CI: unpin pyarrow, fix failing test

.github/actions/setup-conda/action.yml

@@ -18,7 +18,7 @@ runs:
     - name: Set Arrow version in ${{ inputs.environment-file }} to ${{ inputs.pyarrow-version }}
       run: |
         grep -q '  - pyarrow' ${{ inputs.environment-file }}
-        sed -i"" -e "s/  - pyarrow<11/  - pyarrow=${{ inputs.pyarrow-version }}/" ${{ inputs.environment-file }}
+        sed -i"" -e "s/  - pyarrow/  - pyarrow=${{ inputs.pyarrow-version }}/" ${{ inputs.environment-file }}
         cat ${{ inputs.environment-file }}
       shell: bash
       if: ${{ inputs.pyarrow-version }}

ci/deps/actions-310.yaml

@@ -42,7 +42,7 @@ dependencies:
   - psycopg2
   - pymysql
   - pytables
-  - pyarrow<11
+  - pyarrow
   - pyreadstat
   - python-snappy
   - pyxlsb

ci/deps/actions-311.yaml

@@ -42,7 +42,7 @@ dependencies:
   - psycopg2
   - pymysql
   # - pytables>=3.8.0  # first version that supports 3.11
-  - pyarrow<11
+  - pyarrow
   - pyreadstat
   - python-snappy
   - pyxlsb

ci/deps/actions-38-downstream_compat.yaml

@@ -40,7 +40,7 @@ dependencies:
   - openpyxl
   - odfpy
   - psycopg2
-  - pyarrow<11
+  - pyarrow
   - pymysql
   - pyreadstat
   - pytables

ci/deps/actions-38.yaml

@@ -40,7 +40,7 @@ dependencies:
   - odfpy
   - pandas-gbq
   - psycopg2
-  - pyarrow<11
+  - pyarrow
   - pymysql
   - pyreadstat
   - pytables

ci/deps/actions-39.yaml

@@ -41,7 +41,7 @@ dependencies:
   - pandas-gbq
   - psycopg2
   - pymysql
-  - pyarrow<11
+  - pyarrow
   - pyreadstat
   - pytables
   - python-snappy

ci/deps/circle-38-arm64.yaml

@@ -40,7 +40,7 @@ dependencies:
   - odfpy
   - pandas-gbq
   - psycopg2
-  - pyarrow<11
+  - pyarrow
   - pymysql
   # Not provided on ARM
   #- pyreadstat

environment.yml

@@ -43,7 +43,7 @@ dependencies:
   - odfpy
   - py
   - psycopg2
-  - pyarrow<11
+  - pyarrow
   - pymysql
   - pyreadstat
   - pytables

pandas/compat/__init__.py

@@ -31,6 +31,7 @@
     pa_version_under7p0,
     pa_version_under8p0,
     pa_version_under9p0,
+    pa_version_under11p0,
 )
 
 
@@ -161,6 +162,7 @@ def get_lzma_file() -> type[pandas.compat.compressors.LZMAFile]:
     "pa_version_under7p0",
     "pa_version_under8p0",
     "pa_version_under9p0",
+    "pa_version_under11p0",
     "IS64",
     "PY39",
     "PY310",

pandas/compat/pyarrow.py

@@ -14,9 +14,11 @@
     pa_version_under8p0 = _palv < Version("8.0.0")
     pa_version_under9p0 = _palv < Version("9.0.0")
     pa_version_under10p0 = _palv < Version("10.0.0")
+    pa_version_under11p0 = _palv < Version("11.0.0")
 except ImportError:
     pa_version_under6p0 = True
     pa_version_under7p0 = True
     pa_version_under8p0 = True
     pa_version_under9p0 = True
     pa_version_under10p0 = True
+    pa_version_under11p0 = True

pandas/core/arrays/arrow/array.py

@@ -258,9 +258,19 @@ def _from_sequence_of_strings(
 
             scalars = to_datetime(strings, errors="raise").date
         elif pa.types.is_duration(pa_type):
-            from pandas.core.tools.timedeltas import to_timedelta
+            try:
+                # GH51175: test_from_sequence_of_strings_pa_array
+                # attempt to parse as int64 reflecting pyarrow's
+                # duration to string casting behavior
+                if isinstance(strings, (pa.Array, pa.ChunkedArray)):
+                    scalars = strings
+                else:
+                    scalars = pa.array(strings, type=pa.string(), from_pandas=True)
+                scalars = scalars.cast(pa.int64())
+            except pa.ArrowInvalid:
+                from pandas.core.tools.timedeltas import to_timedelta
 
-            scalars = to_timedelta(strings, errors="raise")
+                scalars = to_timedelta(strings, errors="raise")
         elif pa.types.is_time(pa_type):
             from pandas.core.tools.times import to_time
 

pandas/core/tools/timedeltas.py

@@ -240,7 +240,9 @@ def _convert_listlike(
         #  returning arg (errors == "ignore"), and where the input is a
         #  generator, we return a useful list-like instead of a
         #  used-up generator
-        arg = np.array(list(arg), dtype=object)
+        if not hasattr(arg, "__array__"):
+            arg = list(arg)
+        arg = np.array(arg, dtype=object)
 
     try:
         td64arr = sequence_to_td64ns(arg, unit=unit, errors=errors, copy=False)[0]

pandas/tests/extension/test_arrow.py

@@ -33,6 +33,7 @@
     pa_version_under7p0,
     pa_version_under8p0,
     pa_version_under9p0,
+    pa_version_under11p0,
 )
 from pandas.errors import PerformanceWarning
 
@@ -295,7 +296,7 @@ def test_from_sequence_of_strings_pa_array(self, data, request):
                     reason="Nanosecond time parsing not supported.",
                 )
             )
-        elif pa.types.is_duration(pa_dtype):
+        elif pa_version_under11p0 and pa.types.is_duration(pa_dtype):
             request.node.add_marker(
                 pytest.mark.xfail(
                     raises=pa.ArrowNotImplementedError,

requirements-dev.txt

@@ -32,7 +32,7 @@ openpyxl
 odfpy
 py
 psycopg2-binary
-pyarrow<11
+pyarrow
 pymysql
 pyreadstat
 tables