All test cases running again fine!

oups/aggstream/aggstream.py

@@ -1326,19 +1326,19 @@ def agg(
 # works same as for bin: that empty snapshots are generated between 2 row groups.
 # - test case, test parameter value not in 'streamagg' nor in 'write' signature.
 # - un test snap + bin avec deux ou trois chunks dans lesquels il n'y a ni snap, ni bin qui se finissent.
-# - bien faire un cas test snapshot ou le 2nd seed chunk démarre sur une nouvelle bin,
-# straight away / cal pose des problèmes quand c'est simplement bin,
-# alors bin+snapshot, il y a des chances que ça ne marche pas non plus.
+# - bien faire un cas test snapshot ou le 2nd seed chunk démarre sur une nouvelle bin:
+#    straight away / ça pose des problèmes quand c'est simplement bin,
+#    alors bin+snapshot, il y a des chances que ça ne marche pas non plus.
 # - Do a test case to check that if in 'post' an exception is raised for instance,
-# then values in class are properly set: segagg_buffer, post_buffer, seed_index_restart,
-# and agg_res_buffer and bin_res_buffer are empty lists.
+#    then values in class are properly set: segagg_buffer, post_buffer, seed_index_restart,
+#    and agg_res_buffer and bin_res_buffer are empty lists.
 # - Test new parameters: 'final_write' and seed check exception
-#   for seed check exception, check the last '_last_seed_index' has been correctly recorded
-#   and aggregation results integrate results from last seed chunk.
+#    for seed check exception, check the last '_last_seed_index' has been correctly recorded
+#    and aggregation results integrate results from last seed chunk.
 # - A test that with a given filter release an empty dataframe.
 # - A test with a check function that raises an exception: check that last seed index has been
-#   that of the last iteration, prior to the exception.
+#    that of the last iteration, prior to the exception.
 # - A test for filter_dataframe
-#      check when some chunk are empty for a given filter id
+#    check when some chunk are empty for a given filter id
 # - A test with first seed generator like Parquet File, then 1 dataframe each time.
 # - Test with different 'ordered_on' values for a key vs seed

tests/test_aggstream/test_aggstream_advanced.py

@@ -219,7 +219,7 @@ def get_ref_results(seed_df):
         assert rec_res.equals(ref_df)
 
 
-def test_exception_different_indexes_at_restart(tmp_path):
+def test_exception_different_indexes_at_restart(store, seed_path):
     # Test exception at restart with 2 different 'seed_index_restart' for 2
     # different keys.
     # - key 1: time grouper '2T', agg 'first', and 'last',
@@ -242,7 +242,6 @@ def test_exception_different_indexes_at_restart(tmp_path):
     rand_ints.sort()
     ts = [start + Timedelta(f"{mn}T") for mn in rand_ints]
     seed_df = pDataFrame({ordered_on: ts, "val": rand_ints})
-    seed_path = os_path.join(tmp_path, "seed")
     fp_write(seed_path, seed_df, row_group_offsets=max_row_group_size, file_scheme="hive")
     seed = ParquetFile(seed_path).iter_row_groups()
     # Streamed aggregation for 'key1'.

tests/test_aggstream/test_jcumsegagg.py

@@ -13,12 +13,12 @@
 from numpy import ones
 from numpy import zeros
 
-from oups.streamagg.jcumsegagg import jcsagg
-from oups.streamagg.jcumsegagg import jfirst
-from oups.streamagg.jcumsegagg import jlast
-from oups.streamagg.jcumsegagg import jmax
-from oups.streamagg.jcumsegagg import jmin
-from oups.streamagg.jcumsegagg import jsum
+from oups.aggstream.jcumsegagg import jcsagg
+from oups.aggstream.jcumsegagg import jfirst
+from oups.aggstream.jcumsegagg import jlast
+from oups.aggstream.jcumsegagg import jmax
+from oups.aggstream.jcumsegagg import jmin
+from oups.aggstream.jcumsegagg import jsum
 
 
 INT64 = "int64"

tests/test_aggstream/test_segmentby.py

@@ -18,20 +18,20 @@
 from pandas import date_range
 from pandas.core.resample import TimeGrouper
 
-from oups.streamagg.segmentby import DTYPE_DATETIME64
-from oups.streamagg.segmentby import DTYPE_INT64
-from oups.streamagg.segmentby import KEY_BIN_BY
-from oups.streamagg.segmentby import KEY_ON_COLS
-from oups.streamagg.segmentby import KEY_ORDERED_ON
-from oups.streamagg.segmentby import LEFT
-from oups.streamagg.segmentby import NULL_INT64_1D_ARRAY
-from oups.streamagg.segmentby import RIGHT
-from oups.streamagg.segmentby import _next_chunk_starts
-from oups.streamagg.segmentby import by_scale
-from oups.streamagg.segmentby import by_x_rows
-from oups.streamagg.segmentby import mergesort
-from oups.streamagg.segmentby import segmentby
-from oups.streamagg.segmentby import setup_segmentby
+from oups.aggstream.segmentby import DTYPE_DATETIME64
+from oups.aggstream.segmentby import DTYPE_INT64
+from oups.aggstream.segmentby import KEY_BIN_BY
+from oups.aggstream.segmentby import KEY_ON_COLS
+from oups.aggstream.segmentby import KEY_ORDERED_ON
+from oups.aggstream.segmentby import LEFT
+from oups.aggstream.segmentby import NULL_INT64_1D_ARRAY
+from oups.aggstream.segmentby import RIGHT
+from oups.aggstream.segmentby import _next_chunk_starts
+from oups.aggstream.segmentby import by_scale
+from oups.aggstream.segmentby import by_x_rows
+from oups.aggstream.segmentby import mergesort
+from oups.aggstream.segmentby import segmentby
+from oups.aggstream.segmentby import setup_segmentby
 
 
 # from pandas.testing import assert_frame_equal

tests/test_aggstream/test_segmentby_restart.py

@@ -16,21 +16,21 @@
 from pandas import date_range
 from pandas.core.resample import TimeGrouper
 
-from oups.streamagg.segmentby import DTYPE_DATETIME64
-from oups.streamagg.segmentby import DTYPE_INT64
-from oups.streamagg.segmentby import KEY_BIN
-from oups.streamagg.segmentby import KEY_LAST_BIN_END
-from oups.streamagg.segmentby import KEY_LAST_BIN_LABEL
-from oups.streamagg.segmentby import KEY_LAST_ON_VALUE
-from oups.streamagg.segmentby import KEY_RESTART_KEY
-from oups.streamagg.segmentby import KEY_SNAP
-from oups.streamagg.segmentby import LEFT
-from oups.streamagg.segmentby import NULL_INT64_1D_ARRAY
-from oups.streamagg.segmentby import RIGHT
-from oups.streamagg.segmentby import by_scale
-from oups.streamagg.segmentby import by_x_rows
-from oups.streamagg.segmentby import segmentby
-from oups.streamagg.segmentby import setup_segmentby
+from oups.aggstream.segmentby import DTYPE_DATETIME64
+from oups.aggstream.segmentby import DTYPE_INT64
+from oups.aggstream.segmentby import KEY_BIN
+from oups.aggstream.segmentby import KEY_LAST_BIN_END
+from oups.aggstream.segmentby import KEY_LAST_BIN_LABEL
+from oups.aggstream.segmentby import KEY_LAST_ON_VALUE
+from oups.aggstream.segmentby import KEY_RESTART_KEY
+from oups.aggstream.segmentby import KEY_SNAP
+from oups.aggstream.segmentby import LEFT
+from oups.aggstream.segmentby import NULL_INT64_1D_ARRAY
+from oups.aggstream.segmentby import RIGHT
+from oups.aggstream.segmentby import by_scale
+from oups.aggstream.segmentby import by_x_rows
+from oups.aggstream.segmentby import segmentby
+from oups.aggstream.segmentby import setup_segmentby
 
 
 # from pandas.testing import assert_frame_equal
@@ -995,9 +995,9 @@ def by_not_restarting_with_same_bin_label(on, buffer=None):
             [None] * 3,
             [0] * 3,
             [
-                {KEY_RESTART_KEY: 3, KEY_LAST_BIN_LABEL: 0},
-                {KEY_RESTART_KEY: 2, KEY_LAST_BIN_LABEL: 4},
-                {KEY_RESTART_KEY: 1, KEY_LAST_BIN_LABEL: 8},
+                {KEY_BIN: {KEY_RESTART_KEY: 3, KEY_LAST_BIN_LABEL: 0}, KEY_LAST_BIN_LABEL: 0},
+                {KEY_BIN: {KEY_RESTART_KEY: 2, KEY_LAST_BIN_LABEL: 4}, KEY_LAST_BIN_LABEL: 4},
+                {KEY_BIN: {KEY_RESTART_KEY: 1, KEY_LAST_BIN_LABEL: 8}, KEY_LAST_BIN_LABEL: 8},
             ],
         ),
         (