collection_data_type: DFCollectionType, offset: str, df_collection
+ self,
+ df_collection_data_type: DFCollectionType,
+ offset: str,
+ df_collection: DFCollection,
):
assert isinstance(df_collection_data_type, DFCollectionType)
assert isinstance(df_collection, DFCollection)
@@ -52,12 +49,12 @@ class TestDFCollection_thl_web:
)
class TestDFCollection_thl_web_Properties:
- def test_items(self, df_collection):
+ def test_items(self, df_collection: DFCollection):
assert isinstance(df_collection.items, list)
for i in df_collection.items:
assert i._collection == df_collection
- def test__schema(self, df_collection):
+ def test__schema(self, df_collection: DFCollection):
assert isinstance(df_collection._schema, DataFrameSchema)
@@ -67,16 +64,16 @@ class TestDFCollection_thl_web_Properties:
class TestDFCollection_thl_web_BaseProperties:
@pytest.mark.skip
- def test__interval_range(self, df_collection):
+ def test__interval_range(self, df_collection: DFCollection):
pass
- def test_interval_start(self, df_collection):
+ def test_interval_start(self, df_collection: DFCollection):
assert isinstance(df_collection.interval_start, datetime)
- def test_interval_range(self, df_collection):
+ def test_interval_range(self, df_collection: DFCollection):
assert isinstance(df_collection.interval_range, list)
- def test_progress(self, df_collection):
+ def test_progress(self, df_collection: DFCollection):
assert isinstance(df_collection.progress, pd.DataFrame)
@@ -86,17 +83,21 @@ class TestDFCollection_thl_web_BaseProperties:
class TestDFCollection_thl_web_Methods:
@pytest.mark.skip
- def test_initial_loads(self, df_collection_data_type, df_collection, offset):
+ def test_initial_loads(
+ self, df_collection_data_type, df_collection: DFCollection, offset: str
+ ):
pass
@pytest.mark.skip
def test_fetch_force_rr_latest(
- self, df_collection_data_type, df_collection, offset: str
+ self, df_collection_data_type, df_collection: DFCollection, offset: str
):
pass
@pytest.mark.skip
- def test_force_rr_latest(self, df_collection_data_type, df_collection, offset):
+ def test_force_rr_latest(
+ self, df_collection_data_type, df_collection: DFCollection, offset: str
+ ):
pass
@@ -105,63 +106,108 @@ class TestDFCollection_thl_web_Methods:
)
class TestDFCollection_thl_web_BaseMethods:
- def test_fetch_all_paths(self, df_collection_data_type, offset: str, df_collection):
+ def test_fetch_all_paths(
+ self,
+ df_collection_data_type: DFCollectionType,
+ offset: str,
+ df_collection: DFCollection,
+ ):
res = df_collection.fetch_all_paths(
items=None, force_rr_latest=False, include_partial=False
)
assert isinstance(res, list)
@pytest.mark.skip
- def test_ddf(self, df_collection_data_type, offset: str, df_collection):
+ def test_ddf(
+ self,
+ df_collection_data_type: DFCollectionType,
+ offset: str,
+ df_collection: DFCollection,
+ ):
res = df_collection.ddf()
assert isinstance(res, dd.DataFrame)
# -- cleanup --
@pytest.mark.skip
def test_schedule_cleanup(
- self, df_collection_data_type, offset: str, df_collection
+ self,
+ df_collection_data_type: DFCollectionType,
+ offset: str,
+ df_collection: DFCollection,
):
pass
@pytest.mark.skip
- def test_cleanup(self, df_collection_data_type, offset: str, df_collection):
+ def test_cleanup(
+ self,
+ df_collection_data_type: DFCollectionType,
+ offset: str,
+ df_collection: DFCollection,
+ ):
pass
@pytest.mark.skip
def test_cleanup_partials(
- self, df_collection_data_type, offset: str, df_collection
+ self,
+ df_collection_data_type: DFCollectionType,
+ offset: str,
+ df_collection: DFCollection,
):
pass
@pytest.mark.skip
def test_clear_tmp_archives(
- self, df_collection_data_type, offset: str, df_collection
+ self,
+ df_collection_data_type: DFCollectionType,
+ offset: str,
+ df_collection: DFCollection,
):
pass
@pytest.mark.skip
def test_clear_corrupt_archives(
- self, df_collection_data_type, offset: str, df_collection
+ self,
+ df_collection_data_type: DFCollectionType,
+ offset: str,
+ df_collection: DFCollection,
):
pass
@pytest.mark.skip
def test_rebuild_symlinks(
- self, df_collection_data_type, offset: str, df_collection
+ self,
+ df_collection_data_type: DFCollectionType,
+ offset: str,
+ df_collection: DFCollection,
):
pass
# -- Source timing --
@pytest.mark.skip
- def test_get_item(self, df_collection_data_type, offset: str, df_collection):
+ def test_get_item(
+ self,
+ df_collection_data_type: DFCollectionType,
+ offset: str,
+ df_collection: DFCollection,
+ ):
pass
@pytest.mark.skip
- def test_get_item_start(self, df_collection_data_type, offset: str, df_collection):
+ def test_get_item_start(
+ self,
+ df_collection_data_type: DFCollectionType,
+ offset: str,
+ df_collection: DFCollection,
+ ):
pass
@pytest.mark.skip
- def test_get_items(self, df_collection_data_type, offset: str, df_collection):
+ def test_get_items(
+ self,
+ df_collection_data_type: DFCollectionType,
+ offset: str,
+ df_collection: DFCollection,
+ ):
# If we get all the items from the start of the collection, it
# should include all the items!
res1 = df_collection.items
@@ -170,18 +216,27 @@ class TestDFCollection_thl_web_BaseMethods:
@pytest.mark.skip
def test_get_items_from_year(
- self, df_collection_data_type, offset: str, df_collection
+ self,
+ df_collection_data_type: DFCollectionType,
+ offset: str,
+ df_collection: DFCollection,
):
pass
@pytest.mark.skip
def test_get_items_last90(
- self, df_collection_data_type, offset: str, df_collection
+ self,
+ df_collection_data_type: DFCollectionType,
+ offset: str,
+ df_collection: DFCollection,
):
pass
@pytest.mark.skip
def test_get_items_last365(
- self, df_collection_data_type, offset: str, df_collection
+ self,
+ df_collection_data_type: DFCollectionType,
+ offset: str,
+ df_collection: DFCollection,
):
pass