…apache#48925)

### What changes are included in this PR?

Bug fixes and robustness improvements in the IPC file reader:
* Fix bug reading variadic buffers with pre-buffering enabled
* Fix bug reading dictionaries with pre-buffering enabled
* Validate IPC buffer offsets and lengths

Testing improvements:
* Exercise pre-buffering in IPC tests
* Actually exercise variadic buffers in IPC tests, by ensuring non-inline binary views are generated
* Run fuzz targets on golden IPC integration files in ASAN/UBSAN CI job
* Exercise pre-buffering in the IPC file fuzz target

Miscellaneous:
* Add convenience functions for integer overflow checking

### Are these changes tested?

Yes, by existing and improved tests.

### Are there any user-facing changes?

Bug fixes.

**This PR contains a "Critical Fix".** Fixes a potential crash reading variadic buffers with pre-buffering enabled.

* GitHub Issue: apache#48924

Authored-by: Antoine Pitrou <[email protected]>
Signed-off-by: Antoine Pitrou <[email protected]>

…river and the Flight Client (apache#48967)

### Rationale for this change

The bug breaks a Flight SQL server that refreshens the auth token when cookie authentication is enabled

### What changes are included in this PR?

1. In the ODBC layer, removed the code that adds a 2nd ClientCookieMiddlewareFactory in the client options (the 1st one is registered in `BuildFlightClientOptions`). This fixes the issue of the duplicate header cookie fields.
2. In the flight client layer, uses the case-insensitive equality comparator instead of the case-insensitive less-than comparator for the cookies cache which is an unordered map. This fixes the issue of duplicate cookie keys.

### Are these changes tested?
Manually on Windows, and CI

### Are there any user-facing changes?

No
* GitHub Issue: apache#48966

Authored-by: jianfengmao <[email protected]>
Signed-off-by: David Li <[email protected]>

…e buffer is empty (apache#48692)

### Rationale for this change
WriteArrowSerialize could unconditionally read values from the Arrow array even for null rows. Since it's possible the caller could provided a zero-sized dummy buffer for all-null arrays, this caused an ASAN heap-buffer-overflow.

### What changes are included in this PR?
Early check the array is not all null values before serialize it

### Are these changes tested?

Added tests.
### Are there any user-facing changes?

No

* GitHub Issue: apache#48691

Authored-by: rexan <[email protected]>
Signed-off-by: Gang Wu <[email protected]>

…r.msix to fix docker rebuild on Windows wheels (apache#48948)

### Rationale for this change

As soon as we have to rebuild our Windows docker images they will fail installing python-manager-25.0.msix

### What changes are included in this PR?

- Use `pymanager.msi` to install python version instead of `pymanager.msix` which has problems on Docker.
- Update `pymanager install` command to use newer API (old command fails with missing flags)
- Update default python command to use the free-threaded required suffix if free-threaded wheels

### Are these changes tested?

Yes via archery

### Are there any user-facing changes?

No
* GitHub Issue: apache#48947

Authored-by: Raúl Cumplido <[email protected]>
Signed-off-by: Raúl Cumplido <[email protected]>

)

### Rationale for this change

There are date32 and date64 variants for date arrays.

### What changes are included in this PR?

* Add `ArrowFormat::DateType#to_flatbuffers`

### Are these changes tested?

Yes.

### Are there any user-facing changes?

Yes.
* GitHub Issue: apache#48990

Authored-by: Sutou Kouhei <[email protected]>
Signed-off-by: Sutou Kouhei <[email protected]>

…che#48993)

### Rationale for this change

It's a large variant of UTF-8 array.

### What changes are included in this PR?

* Add `ArrowFormat::LargeUTF8Type#to_flatbuffers`
* Add support for large UTF-8 array of `#values` and `#raw_records`

### Are these changes tested?

Yes.

### Are there any user-facing changes?

Yes.
* GitHub Issue: apache#48992

Authored-by: Sutou Kouhei <[email protected]>
Signed-off-by: Sutou Kouhei <[email protected]>

…::FileReader::ReadRowGroup(s) (apache#48982)

### Rationale for this change
`FileReader::ReadRowGroup(s)` previously returned `Status` and required callers to pass an `out` parameter.
### What changes are included in this PR?
Introduce `Result<std::shared_ptr<Table>>` returning APIs to allow clearer error propagation:
  - Add new Result-returning `ReadRowGroup()` / `ReadRowGroups()` methods
  - Deprecate the old Status/out-parameter overloads
  - Update C++ callers and R/Python/GLib bindings to use the new API
### Are these changes tested?
Yes.
### Are there any user-facing changes?
Yes.
Status versions of FileReader::ReadRowGroup(s) have been deprecated.
```cpp
virtual ::arrow::Status ReadRowGroup(int i, const std::vector<int>& column_indices,
                                     std::shared_ptr<::arrow::Table>* out);
virtual ::arrow::Status ReadRowGroup(int i, std::shared_ptr<::arrow::Table>* out);

virtual ::arrow::Status ReadRowGroups(const std::vector<int>& row_groups,
                                      const std::vector<int>& column_indices,
                                      std::shared_ptr<::arrow::Table>* out);
virtual ::arrow::Status ReadRowGroups(const std::vector<int>& row_groups,
                                      std::shared_ptr<::arrow::Table>* out);
```
* GitHub Issue: apache#48949

Lead-authored-by: fenfeng9 <[email protected]>
Co-authored-by: fenfeng9 <[email protected]>
Co-authored-by: Sutou Kouhei <[email protected]>
Co-authored-by: Gang Wu <[email protected]>
Signed-off-by: Sutou Kouhei <[email protected]>

…essions (apache#48989)

### Rationale for this change

Some node options and expressions miss arguments reference. If they miss, arguments may be freed by GC.

### What changes are included in this PR?

* Refer arguments of `garrow_filter_node_options_new()`
* Refer arguments of `garrow_project_node_options_new()`
* Refer arguments of `garrow_aggregate_node_options_new()`
* Refer arguments of `garrow_literal_expression_new()`
* Refer arguments of `garrow_call_expression_new()`
 
### Are these changes tested?

Yes.

### Are there any user-facing changes?

Yes.
* GitHub Issue: apache#48985

Authored-by: Sutou Kouhei <[email protected]>
Signed-off-by: Sutou Kouhei <[email protected]>

…s found on emscripten jobs (apache#49007)

### Rationale for this change

When looking for the wheel the script was falling back to returning a 404 even when the wheel was found:
```
 + python scripts/run_emscripten_tests.py dist/pyarrow-24.0.0.dev31-cp312-cp312-pyodide_2024_0_wasm32.whl --dist-dir=/pyodide --runtime=chrome
127.0.0.1 - - [27/Jan/2026 01:14:50] code 404, message File not found
```
Timing out the job and failing.

### What changes are included in this PR?

Correct logic and only return 404 if the file requested wasn't found.

### Are these changes tested?

Yes via archery

### Are there any user-facing changes?

No
* GitHub Issue: apache#47692

Authored-by: Raúl Cumplido <[email protected]>
Signed-off-by: Raúl Cumplido <[email protected]>

…king (apache#48974)

### Rationale for this change

Benchmark failing since C++20 upgrade due to lack of C++20 configuration

### What changes are included in this PR?

Changes entirely from 🤖 (Claude) with discussion from me regarding optimal approach.  

Description as follows:

> conda-forge's R package doesn't have CXX20 configured in Makeconf, even though the compiler (gcc 14.3.0) supports C++20. This causes Arrow R package installation to fail with "a C++20 compiler is required" because `R CMD config CXX20` returns empty. 
>
> This PR adds CXX20 configuration to R's Makeconf before building the Arrow R package in the benchmark hooks, if not already present.                                                               

### Are these changes tested?

I got 🤖  to try it locally in a container but I'm not convinced we'll know for sure til we try it out properly.

>  Tested in Docker container with Amazon Linux 2023 + conda-forge R - confirmed `R CMD config CXX20` returns empty before patch and `g++` after patch.
>
> The only thing we didn't test end-to-end was actually building Arrow R, but that would have taken much longer and the configure check (R CMD config CXX20 returning non-empty) is exactly what Arrow's configure script tests before proceeding.                                       

### Are there any user-facing changes?

Nope
* GitHub Issue: apache#48912

Authored-by: Nic Crane <[email protected]>
Signed-off-by: Nic Crane <[email protected]>

…ch is empty (apache#48718)

### Rationale for this change

Fixes apache#36889

When writing CSV from a table where the first batch is empty, the header gets written twice:

```python
table = pa.table({"col1": ["a", "b", "c"]})
combined = pa.concat_tables([table.schema.empty_table(), table])
write_csv(combined, buf)
# Result: "col1"\n"col1"\n"a"\n"b"\n"c"\n  <-- header appears twice
```

### What changes are included in this PR?

The bug happens because:
1. Header is written to `data_buffer_` and flushed during `CSVWriterImpl` initialization
2. The buffer is not cleared after flush
3. When the next batch is empty, `TranslateMinimalBatch` returns early without modifying `data_buffer_`
4. The write loop then writes `data_buffer_` which still contains stale content

The fix introduces a `WriteAndClearBuffer()` helper that writes the buffer to sink and clears it. This helper is used in all write paths:
- `WriteHeader()`
- `WriteRecordBatch()`
- `WriteTable()`

This ensures the buffer is always clean after any flush, making it impossible for stale content to be written again.

### Are these changes tested?

Yes. Added C++ tests in `writer_test.cc` and Python tests in `test_csv.py`:
- Empty batch at start of table
- Empty batch in middle of table

### Are there any user-facing changes?

No API changes. This is a bug fix that prevents duplicate headers when writing CSV from tables with empty batches.

* GitHub Issue: apache#36889

Lead-authored-by: Ruiyang Wang <[email protected]>
Co-authored-by: Ruiyang Wang <[email protected]>
Co-authored-by: Gang Wu <[email protected]>
Signed-off-by: Gang Wu <[email protected]>

…DBC Nightly Package (apache#48933)

### Rationale for this change
apache#48932
### What changes are included in this PR?
- Fix `rsync` build error ODBC Nightly Package 
### Are these changes tested?
- tested in CI
### Are there any user-facing changes?
- After fix, users should be able to get Nightly ODBC package release

* GitHub Issue: apache#48932

Authored-by: Alina (Xi) Li <[email protected]>
Signed-off-by: Sutou Kouhei <[email protected]>

…he#48952)

### Rationale for this change

Add guidance re AI tooling

### What changes are included in this PR?

Updates to main docs and links to it from new contributor's guide

### Are these changes tested?

No but I'll built the docs

### Are there any user-facing changes?

Just docs

:robot: Changes generated using Claude Code - I took the discussion from the mailing list, asked it to add the original text and then apply suggested changes one at a time, made a few of my own tweaks, and then instructed it to edit things down a bit for clarity and conciseness.
* GitHub Issue: apache#48951

Lead-authored-by: Nic Crane <[email protected]>
Co-authored-by: Rok Mihevc <[email protected]>
Co-authored-by: Andrew Lamb <[email protected]>
Signed-off-by: Nic Crane <[email protected]>

### Rationale for this change

`sphinx-build` allows for parallel operation, but it builds serially by default and that can be very slow on our docs given the amount of documents (many of them auto-generated from API docs).

### Are these changes tested?

By existing CI jobs.

### Are there any user-facing changes?

No.
* GitHub Issue: apache#49029

Authored-by: Antoine Pitrou <[email protected]>
Signed-off-by: Raúl Cumplido <[email protected]>

### Rationale for this change

This functionality is unused now that we have a proper atfork facility.

### Are these changes tested?

By existing CI tests.

### Are there any user-facing changes?

Removing an API that was always meant for internal use (though we didn't flag it explicitly as internal).

* GitHub Issue: apache#33450

Authored-by: Antoine Pitrou <[email protected]>
Signed-off-by: Antoine Pitrou <[email protected]>

…t& return types (apache#48956)

### Rationale for this change

The TODO comment in `vector_array_sort.cc` asking whether `DictionaryArray::dictionary()` and `DictionaryArray::indices()` should return `const&` has been obsolete.

It was added in commit 6ceb12f when dictionary array sorting was implemented. At that time, these methods returned `std::shared_ptr<Array>` by value, causing unnecessary copies.

The issue was fixed in commit 95a8bfb which changed both methods to return `const std::shared_ptr<Array>&`, removing the copies. However, the TODO comment was left unremoved.

### What changes are included in this PR?

Removed the outdated TODO comment that referenced apacheGH-35437.

### Are these changes tested?

I did not test.

### Are there any user-facing changes?

No.
* GitHub Issue: apache#35437

Authored-by: Hyukjin Kwon <[email protected]>
Signed-off-by: Antoine Pitrou <[email protected]>

…che#49008)

### Rationale for this change

When running the python-sdist job we are currently not uploading the build artifact to the job.

### What changes are included in this PR?

Upload artifact as part of building the job so it's easier to test and validate contents if necessary.

### Are these changes tested?

Yes via archery.

### Are there any user-facing changes?

No

* GitHub Issue: apache#48586

Authored-by: Raúl Cumplido <[email protected]>
Signed-off-by: Raúl Cumplido <[email protected]>

### Rationale for this change

CI needs updating to test old R package versions

### What changes are included in this PR?

Add 22.0.0.1

### Are these changes tested?

Nah, it's CI stuff

### Are there any user-facing changes?

No

Authored-by: Nic Crane <[email protected]>
Signed-off-by: Raúl Cumplido <[email protected]>

)

### Rationale for this change
See issue apache#48961
Pandas 3.0.0 string storage type changes https://github.com/pandas-dev/pandas/pull/62118/changes
and https://pandas.pydata.org/docs/whatsnew/v3.0.0.html#dedicated-string-data-type-by-default

### What changes are included in this PR?
Updating several doctest examples from `string` to `large_string`.

### Are these changes tested?
Yes, locally.

### Are there any user-facing changes?
No.

Closes apache#48961 
* GitHub Issue: apache#48961

Authored-by: Tadeja Kadunc <[email protected]>
Signed-off-by: AlenkaF <[email protected]>

…nchmarking (apache#49038)

### Rationale for this change

Slow benchmarks due to conda duckdb building from source

### What changes are included in this PR?

Try ditching conda and installing R via rig and using PPM binaries

### Are these changes tested?

I'll try running

### Are there any user-facing changes?
 
Nope
* GitHub Issue: apache#49037

Authored-by: Nic Crane <[email protected]>
Signed-off-by: Nic Crane <[email protected]>

### Rationale for this change

This patch was integrated upstream in microsoft/mimalloc#1139

### Are these changes tested?

By existing CI.

### Are there any user-facing changes?

No.
* GitHub Issue: apache#49042

Authored-by: Antoine Pitrou <[email protected]>
Signed-off-by: Sutou Kouhei <[email protected]>

### Rationale for this change

Default Debian version in `.env` now maps to oldstable, we should use stable instead.
Also prune entries that are not used anymore.

### Are these changes tested?

By existing CI jobs.

### Are there any user-facing changes?

No.
* GitHub Issue: apache#49024

Authored-by: Antoine Pitrou <[email protected]>
Signed-off-by: Sutou Kouhei <[email protected]>

)

### Rationale for this change

There are 32/64 bit and second/millisecond/microsecond/nanosecond variants for time arrays.

### What changes are included in this PR?

* Add `ArrowFormat::TimeType#to_flatbuffers`
* Add bit width information to `ArrowFormat::TimeType`

### Are these changes tested?

Yes.

### Are there any user-facing changes?

Yes.

* GitHub Issue: apache#49027

Authored-by: Sutou Kouhei <[email protected]>
Signed-off-by: Sutou Kouhei <[email protected]>

apache#49031)

### Rationale for this change

It's a fixed size variant of binary array.

### What changes are included in this PR?

* Add `ArrowFormat::FixedSizeBinaryType#to_flatbuffers`
* Add `ArrowFormat::FixedSizeBinaryArray#each_buffer`

### Are these changes tested?

Yes.

### Are there any user-facing changes?

Yes.
* GitHub Issue: apache#49030

Authored-by: Sutou Kouhei <[email protected]>
Signed-off-by: Sutou Kouhei <[email protected]>

…s in `castTIMESTAMP_utf8` and `castTIME_utf8` (apache#48867)

### Rationale for this change

Fixes apache#48866. The Gandiva precompiled time functions `castTIMESTAMP_utf8` and `castTIME_utf8` currently reject timestamp and time string literals with more than 3 subsecond digits (beyond millisecond precision), throwing an "Invalid millis" error. This behavior is inconsistent with other implementations.

### What changes are included in this PR?

- Fixed `castTIMESTAMP_utf8` and `castTIME_utf8` functions to truncate subseconds beyond 3 digits instead of throwing an error
- Updated tests. Replaced error-expecting tests with truncation verification tests and added edge cases

### Are these changes tested?

Yes

### Are there any user-facing changes?

No
* GitHub Issue: apache#48866

Authored-by: Arkadii Kravchuk <[email protected]>
Signed-off-by: Sutou Kouhei <[email protected]>

…d+ pattern before removing lines (apache#48674)

### Rationale for this change

This PR proposes to fix the todo https://github.com/apache/arrow/blob/7ebc88c8fae62ed97bc30865c845c8061132af7e/cpp/src/arrow/status.cc#L131-L134 which would allows a better parsing for line numbers.

I could not find the relevant example to demonstrate within this project but assume that we have a test such as:

(Generated by ChatGPT)

```cpp
TEST(BlockParser, ErrorMessageWithColonsPreserved) {
  Status st(StatusCode::Invalid,
            "CSV parse error: Row #2: Expected 2 columns, got 3: 12:34:56,key:value,data\n"
            "Error details: Time format: 12:34:56, Key: value\n"
            "parser_test.cc:940  Parse(parser, csv, &out_size)");

  std::string expected_msg =
      "Invalid: CSV parse error: Row #2: Expected 2 columns, got 3: 12:34:56,key:value,data\n"
      "Error details: Time format: 12:34:56, Key: value";

  ASSERT_RAISES_WITH_MESSAGE(Invalid, expected_msg, st);
}

// Test with URL-like data (another common case with colons)
TEST(BlockParser, ErrorMessageWithURLPreserved) {
  Status st(StatusCode::Invalid,
            "CSV parse error: Row #2: Expected 1 columns, got 2: http://arrow.apache.org:8080/api,data\n"
            "URL: http://arrow.apache.org:8080/api\n"
            "parser_test.cc:974  Parse(parser, csv, &out_size)");

  std::string expected_msg =
      "Invalid: CSV parse error: Row #2: Expected 1 columns, got 2: http://arrow.apache.org:8080/api,data\n"
      "URL: http://arrow.apache.org:8080/api";

  ASSERT_RAISES_WITH_MESSAGE(Invalid, expected_msg, st);
}
```

then it fails.

### What changes are included in this PR?

Fixed `Status::ToStringWithoutContextLines()` to only remove context lines matching the `filename:line` pattern (`:\d+`), preventing legitimate error messages containing colons from being incorrectly stripped.

### Are these changes tested?

Manually tested, and unittests were added, with `cmake .. --preset ninja-debug -DARROW_EXTRA_ERROR_CONTEXT=ON`.

### Are there any user-facing changes?

No, test-only.

* GitHub Issue: apache#48673

Authored-by: Hyukjin Kwon <[email protected]>
Signed-off-by: Sutou Kouhei <[email protected]>

…dding required user-agent on urllib request (apache#49052)

### Rationale for this change

See: apache#49044

### What changes are included in this PR?

Urllib now request with `"user-agent": "pyarrow"`

### Are these changes tested?

It's a CI fix.

### Are there any user-facing changes?

No, just a CI test fix.
* GitHub Issue: apache#49044

Authored-by: Rok Mihevc <[email protected]>
Signed-off-by: Raúl Cumplido <[email protected]>

…d and add check to validate LICENSE.txt and NOTICE.txt are part of the wheel contents (apache#48988)

### Rationale for this change

Currently the files are missing from the published wheels.

### What changes are included in this PR?

- Ensure the license and notice files are part of the wheels
- Use build frontend to build wheels
- Build wheel from sdist

### Are these changes tested?

Yes, via archery.
I've validated all wheels will fail with the new check if LICENSE.txt or NOTICE.txt are missing:
```
 AssertionError: LICENSE.txt is missing from the wheel.
```

### Are there any user-facing changes?

No

* GitHub Issue: apache#48983

Lead-authored-by: Raúl Cumplido <[email protected]>
Co-authored-by: Antoine Pitrou <[email protected]>
Co-authored-by: Rok Mihevc <[email protected]>
Signed-off-by: Raúl Cumplido <[email protected]>

…che#49060)

### Rationale for this change

Fix two issues found by OSS-Fuzz in the IPC reader:

* a controlled abort on invalid IPC metadata: https://oss-fuzz.com/testcase-detail/5301064831401984
* a nullptr dereference on invalid IPC metadata: https://oss-fuzz.com/testcase-detail/5091511766417408

None of these two issues is a security issue.

### Are these changes tested?

Yes, by new unit tests and new fuzz regression files.

### Are there any user-facing changes?

No.

**This PR contains a "Critical Fix".** (If the changes fix either (a) a security vulnerability, (b) a bug that caused incorrect or invalid data to be produced, or (c) a bug that causes a crash (even when the API contract is upheld), please provide explanation. If not, you can remove this.)

* GitHub Issue: apache#49059

Authored-by: Antoine Pitrou <[email protected]>
Signed-off-by: Antoine Pitrou <[email protected]>

- Implemented BuildSchemaFieldRecursive helper for depth-first traversal
- Walks Arrow schema and ORC type tree in parallel
- Assigns column indices using ORC depth-first pre-order (col 0 = root struct)
- Marks leaf nodes (primitives) with column_index for statistics
- Marks container nodes (struct/list/map) with column_index = -1
- Builds column_index_to_field and child_to_parent lookup maps
- Handles struct, list, and map types with proper child matching
- Added orc/Type.hh include for ORC type information

Implementation details:
- Column indexing starts at 1 (column 0 is root struct)
- Leaf nodes are primitives that have statistics
- Container types recursively process children
- Validates ORC root type is STRUCT

Verified: Manual code review - follows ORC depth-first pre-order pattern

Co-authored-by: Claude Sonnet 4.5 <[email protected]>

- Implemented GetOrcColumnIndex helper function that:
  - Resolves FieldRef to ORC column index using manifest
  - Uses FieldRef.FindOne() to locate field in schema
  - Traverses manifest tree following field path indices
  - Handles both top-level and nested fields
  - Returns column_index for leaf nodes (primitives with statistics)
  - Returns std::nullopt for containers or not found

- Added necessary includes:
  - <optional> for std::optional return type
  - arrow/compute/api_scalar.h for FieldRef and FieldPath

Implementation details:
- Top-level fields accessed via manifest.schema_fields[index]
- Nested fields traversed via current_field->children[index]
- Validates indices at each level to prevent out-of-bounds
- Only returns column_index if field is leaf (has statistics)
- Containers (struct/list/map) return nullopt

Verified: Manual code review - follows FieldRef resolution pattern

Co-authored-by: Claude Sonnet 4.5 <[email protected]>

- Add OrcFileFragment class extending FileFragment with ORC-specific predicate pushdown
- Add StripeStatisticsCache structure for caching stripe-level statistics expressions
- Add OrcCacheStatus enum for tracking metadata loading state
- Implement public API: stripes(), metadata(), EnsureCompleteMetadata(), Subset(), SplitByStripe()
- Add private methods: FilterStripes(), TestStripes(), TryCountRows() (stubs for later tasks)
- Mirror ParquetFileFragment structure adapted for ORC stripes
- Use void* for orc::Reader to avoid exposing ORC headers in public API
- Thread-safe with mutex locking following Fragment base class pattern

Implementation notes:
- stripes_ replaces Parquet's row_groups_
- statistics_cache_ stores per-stripe guarantee expressions
- Forward declarations for ORC types to minimize header dependencies
- Most methods are stubs marked with TODO for subsequent tasks

Verified: Manual code review following Parquet reference patterns

Co-Authored-By: Claude Sonnet 4.5 <[email protected]>