Merge branch 'main' into mutation

Author: jsngn

ci/code_checks.sh

@@ -71,7 +71,6 @@ if [[ -z "$CHECK" || "$CHECK" == "docstrings" ]]; then
         -i ES01 `# For now it is ok if docstrings are missing the extended summary` \
         -i "pandas.Series.dt PR01" `# Accessors are implemented as classes, but we do not document the Parameters section` \
         -i "pandas.MultiIndex.reorder_levels RT03,SA01" \
-        -i "pandas.MultiIndex.to_frame RT03" \
         -i "pandas.NA SA01" \
         -i "pandas.NaT SA01" \
         -i "pandas.Period.freq GL08" \
@@ -81,27 +80,6 @@ if [[ -z "$CHECK" || "$CHECK" == "docstrings" ]]; then
         -i "pandas.Period.to_timestamp SA01" \
         -i "pandas.PeriodDtype SA01" \
         -i "pandas.PeriodDtype.freq SA01" \
-        -i "pandas.PeriodIndex.day SA01" \
-        -i "pandas.PeriodIndex.day_of_week SA01" \
-        -i "pandas.PeriodIndex.day_of_year SA01" \
-        -i "pandas.PeriodIndex.dayofweek SA01" \
-        -i "pandas.PeriodIndex.dayofyear SA01" \
-        -i "pandas.PeriodIndex.days_in_month SA01" \
-        -i "pandas.PeriodIndex.daysinmonth SA01" \
-        -i "pandas.PeriodIndex.from_fields PR07,SA01" \
-        -i "pandas.PeriodIndex.from_ordinals SA01" \
-        -i "pandas.PeriodIndex.hour SA01" \
-        -i "pandas.PeriodIndex.is_leap_year SA01" \
-        -i "pandas.PeriodIndex.minute SA01" \
-        -i "pandas.PeriodIndex.month SA01" \
-        -i "pandas.PeriodIndex.quarter SA01" \
-        -i "pandas.PeriodIndex.qyear GL08" \
-        -i "pandas.PeriodIndex.second SA01" \
-        -i "pandas.PeriodIndex.to_timestamp RT03,SA01" \
-        -i "pandas.PeriodIndex.week SA01" \
-        -i "pandas.PeriodIndex.weekday SA01" \
-        -i "pandas.PeriodIndex.weekofyear SA01" \
-        -i "pandas.PeriodIndex.year SA01" \
         -i "pandas.RangeIndex PR07" \
         -i "pandas.RangeIndex.from_range PR01,SA01" \
         -i "pandas.RangeIndex.start SA01" \
@@ -124,7 +102,6 @@ if [[ -z "$CHECK" || "$CHECK" == "docstrings" ]]; then
         -i "pandas.Series.dt.month_name PR01,PR02" \
         -i "pandas.Series.dt.nanoseconds SA01" \
         -i "pandas.Series.dt.normalize PR01" \
-        -i "pandas.Series.dt.qyear GL08" \
         -i "pandas.Series.dt.round PR01,PR02" \
         -i "pandas.Series.dt.seconds SA01" \
         -i "pandas.Series.dt.strftime PR01,PR02" \
@@ -134,12 +111,10 @@ if [[ -z "$CHECK" || "$CHECK" == "docstrings" ]]; then
         -i "pandas.Series.dt.tz_localize PR01,PR02" \
         -i "pandas.Series.dt.unit GL08" \
         -i "pandas.Series.pad PR01,SA01" \
-        -i "pandas.Series.sparse PR01,SA01" \
         -i "pandas.Series.sparse.fill_value SA01" \
         -i "pandas.Series.sparse.from_coo PR07,SA01" \
         -i "pandas.Series.sparse.npoints SA01" \
         -i "pandas.Series.sparse.sp_values SA01" \
-        -i "pandas.Series.sparse.to_coo PR07,RT03,SA01" \
         -i "pandas.Timedelta.asm8 SA01" \
         -i "pandas.Timedelta.ceil SA01" \
         -i "pandas.Timedelta.components SA01" \
@@ -152,7 +127,6 @@ if [[ -z "$CHECK" || "$CHECK" == "docstrings" ]]; then
         -i "pandas.Timedelta.to_timedelta64 SA01" \
         -i "pandas.Timedelta.total_seconds SA01" \
         -i "pandas.Timedelta.view SA01" \
-        -i "pandas.TimedeltaIndex.as_unit RT03,SA01" \
         -i "pandas.TimedeltaIndex.components SA01" \
         -i "pandas.TimedeltaIndex.microseconds SA01" \
         -i "pandas.TimedeltaIndex.nanoseconds SA01" \
@@ -253,11 +227,6 @@ if [[ -z "$CHECK" || "$CHECK" == "docstrings" ]]; then
         -i "pandas.core.resample.Resampler.sum SA01" \
         -i "pandas.core.resample.Resampler.transform PR01,RT03,SA01" \
         -i "pandas.core.resample.Resampler.var SA01" \
-        -i "pandas.core.window.expanding.Expanding.corr PR01" \
-        -i "pandas.core.window.expanding.Expanding.count PR01" \
-        -i "pandas.core.window.rolling.Rolling.max PR01" \
-        -i "pandas.core.window.rolling.Window.std PR01" \
-        -i "pandas.core.window.rolling.Window.var PR01" \
         -i "pandas.date_range RT03" \
         -i "pandas.errors.AbstractMethodError PR01,SA01" \
         -i "pandas.errors.AttributeConflictWarning SA01" \

doc/source/whatsnew/v2.3.0.rst

@@ -102,7 +102,7 @@ Conversion
 
 Strings
 ^^^^^^^
--
+- Bug in :meth:`Series.str.replace` when ``n < 0`` for :class:`StringDtype` with ``storage="pyarrow"`` (:issue:`59628`)
 -
 
 Interval

pandas/_libs/internals.pyx

@@ -1,9 +1,13 @@
 from collections import defaultdict
-import weakref
 
 cimport cython
+from cpython.object cimport PyObject
 from cpython.pyport cimport PY_SSIZE_T_MAX
 from cpython.slice cimport PySlice_GetIndicesEx
+from cpython.weakref cimport (
+    PyWeakref_GetObject,
+    PyWeakref_NewRef,
+)
 from cython cimport Py_ssize_t
 
 import numpy as np
@@ -26,6 +30,10 @@ from pandas._libs.util cimport (
 )
 
 
+cdef extern from "Python.h":
+    PyObject* Py_None
+
+
 @cython.final
 @cython.freelist(32)
 cdef class BlockPlacement:
@@ -746,7 +754,7 @@ cdef class BlockManager:
     # -------------------------------------------------------------------
     # Block Placement
 
-    def _rebuild_blknos_and_blklocs(self) -> None:
+    cpdef _rebuild_blknos_and_blklocs(self):
         """
         Update mgr._blknos / mgr._blklocs.
         """
@@ -890,27 +898,28 @@ cdef class BlockValuesRefs:
 
     def __cinit__(self, blk: Block | None = None) -> None:
         if blk is not None:
-            self.referenced_blocks = [weakref.ref(blk)]
+            self.referenced_blocks = [PyWeakref_NewRef(blk, None)]
         else:
             self.referenced_blocks = []
         self.clear_counter = 500  # set reasonably high
 
-    def _clear_dead_references(self, force=False) -> None:
+    cdef _clear_dead_references(self, bint force=False):
         # Use exponential backoff to decide when we want to clear references
         # if force=False. Clearing for every insertion causes slowdowns if
         # all these objects stay alive, e.g. df.items() for wide DataFrames
         # see GH#55245 and GH#55008
         if force or len(self.referenced_blocks) > self.clear_counter:
             self.referenced_blocks = [
-                ref for ref in self.referenced_blocks if ref() is not None
+                ref for ref in self.referenced_blocks
+                if PyWeakref_GetObject(ref) != Py_None
             ]
             nr_of_refs = len(self.referenced_blocks)
             if nr_of_refs < self.clear_counter // 2:
                 self.clear_counter = max(self.clear_counter // 2, 500)
             elif nr_of_refs > self.clear_counter:
                 self.clear_counter = max(self.clear_counter * 2, nr_of_refs)
 
-    def add_reference(self, blk: Block) -> None:
+    cpdef add_reference(self, Block blk):
         """Adds a new reference to our reference collection.
 
         Parameters
@@ -919,7 +928,7 @@ cdef class BlockValuesRefs:
             The block that the new references should point to.
         """
         self._clear_dead_references()
-        self.referenced_blocks.append(weakref.ref(blk))
+        self.referenced_blocks.append(PyWeakref_NewRef(blk, None))
 
     def add_index_reference(self, index: object) -> None:
         """Adds a new reference to our reference collection when creating an index.
@@ -930,7 +939,7 @@ cdef class BlockValuesRefs:
             The index that the new reference should point to.
         """
         self._clear_dead_references()
-        self.referenced_blocks.append(weakref.ref(index))
+        self.referenced_blocks.append(PyWeakref_NewRef(index, None))
 
     def has_reference(self) -> bool:
         """Checks if block has foreign references.

pandas/core/arrays/_arrow_string_mixins.py

@@ -23,6 +23,14 @@ class ArrowStringArrayMixin:
     def __init__(self, *args, **kwargs) -> None:
         raise NotImplementedError
 
+    def _convert_bool_result(self, result):
+        # Convert a bool-dtype result to the appropriate result type
+        raise NotImplementedError
+
+    def _convert_int_result(self, result):
+        # Convert an integer-dtype result to the appropriate result type
+        raise NotImplementedError
+
     def _str_pad(
         self,
         width: int,

pandas/core/arrays/arrow/array.py

@@ -2311,6 +2311,12 @@ def _apply_elementwise(self, func: Callable) -> list[list[Any]]:
             for chunk in self._pa_array.iterchunks()
         ]
 
+    def _convert_bool_result(self, result):
+        return type(self)(result)
+
+    def _convert_int_result(self, result):
+        return type(self)(result)
+
     def _str_count(self, pat: str, flags: int = 0) -> Self:
         if flags:
             raise NotImplementedError(f"count not implemented with {flags=}")