CHANGELOG.md

Apache Arrow 0.9.0 (16 March 2018)

Bug

• ARROW-1345 - [Python] Conversion from nested NumPy arrays fails on integers other than int64, float32
• ARROW-1646 - [Python] pyarrow.array cannot handle NumPy scalar types
• ARROW-1856 - [Python] Auto-detect Parquet ABI version when using PARQUET_HOME
• ARROW-1909 - [C++] Bug: Build fails on windows with "-DARROW_BUILD_BENCHMARKS=ON"
• ARROW-1912 - [Website] Add org affiliations to committers.html
• ARROW-1919 - Plasma hanging if object id is not 20 bytes
• ARROW-1924 - [Python] Bring back pickle=True option for serialization
• ARROW-1933 - [GLib] Build failure with --with-arrow-cpp-build-dir and GPU enabled Arrow C++
• ARROW-1940 - [Python] Extra metadata gets added after multiple conversions between pd.DataFrame and pa.Table
• ARROW-1941 - Table <–> DataFrame roundtrip failing
• ARROW-1943 - Handle setInitialCapacity() for deeply nested lists of lists
• ARROW-1944 - FindArrow has wrong ARROW_STATIC_LIB
• ARROW-1945 - [C++] Fix doxygen documentation of array.h
• ARROW-1946 - Add APIs to decimal vector for writing big endian data
• ARROW-1948 - [Java] ListVector does not handle ipc with all non-null values with none set
• ARROW-1950 - [Python] pandas_type in pandas metadata incorrect for List types
• ARROW-1953 - [JS] JavaScript builds broken on master
• ARROW-1958 - [Python] Error in pandas conversion for datetimetz row index
• ARROW-1961 - [Python] Writing Parquet file with flavor='spark' loses pandas schema metadata
• ARROW-1966 - [C++] Support JAVA_HOME paths in HDFS libjvm loading that include the jre directory
• ARROW-1971 - [Python] Add pandas serialization to the default
• ARROW-1972 - Deserialization of buffer objects (and pandas dataframes) segfaults on different processes.
• ARROW-1973 - [Python] Memory leak when converting Arrow tables with array columns to Pandas dataframes.
• ARROW-1976 - [Python] Handling unicode pandas columns on parquet.read_table
• ARROW-1979 - [JS] JS builds handing in es2015:umd tests
• ARROW-1980 - [Python] Race condition in write\_to\_dataset
• ARROW-1982 - [Python] Return parquet statistics min/max as values instead of strings
• ARROW-1991 - [GLib] Docker-based documentation build is broken
• ARROW-1992 - [Python] to_pandas crashes when using strings_to_categoricals on empty string cols on 0.8.0
• ARROW-1997 - [Python] to_pandas with strings_to_categorical fails
• ARROW-1998 - [Python] Table.from_pandas crashes when data frame is empty
• ARROW-1999 - [Python] from_numpy_dtype returns wrong types
• ARROW-2000 - Deduplicate file descriptors when plasma store replies to get request.
• ARROW-2002 - use pyarrow download file will raise queue.Full exceptions sometimes
• ARROW-2003 - [Python] Do not use deprecated kwarg in pandas.core.internals.make_block
• ARROW-2005 - [Python] pyflakes warnings on Cython files not failing build
• ARROW-2008 - [Python] Type inference for int32 NumPy arrays (expecting list) returns int64 and then conversion fails
• ARROW-2010 - [C++] Compiler warnings with CHECKIN warning level in ORC adapter
• ARROW-2017 - Array initialization with large (>2**31-1) uint64 values fails
• ARROW-2023 - [C++] Test opening IPC stream reader or file reader on an empty InputStream
• ARROW-2025 - [Python/C++] HDFS Client disconnect closes all open clients
• ARROW-2029 - [Python] Program crash on HdfsFile.tell if file is closed
• ARROW-2032 - [C++] ORC ep installs on each call to ninja build (even if no work to do)
• ARROW-2033 - pa.array() doesn't work with iterators
• ARROW-2039 - [Python] pyarrow.Buffer().to_pybytes() segfaults
• ARROW-2040 - [Python] Deserialized Numpy array must keep ref to underlying tensor
• ARROW-2047 - [Python] test_serialization.py uses a python executable in PATH rather than that used for a test run
• ARROW-2049 - ARROW-2049: [Python] Use python -m cython to run Cython, instead of CYTHON_EXECUTABLE
• ARROW-2062 - [C++] Stalled builds in test_serialization.py in Travis CI
• ARROW-2070 - [Python] chdir logic in setup.py buggy
• ARROW-2072 - [Python] decimal128.byte_width crashes
• ARROW-2080 - [Python] Update documentation after ARROW-2024
• ARROW-2085 - HadoopFileSystem.isdir and .isfile should return False if the path doesn't exist
• ARROW-2106 - [Python] pyarrow.array can't take a pandas Series of python datetime objects.
• ARROW-2109 - [C++] Boost 1.66 compilation fails on Windows on linkage stage
• ARROW-2124 - [Python] ArrowInvalid raised if the first item of a nested list of numpy arrays is empty
• ARROW-2128 - [Python] Cannot serialize array of empty lists
• ARROW-2129 - [Python] Segmentation fault on conversion of empty array to Pandas
• ARROW-2131 - [Python] Serialization test fails on Windows when library has been built in place / not installed
• ARROW-2133 - [Python] Segmentation fault on conversion of empty nested arrays to Pandas
• ARROW-2135 - [Python] NaN values silently casted to int64 when passing explicit schema for conversion in Table.from_pandas
• ARROW-2145 - [Python] Decimal conversion not working for NaN values
• ARROW-2150 - [Python] array equality defaults to identity
• ARROW-2151 - [Python] Error when converting from list of uint64 arrays
• ARROW-2153 - [C++/Python] Decimal conversion not working for exponential notation
• ARROW-2157 - [Python] Decimal arrays cannot be constructed from Python lists
• ARROW-2160 - [C++/Python] Fix decimal precision inference
• ARROW-2161 - [Python] Skip test_cython_api if ARROW_HOME isn't defined
• ARROW-2162 - [Python/C++] Decimal Values with too-high precision are multiplied by 100
• ARROW-2167 - [C++] Building Orc extensions fails with the default BUILD_WARNING_LEVEL=Production
• ARROW-2170 - [Python] construct_metadata fails on reading files where no index was preserved
• ARROW-2171 - [Python] OwnedRef is fragile
• ARROW-2172 - [Python] Incorrect conversion from Numpy array when stride % itemsize != 0
• ARROW-2173 - [Python] NumPyBuffer destructor should hold the GIL
• ARROW-2175 - [Python] arrow_ep build is triggering during parquet-cpp build in Travis CI
• ARROW-2178 - [JS] Fix JS html FileReader example
• ARROW-2179 - [C++] arrow/util/io-util.h missing from libarrow-dev
• ARROW-2192 - Commits to master should run all builds in CI matrix
• ARROW-2209 - [Python] Partition columns are not correctly loaded in schema of ParquetDataset
• ARROW-2210 - [C++] TestBuffer_ResizeOOM has a memory leak with jemalloc
• ARROW-2212 - [C++/Python] Build Protobuf in base manylinux 1 docker image
• ARROW-2223 - [JS] installing umd release throws an error
• ARROW-2227 - [Python] Table.from_pandas does not create chunked_arrays.
• ARROW-2230 - [Python] JS version number is sometimes picked up
• ARROW-2232 - [Python] pyarrow.Tensor constructor segfaults
• ARROW-2234 - [JS] Read timestamp low bits as Uint32s
• ARROW-2240 - [Python] Array initialization with leading numpy nan fails with exception
• ARROW-2244 - [C++] Slicing NullArray should not cause the null count on the internal data to be unknown
• ARROW-2245 - [Python] Revert static linkage of parquet-cpp in manylinux1 wheel
• ARROW-2246 - [Python] Use namespaced boost in manylinux1 package
• ARROW-2251 - [GLib] Destroying GArrowBuffer while GArrowTensor that uses the buffer causes a crash
• ARROW-2254 - [Python] Local in-place dev versions picking up JS tags
• ARROW-2258 - [C++] Appveyor builds failing on master
• ARROW-2263 - [Python] test_cython.py fails if pyarrow is not in import path (e.g. with inplace builds)
• ARROW-2265 - [Python] Serializing subclasses of np.ndarray returns a np.ndarray.
• ARROW-2268 - Remove MD5 checksums from release process
• ARROW-2269 - [Python] Cannot build bdist_wheel for Python
• ARROW-2270 - [Python] ForeignBuffer doesn't tie Python object lifetime to C++ buffer lifetime
• ARROW-2272 - [Python] test_plasma spams /tmp
• ARROW-2275 - [C++] Buffer::mutable_data_ member uninitialized
• ARROW-2280 - [Python] pyarrow.Array.buffers should also include the offsets
• ARROW-2284 - [Python] test_plasma error on plasma_store error
• ARROW-2288 - [Python] slicing logic defective
• ARROW-2297 - [JS] babel-jest is not listed as a dev dependency
• ARROW-2304 - [C++] MultipleClients test in io-hdfs-test fails on trunk
• ARROW-2306 - [Python] HDFS test failures
• ARROW-2307 - [Python] Unable to read arrow stream containing 0 record batches
• ARROW-2311 - [Python] Struct array slicing defective
• ARROW-2312 - [JS] verify-release-candidate-sh must be updated to include JS in integration tests
• ARROW-2313 - [GLib] Release builds must define NDEBUG
• ARROW-2316 - [C++] Revert Buffer::mutable_data member to always inline
• ARROW-2318 - [C++] TestPlasmaStore.MultipleClientTest is flaky (hangs) in release builds
• ARROW-2320 - [C++] Vendored Boost build does not build regex library

Improvement

• ARROW-1021 - [Python] Add documentation about using pyarrow from other Cython and C++ projects
• ARROW-1035 - [Python] Add ASV benchmarks for streaming columnar deserialization
• ARROW-1463 - [JAVA] Restructure ValueVector hierarchy to minimize compile-time generated code
• ARROW-1579 - [Java] Add dockerized test setup to validate Spark integration
• ARROW-1580 - [Python] Instructions for setting up nightly builds on Linux
• ARROW-1623 - [C++] Add convenience method to construct Buffer from a string that owns its memory
• ARROW-1632 - [Python] Permit categorical conversions in Table.to_pandas on a per-column basis
• ARROW-1643 - [Python] Accept hdfs:// prefixes in parquet.read_table and attempt to connect to HDFS
• ARROW-1706 - [Python] StructArray.from_arrays should handle sequences that are coercible to arrays
• ARROW-1712 - [C++] Add method to BinaryBuilder to reserve space for value data
• ARROW-1835 - [C++] Create Arrow schema from std::tuple types
• ARROW-1861 - [Python] Fix up ASV setup, add developer instructions for writing new benchmarks and running benchmark suite locally
• ARROW-1872 - [Website] Populate hard-coded fields for current release from a YAML file
• ARROW-1927 - [Plasma] Implement delete function
• ARROW-1929 - [C++] Move various Arrow testing utility code from Parquet to Arrow codebase
• ARROW-1937 - [Python] Add documentation for different forms of constructing nested arrays from Python data structures
• ARROW-1942 - [C++] Hash table specializations for small integers
• ARROW-1947 - [Plasma] Change Client Create and Get to use Buffers
• ARROW-1951 - Add memcopy_threads to serialization context
• ARROW-1962 - [Java] Add reset() to ValueVector interface
• ARROW-1969 - [C++] Do not build ORC adapter by default
• ARROW-1977 - [C++] Update windows dev docs
• ARROW-1978 - [Website] Add more visible link to "Powered By" page to front page, simplify Powered By
• ARROW-2004 - [C++] Add shrink_to_fit option in BufferBuilder::Resize
• ARROW-2007 - [Python] Sequence converter for float32 not implemented
• ARROW-2011 - Allow setting the pickler to use in pyarrow serialization.
• ARROW-2012 - [GLib] Support "make distclean"
• ARROW-2018 - [C++] Build instruction on macOS and Homebrew is incomplete
• ARROW-2019 - Control the memory allocated for inner vector in LIST
• ARROW-2024 - [Python] Remove global SerializationContext variables
• ARROW-2028 - [Python] extra_cmake_args needs to be passed through shlex.split
• ARROW-2031 - HadoopFileSystem isn't pickleable
• ARROW-2035 - [C++] Update vendored cpplint.py to a Py3-compatible one
• ARROW-2036 - NativeFile should support standard IOBase methods
• ARROW-2042 - [Plasma] Revert API change of plasma::Create to output a MutableBuffer
• ARROW-2043 - [C++] Change description from OS X to macOS
• ARROW-2046 - [Python] Add support for PEP519 - pathlib and similar objects
• ARROW-2048 - [Python/C++] Upate Thrift pin to 0.11
• ARROW-2050 - Support setup.py pytest to automatically fetch the test dependencies
• ARROW-2064 - [GLib] Add common build problems link to the install section
• ARROW-2065 - Fix bug in SerializationContext.clone().
• ARROW-2068 - [Python] Expose Array's buffers to Python users
• ARROW-2069 - [Python] Document that Plasma is not (yet) supported on Windows
• ARROW-2071 - [Python] Reduce runtime of builds in Travis CI
• ARROW-2073 - [Python] Create StructArray from sequence of tuples given a known data type
• ARROW-2076 - [Python] Display slowest test durations
• ARROW-2083 - Support skipping builds
• ARROW-2084 - [C++] Support newer Brotli static library names
• ARROW-2086 - [Python] Shrink size of arrow_manylinux1_x86_64_base docker image
• ARROW-2087 - [Python] Binaries of 3rdparty are not stripped in manylinux1 base image
• ARROW-2088 - [GLib] Add GArrowNumericArray
• ARROW-2089 - [GLib] Rename to GARROW_TYPE_BOOLEAN for consistency
• ARROW-2090 - [Python] Add context manager methods to ParquetWriter
• ARROW-2093 - [Python] Possibly do not test pytorch serialization in Travis CI
• ARROW-2094 - [Python] Use toolchain libraries and PROTOBUF_HOME for protocol buffers
• ARROW-2095 - [C++] Suppress ORC EP build logging by default
• ARROW-2096 - [C++] Turn off Boost_DEBUG to trim build output
• ARROW-2099 - [Python] Support DictionaryArray::FromArrays in Python bindings
• ARROW-2107 - [GLib] Follow arrow::gpu::CudaIpcMemHandle API change
• ARROW-2110 - [Python] Only require pytest-runner on test commands
• ARROW-2111 - [C++] Linting could be faster
• ARROW-2117 - [C++] Pin clang to version 5.0
• ARROW-2118 - [Python] Improve error message when calling parquet.read_table on an empty file
• ARROW-2120 - Add possibility to use empty _MSVC_STATIC_LIB_SUFFIX for Thirdparties
• ARROW-2121 - [Python] Consider special casing object arrays in pandas serializers.
• ARROW-2132 - [Doc] Add links / mentions of Plasma store to main README
• ARROW-2137 - [Python] Don't print paths that are ignored when reading Parquet files
• ARROW-2138 - [C++] Have FatalLog abort instead of exiting
• ARROW-2142 - [Python] Conversion from Numpy struct array unimplemented
• ARROW-2143 - [Python] Provide a manylinux1 wheel for cp27m
• ARROW-2146 - [GLib] Implement Slice for ChunkedArray
• ARROW-2154 - [Python] __eq__ unimplemented on Buffer
• ARROW-2155 - [Python] pa.frombuffer(bytearray) returns immutable Buffer
• ARROW-2163 - Install apt dependencies separate from built-in Travis commands, retry on flakiness
• ARROW-2168 - [C++] Build toolchain builds with jemalloc
• ARROW-2169 - [C++] MSVC is complaining about uncaptured variables
• ARROW-2174 - [JS] Export format and schema enums
• ARROW-2177 - [C++] Remove support for specifying negative scale values in DecimalType
• ARROW-2180 - [C++] Remove APIs deprecated in 0.8.0 release
• ARROW-2181 - [Python] Add concat_tables to API reference, add documentation on use
• ARROW-2184 - [C++] Add static constructor for FileOutputStream returning shared_ptr to base OutputStream
• ARROW-2185 - Remove CI directives from squashed commit messages
• ARROW-2191 - [C++] Only use specific version of jemalloc
• ARROW-2198 - [Python] Docstring for parquet.read_table is misleading or incorrect
• ARROW-2199 - [JAVA] Follow up fixes for ARROW-2019. Ensure density driven capacity is never less than 1 and propagate density throughout the vector tree
• ARROW-2203 - [C++] StderrStream class
• ARROW-2204 - [C++] Build fails with TLS error on parquet-cpp clone
• ARROW-2206 - [JS] Add Perspective as a community project
• ARROW-2218 - [Python] PythonFile should infer mode when not given
• ARROW-2231 - [CI] Use clcache on AppVeyor
• ARROW-2238 - [C++] Detect clcache in cmake configuration
• ARROW-2250 - plasma_store process should cleanup on INT and TERM signals
• ARROW-2261 - [GLib] Can't share the same memory in GArrowBuffer safely
• ARROW-2279 - [Python] Better error message if lib cannot be found
• ARROW-2282 - [Python] Create StringArray from buffers
• ARROW-2283 - [C++] Support Arrow C++ installed in /usr detection by pkg-config
• ARROW-2289 - [GLib] Add Numeric, Integer and FloatingPoint data types
• ARROW-2291 - [C++] README missing instructions for libboost-regex-dev
• ARROW-2292 - [Python] More consistent / intuitive name for pyarrow.frombuffer
• ARROW-2321 - [C++] Release verification script fails with if CMAKE_INSTALL_LIBDIR is not $ARROW_HOME/lib
• ARROW-764 - [C++] Improve performance of CopyBitmap, add benchmarks

New Feature

• ARROW-1394 - [Plasma] Add optional extension for allocating memory on GPUs
• ARROW-1705 - [Python] Create StructArray from sequence of dicts given a known data type
• ARROW-1757 - [C++] Add DictionaryArray::FromArrays alternate ctor that can check or sanitized "untrusted" indices
• ARROW-1832 - [JS] Implement JSON reader for integration tests
• ARROW-1920 - Add support for reading ORC files
• ARROW-1926 - [GLib] Add garrow_timestamp_data_type_get_unit()
• ARROW-1930 - [C++] Implement Slice for ChunkedArray and Column
• ARROW-1931 - [C++] w4996 warning due to std::tr1 failing builds on Visual Studio 2017
• ARROW-1965 - [GLib] Add garrow_array_builder_get_value_data_type() and garrow_array_builder_get_value_type()
• ARROW-1970 - [GLib] Add garrow_chunked_array_get_value_data_type() and garrow_chunked_array_get_value_type()
• ARROW-2166 - [GLib] Implement Slice for Column
• ARROW-2176 - [C++] Extend DictionaryBuilder to support delta dictionaries
• ARROW-2190 - [GLib] Add add/remove field functions for RecordBatch.
• ARROW-2205 - [Python] Option for integer object nulls
• ARROW-2252 - [Python] Create buffer from address, size and base
• ARROW-2253 - [Python] Support __eq__ on scalar values
• ARROW-2262 - [Python] Support slicing on pyarrow.ChunkedArray
• ARROW-232 - C++/Parquet: Support writing chunked arrays as part of a table
• ARROW-633 - [Java] Add support for FixedSizeBinary type
• ARROW-634 - Add integration tests for FixedSizeBinary
• ARROW-969 - [C++/Python] Add add/remove field functions for RecordBatch

Sub-task

• ARROW-1815 - [Java] Rename MapVector to StructVector

Task

• ARROW-2052 - Unify OwnedRef and ScopedRef
• ARROW-2054 - Compilation warnings
• ARROW-2108 - [Python] Update instructions for ASV
• ARROW-2114 - [Python] Pull latest docker manylinux1 image
• ARROW-2123 - [JS] Upgrade to TS 2.7.1
• ARROW-2134 - [CI] Make Travis commit inspection more robust
• ARROW-2149 - [Python] reorganize test_convert_pandas.py
• ARROW-2156 - [CI] Isolate Sphinx dependencies
• ARROW-2197 - Document "undefined symbol" issue and workaround
• ARROW-2239 - [C++] Update build docs for Windows
• ARROW-2309 - [C++] Use std::make_unsigned

Test

• ARROW-1589 - [C++] Fuzzing for certain input formats

Apache Arrow 0.8.0 (12 December 2017)

Bug

• ARROW-1282 - Large memory reallocation by Arrow causes hang in jemalloc
• ARROW-1341 - [C++] Deprecate arrow::MakeTable in favor of new ctor from ARROW-1334
• ARROW-1347 - [JAVA] List null type should use consistent name for inner field
• ARROW-1398 - [Python] No support reading columns of type decimal(19,4)
• ARROW-1409 - [Format] Use for "page" attribute in Buffer in metadata
• ARROW-1540 - [C++] Fix valgrind warnings in cuda-test if possible
• ARROW-1541 - [C++] Race condition with arrow_gpu
• ARROW-1543 - [C++] row_wise_conversion example doesn't correspond to ListBuilder constructor arguments
• ARROW-1555 - [Python] write_to_dataset on s3
• ARROW-1584 - [PYTHON] serialize_pandas on empty dataframe
• ARROW-1585 - serialize_pandas round trip fails on integer columns
• ARROW-1586 - [PYTHON] serialize_pandas roundtrip loses columns name
• ARROW-1609 - Plasma: Build fails with Xcode 9.0
• ARROW-1615 - CXX flags for development more permissive than Travis CI builds
• ARROW-1617 - [Python] Do not use symlinks in python/cmake_modules
• ARROW-1620 - Python: Download Boost in manylinux1 build from bintray
• ARROW-1624 - [C++] Follow up fixes / tweaks to compiler warnings for Plasma / LLVM 4.0, add to readme
• ARROW-1625 - [Serialization] Support OrderedDict properly
• ARROW-1629 - [C++] Fix problematic code paths identified by infer tool
• ARROW-1633 - [Python] numpy "unicode" arrays not understood
• ARROW-1640 - Resolve OpenSSL issues in Travis CI
• ARROW-1647 - [Plasma] Potential bug when reading/writing messages.
• ARROW-1653 - [Plasma] Use static cast to avoid compiler warning.
• ARROW-1656 - [C++] Endianness Macro is Incorrect on Windows And Mac
• ARROW-1657 - [C++] Multithreaded Read Test Failing on Arch Linux
• ARROW-1658 - [Python] Out of bounds dictionary indices causes segfault after converting to pandas
• ARROW-1663 - [Java] Follow up on ARROW-1347 and make schema backward compatible
• ARROW-1670 - [Python] Speed up deserialization code path
• ARROW-1672 - [Python] Failure to write Feather bytes column
• ARROW-1673 - [Python] NumPy boolean arrays get converted to uint8 arrays on NdarrayToTensor roundtrip
• ARROW-1676 - [C++] Correctly truncate oversized validity bitmaps when writing Feather format
• ARROW-1678 - [Python] Incorrect serialization of numpy.float16
• ARROW-1680 - [Python] Timestamp unit change not done in from_pandas() conversion
• ARROW-1686 - Documentation generation script creates "apidocs" directory under site/java
• ARROW-1693 - [JS] Error reading dictionary-encoded integration test files
• ARROW-1695 - [Serialization] Fix reference counting of numpy arrays created in custom serialializer
• ARROW-1698 - [JS] File reader attempts to load the same dictionary batch more than once
• ARROW-1704 - [GLib] Go example in test suite is broken
• ARROW-1708 - [JS] Linter problem breaks master build
• ARROW-1709 - [C++] Decimal.ToString is incorrect for negative scale
• ARROW-1711 - [Python] flake8 checks still not failing builds
• ARROW-1714 - [Python] No named pd.Series name serialized as u'None'
• ARROW-1720 - [Python] Segmentation fault while trying to access an out-of-bound chunk
• ARROW-1723 - Windows: __declspec(dllexport) specified when building arrow static library
• ARROW-1730 - [Python] Incorrect result from pyarrow.array when passing timestamp type
• ARROW-1732 - [Python] RecordBatch.from_pandas fails on DataFrame with no columns when preserve_index=False
• ARROW-1735 - [C++] Cast kernels cannot write into sliced output array
• ARROW-1738 - [Python] Wrong datetime conversion when pa.array with unit
• ARROW-1739 - [Python] Fix usages of assertRaises causing broken build
• ARROW-1742 - C++: clang-format is not detected correct on OSX anymore
• ARROW-1743 - [Python] Table to_pandas fails when index contains categorical column
• ARROW-1745 - Compilation failure on Mac OS in plasma tests
• ARROW-1749 - [C++] Handle range of Decimal128 values that require 39 digits to be displayed
• ARROW-1751 - [Python] Pandas 0.21.0 introduces a breaking API change for MultiIndex construction
• ARROW-1754 - [Python] Fix buggy Parquet roundtrip when an index name is the same as a column name
• ARROW-1756 - [Python] Observed int32 overflow in Feather write/read path
• ARROW-1762 - [C++] unittest failure for language environment
• ARROW-1764 - [Python] Add -c conda-forge for Windows dev installation instructions
• ARROW-1766 - [GLib] Fix failing builds on OSX
• ARROW-1768 - [Python] Fix suppressed exception in ParquetWriter.__del__
• ARROW-1770 - [GLib] Fix GLib compiler warning
• ARROW-1771 - [C++] ARROW-1749 Breaks Public API test in parquet-cpp
• ARROW-1776 - [C++[ arrow::gpu::CudaContext::bytes_allocated() isn't defined
• ARROW-1778 - [Python] Link parquet-cpp statically, privately in manylinux1 wheels
• ARROW-1781 - [CI] OSX Builds on Travis-CI time out often
• ARROW-1788 - Plasma store crashes when trying to abort objects for disconnected client
• ARROW-1791 - Integration tests generate date[DAY] values outside of reasonable range
• ARROW-1793 - [Integration] fix a typo for README.md
• ARROW-1800 - [C++] Fix and simplify random_decimals
• ARROW-1805 - [Python] ignore non-parquet files when exploring dataset
• ARROW-1811 - [C++/Python] Rename all Decimal based APIs to Decimal128
• ARROW-1812 - Plasma store modifies hash table while iterating during client disconnect
• ARROW-1829 - [Plasma] Clean up eviction policy bookkeeping
• ARROW-1830 - [Python] Error when loading all the files in a dictionary
• ARROW-1836 - [C++] Fix C4996 warning from arrow/util/variant.h on MSVC builds
• ARROW-1840 - [Website] The installation command failed on Windows10 anaconda environment.
• ARROW-1845 - [Python] Expose Decimal128Type
• ARROW-1852 - [Plasma] Make retrieving manager file descriptor const
• ARROW-1853 - [Plasma] Fix off-by-one error in retry processing
• ARROW-1863 - [Python] PyObjectStringify could render bytes-like output for more types of objects
• ARROW-1865 - [C++] Adding a column to an empty Table fails
• ARROW-1869 - Fix typo in LowCostIdentityHashMap
• ARROW-1871 - [Python/C++] Appending Python Decimals with different scales requires rescaling
• ARROW-1873 - [Python] Segmentation fault when loading total 2GB of parquet files
• ARROW-1877 - Incorrect comparison in JsonStringArrayList.equals
• ARROW-1879 - [Python] Dask integration tests are not skipped if dask is not installed
• ARROW-1881 - [Python] setuptools_scm picks up JS version tags
• ARROW-1882 - [C++] Reintroduce DictionaryBuilder
• ARROW-1883 - [Python] BUG: Table.to_pandas metadata checking fails if columns are not present
• ARROW-1889 - [Python] --exclude is not available in older git versions
• ARROW-1890 - [Python] Masking for date32 arrays not working
• ARROW-1891 - [Python] NaT date32 values are only converted to nulls if from_pandas is used
• ARROW-1892 - [Python] Unknown list item type: binary
• ARROW-1893 - [Python] test_primitive_serialization fails on Python 2.7.3
• ARROW-1895 - [Python] Add field_name to pandas index metadata
• ARROW-1897 - [Python] Incorrect numpy_type for pandas metadata of Categoricals
• ARROW-1904 - [C++] Deprecate PrimitiveArray::raw_values
• ARROW-1906 - [Python] Creating a pyarrow.Array with timestamp of different unit is not casted
• ARROW-1908 - [Python] Construction of arrow table from pandas DataFrame with duplicate column names crashes
• ARROW-1910 - CPP README Brewfile link incorrect
• ARROW-1914 - [C++] make -j may fail to build with -DARROW_GPU=on
• ARROW-1915 - [Python] Parquet tests should be optional
• ARROW-1916 - [Java] Do not exclude java/dev/checkstyle from source releases
• ARROW-1917 - [GLib] Must set GI_TYPELIB_PATH in verify-release-candidate.sh
• ARROW-226 - [C++] libhdfs: feedback to help determining cause of failure in opening file path
• ARROW-641 - [C++] Do not build/run io-hdfs-test if ARROW_HDFS=off

Improvement

• ARROW-1087 - [Python] add get_include to expose directory containing header files
• ARROW-1134 - [C++] Allow C++/CLI projects to build with Arrow​
• ARROW-1178 - [Python] Create alternative to Table.from_pandas that yields a list of RecordBatch objects with a given chunk size
• ARROW-1226 - [C++] Improve / correct doxygen function documentation in arrow::ipc
• ARROW-1371 - [Website] Add "Powered By" page to the website
• ARROW-1455 - [Python] Add Dockerfile for validating Dask integration outside of usual CI
• ARROW-1488 - [C++] Implement ArrayBuilder::Finish in terms of internal::ArrayData
• ARROW-1498 - [GitHub] Add CONTRIBUTING.md and ISSUE_TEMPLATE.md
• ARROW-1503 - [Python] Add serialization callbacks for pandas objects in pyarrow.serialize
• ARROW-1522 - [C++] Support pyarrow.Buffer as built-in type in pyarrow.serialize
• ARROW-1523 - [C++] Add helper data struct with methods for reading a validity bitmap possibly having a non-zero offset
• ARROW-1524 - [C++] More graceful solution for handling non-zero offsets on inputs and outputs in compute library
• ARROW-1525 - [C++] Change functions in arrow/compare.h to not return Status
• ARROW-1526 - [Python] Unit tests to exercise code path in PARQUET-1100
• ARROW-1535 - [Python] Enable sdist source tarballs to build assuming that Arrow C++ libraries are available on the host system
• ARROW-1538 - [C++] Support Ubuntu 14.04 in .deb packaging automation
• ARROW-1539 - [C++] Remove functions deprecated as of 0.7.0 and prior releases
• ARROW-1556 - [C++] Incorporate AssertArraysEqual function from PARQUET-1100 patch
• ARROW-1588 - [C++/Format] Harden Decimal Format
• ARROW-1593 - [PYTHON] serialize_pandas should pass through the preserve_index keyword
• ARROW-1594 - [Python] Enable multi-threaded conversions in Table.from_pandas
• ARROW-1600 - [C++] Zero-copy Buffer constructor from std::string
• ARROW-1602 - [C++] Add IsValid/IsNotNull method to arrow::Array
• ARROW-1603 - [C++] Add BinaryArray method to get a value as a std::string
• ARROW-1604 - [Python] Support common type aliases in cast(...) and various type= arguments
• ARROW-1605 - [Python] pyarrow.array should be able to yield smaller integer types without an explicit cast
• ARROW-1607 - [C++] Implement DictionaryBuilder for Decimals
• ARROW-1613 - [Java] ArrowReader should not close the input ReadChannel
• ARROW-1616 - [Python] Add "write" method to RecordBatchStreamWriter that dispatches to write_table/write_back as appropriate
• ARROW-1626 - Add make targets to run the inter-procedural static analysis tool called "infer".
• ARROW-1627 - [JAVA] Reduce heap usage(Phase 2) - memory footprint in AllocationManager.BufferLedger
• ARROW-1630 - [Serialization] Support Python datetime objects
• ARROW-1635 - Add release management guide for PMCs
• ARROW-1641 - [C++] Do not include in public headers
• ARROW-1651 - [JS] Lazy row accessor in Table
• ARROW-1652 - [JS] Separate Vector into BatchVector and CompositeVector
• ARROW-1654 - [Python] pa.DataType cannot be pickled
• ARROW-1662 - Move OSX Dependency management into brew bundle Brewfiles
• ARROW-1665 - [Serialization] Support more custom datatypes in the default serialization context
• ARROW-1666 - [GLib] Enable gtk-doc on Travis CI Mac environment
• ARROW-1671 - [C++] Change arrow::MakeArray to not return Status
• ARROW-1675 - [Python] Use RecordBatch.from_pandas in FeatherWriter.write
• ARROW-1677 - [Blog] Add blog post on Ray and Arrow Python serialization
• ARROW-1679 - [GLib] Add garrow_record_batch_reader_read_next()
• ARROW-1683 - [Python] Restore "TimestampType" to pyarrow namespace
• ARROW-1684 - [Python] Simplify user API for reading nested Parquet columns
• ARROW-1689 - [Python] Categorical Indices Should Be Zero-Copy
• ARROW-1691 - [Java] Conform Java Decimal type implementation to format decisions in ARROW-1588
• ARROW-1701 - [Serialization] Support zero copy PyTorch Tensor serialization
• ARROW-1702 - Update jemalloc in manylinux1 build
• ARROW-1703 - [C++] Vendor exact version of jemalloc we depend on
• ARROW-1707 - Update dev README after movement to GitBox
• ARROW-1716 - [Format/JSON] Use string integer value for Decimals in JSON
• ARROW-1721 - [Python] Support null mask in places where it isn't supported in numpy_to_arrow.cc
• ARROW-1724 - [Packaging] Support Ubuntu 17.10
• ARROW-1725 - [Packaging] Upload .deb for Ubuntu 17.10
• ARROW-1726 - [GLib] Add setup description to verify C GLib build
• ARROW-1727 - [Format] Expand Arrow streaming format to permit new dictionaries and deltas / additions to existing dictionaries
• ARROW-1728 - [C++] Run clang-format checks in Travis CI
• ARROW-1737 - [GLib] Use G_DECLARE_DERIVABLE_TYPE
• ARROW-1746 - [Python] Add build dependencies for Arch Linux
• ARROW-1747 - [C++] Don't export symbols of statically linked libraries
• ARROW-1750 - [C++] Remove the need for arrow/util/random.h
• ARROW-1753 - [Python] Provide for matching subclasses with register_type in serialization context
• ARROW-1755 - [C++] Add build options for MSVC to use static runtime libraries
• ARROW-1758 - [Python] Remove pickle=True option for object serialization
• ARROW-1763 - [Python] DataType should be hashable
• ARROW-1765 - [Doc] Use dependencies from conda in C++ docker build
• ARROW-1785 - [Format/C++/Java] Remove VectorLayout metadata from Flatbuffers metadata
• ARROW-1787 - [Python] Support reading parquet files into DataFrames in a backward compatible way
• ARROW-1794 - [C++/Python] Rename DecimalArray to Decimal128Array
• ARROW-1801 - [Docs] Update install instructions to use red-data-tools repos
• ARROW-1808 - [C++] Make RecordBatch interface virtual to permit record batches that lazy-materialize columns
• ARROW-1809 - [GLib] Use .xml instead of .sgml for GTK-Doc main file
• ARROW-1810 - [Plasma] Remove test shell scripts
• ARROW-1817 - Configure JsonFileReader to read NaN for floats
• ARROW-1826 - [JAVA] Avoid branching at cell level (copyFrom)
• ARROW-1828 - [C++] Implement hash kernel specialization for BooleanType
• ARROW-1834 - [Doc] Build documentation in separate build folders
• ARROW-1838 - [C++] Use compute::Datum uniformly for input argument to kernels
• ARROW-1841 - [JS] Update text-encoding-utf-8 and tslib for node ESModules support
• ARROW-1849 - [GLib] Add input checks to GArrowRecordBatch
• ARROW-1850 - [C++] Use const void* in Writable::Write instead of const uint8_t*
• ARROW-1854 - [Python] Improve performance of serializing object dtype ndarrays
• ARROW-1855 - [GLib] Add workaround for build failure on macOS
• ARROW-1864 - [Java] Upgrade Netty to 4.1.x
• ARROW-1884 - [C++] Make JsonReader/JsonWriter classes internal APIs
• ARROW-1901 - [Python] Support recursive mkdir for DaskFilesystem
• ARROW-1902 - [Python] Remove mkdir race condition from write_to_dataset
• ARROW-1905 - [Python] Add more functions for checking exact types in pyarrow.types
• ARROW-1911 - Add Graphistry to Arrow JS proof points
• ARROW-905 - [Docs] Add Dockerfile for reproducible documentation generation
• ARROW-942 - Support integration testing on Python 2.7
• ARROW-950 - [Site] Add Google Analytics tag

New Feature

• ARROW-1032 - [JS] Support custom_metadata
• ARROW-1047 - [Java] Add generalized stream writer and reader interfaces that are decoupled from IO / message framing
• ARROW-1114 - [C++] Create Record Batch Builder class as a reusable and efficient way to transpose row-by-row data to columns
• ARROW-1250 - [Python] Define API for user type checking of array types
• ARROW-1482 - [C++] Implement casts between date32 and date64
• ARROW-1483 - [C++] Implement casts between time32 and time64
• ARROW-1484 - [C++] Implement (safe and unsafe) casts between timestamps and times of different units
• ARROW-1486 - [C++] Decide if arrow::RecordBatch needs to be copyable
• ARROW-1487 - [C++] Implement casts from List to List, where a cast function is defined from any A to B
• ARROW-1559 - [C++] Kernel implementations for "unique" (compute distinct elements of array)
• ARROW-1573 - [C++] Implement stateful kernel function that uses DictionaryBuilder to compute dictionary indices
• ARROW-1575 - [Python] Add pyarrow.column factory function
• ARROW-1577 - [JS] Package release script for NPM modules
• ARROW-1631 - [C++] Add GRPC to ThirdpartyToolchain.cmake
• ARROW-1637 - [C++] IPC round-trip for null type
• ARROW-1648 - C++: Add cast from Dictionary[NullType] to NullType
• ARROW-1649 - C++: Print number of nulls in PrettyPrint for NullArray
• ARROW-1667 - [GLib] Support Meson
• ARROW-1685 - [GLib] Add GArrowTableReader
• ARROW-1690 - [GLib] Add garrow_array_is_valid()
• ARROW-1697 - [GitHub] Add ISSUE_TEMPLATE.md
• ARROW-1718 - [Python] Implement casts from timestamp to date32/date64 and support in Array.from_pandas
• ARROW-1734 - C++/Python: Add cast function on Column-level
• ARROW-1736 - [GLib] Add GArrowCastOptions:allow-time-truncate
• ARROW-1748 - [GLib] Add GArrowRecordBatchBuilder
• ARROW-1752 - [Packaging] Add GPU packages for Debian and Ubuntu
• ARROW-1767 - [C++] Support file reads and writes over 2GB on Windows
• ARROW-1772 - [C++] Add public-api-test module in style of parquet-cpp
• ARROW-1773 - [C++] Add casts from date/time types to compatible signed integers
• ARROW-1775 - Ability to abort created but unsealed Plasma objects
• ARROW-1777 - [C++] Add static ctor ArrayData::Make for nicer syntax in places
• ARROW-1782 - [Python] Expose compressors as pyarrow.compress, pyarrow.decompress
• ARROW-1783 - [Python] Convert SerializedPyObject to/from sequence of component buffers with minimal memory allocation / copying
• ARROW-1784 - [Python] Read and write pandas.DataFrame in pyarrow.serialize by decomposing the BlockManager rather than coercing to Arrow format
• ARROW-1802 - [GLib] Add Arrow GPU support
• ARROW-1806 - [GLib] Add garrow_record_batch_writer_write_table()
• ARROW-1844 - [C++] Basic benchmark suite for hash kernels
• ARROW-1857 - [Python] Add switch for boost linkage with static parquet in wheels
• ARROW-1859 - [GLib] Add GArrowDictionaryDataType
• ARROW-1862 - [GLib] Add GArrowDictionaryArray
• ARROW-1874 - [GLib] Add garrow_array_unique()
• ARROW-1878 - [GLib] Add garrow_array_dictionary_encode()
• ARROW-480 - [Python] Add accessors for Parquet column statistics
• ARROW-504 - [Python] Add adapter to write pandas.DataFrame in user-selected chunk size to streaming format
• ARROW-507 - [C++/Python] Construct List container from offsets and values subarrays
• ARROW-541 - [JS] Implement JavaScript-compatible implementation
• ARROW-571 - [Python] Add APIs to build Parquet files incrementally from Arrow tables
• ARROW-587 - Add JIRA fix version to merge tool
• ARROW-609 - [C++] Function for casting from days since UNIX epoch to int64 date
• ARROW-838 - [Python] Efficient construction of arrays from non-pandas 1D NumPy arrays
• ARROW-972 - [Python] Add test cases and basic APIs for UnionArray

Sub-task

• ARROW-1471 - [JAVA] Document requirements and non/requirements for ValueVector updates
• ARROW-1472 - [JAVA] Design updated ValueVector Object Hierarchy
• ARROW-1473 - [JAVA] Create Prototype Code Hierarchy (Implementation Phase 1)
• ARROW-1474 - [JAVA] ValueVector hierarchy (Implementation Phase 2)
• ARROW-1476 - [JAVA] Implement final ValueVector updates
• ARROW-1710 - [Java] Remove non-nullable vectors in new vector class hierarchy
• ARROW-1717 - [Java] Remove public static helper method in vector classes for JSONReader/Writer
• ARROW-1719 - [Java] Remove accessor/mutator
• ARROW-1779 - [Java] Integration test breaks without zeroing out validity vectors
• ARROW-1819 - [Java] Remove legacy vector classes
• ARROW-1867 - [Java] Add BitVector APIs from old vector class
• ARROW-1885 - [Java] Restore previous MapVector class names

Task

• ARROW-1369 - Support boolean types in the javascript arrow reader library
• ARROW-1818 - Examine Java Dependencies
• ARROW-1827 - [Java] Add checkstyle config file and header file

Test

• ARROW-1549 - [JS] Integrate auto-generated Arrow test files
• ARROW-1821 - Add integration test case to explicitly check for optional validity buffer
• ARROW-1839 - [C++/Python] Add Decimal Parquet Read/Write Tests

Apache Arrow 0.7.1 (27 September 2017)

Bug

• ARROW-1497 - [Java] JsonFileReader doesn't set value count for some vectors
• ARROW-1500 - [C++] Result of ftruncate ignored in MemoryMappedFile::Create
• ARROW-1536 - [C++] Do not transitively depend on libboost_system
• ARROW-1542 - [C++] Windows release verification script should not modify conda environment
• ARROW-1544 - [JS] Export Vector type definitions
• ARROW-1545 - Int64Builder should not need int64() as arg
• ARROW-1550 - [Python] Fix flaky test on Windows
• ARROW-1554 - [Python] Document that pip wheels depend on MSVC14 runtime
• ARROW-1557 - [PYTHON] pyarrow.Table.from_arrays doesn't validate names length
• ARROW-1591 - C++: Xcode 9 is not correctly detected
• ARROW-1595 - [Python] Fix package dependency issues causing build failures
• ARROW-1601 - [C++] READ_NEXT_BITSET reads one byte past the last byte on last iteration
• ARROW-1606 - Python: Windows wheels don't include .lib files.
• ARROW-1610 - C++/Python: Only call python-prefix if the default PYTHON_LIBRARY is not present
• ARROW-1611 - Crash in BitmapReader when length is zero

Improvement

• ARROW-1537 - [C++] Support building with full path install_name on macOS
• ARROW-1546 - [GLib] Support GLib 2.40 again
• ARROW-1578 - [C++/Python] Run lint checks in Travis CI to fail for linting issues as early as possible
• ARROW-1608 - Support Release verification script on macOS
• ARROW-1612 - [GLib] add how to install for mac os to README

New Feature

• ARROW-1548 - [GLib] Support build append in builder
• ARROW-1592 - [GLib] Add GArrowUIntArrayBuilder

Test

• ARROW-1529 - [GLib] Fix failure on macOS on Travis CI

Wish

• ARROW-559 - Script to easily verify release in all languages

Apache Arrow 0.7.0 (12 September 2017)

Bug

• ARROW-1302 - C++: ${MAKE} variable not set sometimes on older MacOS installations
• ARROW-1354 - [Python] Segfault in Table.from_pandas with Mixed-Type Categories
• ARROW-1357 - [Python] Data corruption in reading multi-file parquet dataset
• ARROW-1363 - [C++] IPC writer sends buffer layout for dictionary rather than indices
• ARROW-1365 - [Python] Remove usage of removed jemalloc_memory_pool in Python API docs
• ARROW-1373 - [Java] Implement getBuffer() methods at the ValueVector interface
• ARROW-1375 - [C++] Visual Studio 2017 Appveyor builds failing
• ARROW-1379 - [Java] maven dependency issues - both unused and undeclared
• ARROW-1407 - Dictionaries can only hold a maximum of 4096 indices
• ARROW-1411 - [Python] Booleans in Float Columns cause Segfault
• ARROW-1414 - [GLib] Cast after status check
• ARROW-1421 - [Python] pyarrow.serialize cannot serialize a Python dict input
• ARROW-1426 - [Website] The title element of the top page is empty
• ARROW-1429 - [Python] Error loading parquet file with _metadata from HDFS
• ARROW-1430 - [Python] flake8 warnings are not failing CI builds
• ARROW-1434 - [C++/Python] pyarrow.Array.from_pandas does not support datetime64[D] arrays
• ARROW-1435 - [Python] PyArrow not propagating timezone information from Parquet to Python
• ARROW-1439 - [Packaging] Automate updating RPM in RPM build
• ARROW-1443 - [Java] Bug on ArrowBuf.setBytes with unsliced ByteBuffers
• ARROW-1444 - BitVector.splitAndTransfer copies last byte incorrectly
• ARROW-1446 - Python: Writing more than 2^31 rows from pandas dataframe causes row count overflow error
• ARROW-1450 - [Python] Raise proper error if custom serialization handler fails
• ARROW-1452 - [C++] Make UNUSED macro name more unique so it does not conflict with thirdparty projects
• ARROW-1453 - [Python] Implement WriteTensor for non-contiguous tensors
• ARROW-1458 - [Python] Document that HadoopFileSystem.mkdir with create_parents=False has no effect
• ARROW-1459 - [Python] PyArrow fails to load partitioned parquet files with non-primitive types
• ARROW-1461 - [C++] Disable builds using LLVM apt packages temporarily
• ARROW-1467 - [JAVA]: Fix reset() and allocateNew() in Nullable Value Vectors template
• ARROW-1490 - [Java] Allow Travis CI failures for JDK9 for now
• ARROW-1493 - [C++] Flush the output stream at the end of each PrettyPrint function
• ARROW-1495 - [C++] Store shared_ptr to boxed arrays in RecordBatch
• ARROW-1507 - [C++] arrow/compute/api.h can't be used without arrow/array.h
• ARROW-1512 - [Docs] NumericArray has no member named 'raw_data'
• ARROW-1514 - [C++] Fix a typo in document
• ARROW-1527 - Fix Travis JDK9 build
• ARROW-1531 - [C++] Return ToBytes by value from Decimal128
• ARROW-1532 - [Python] Referencing an Empty Schema causes a SegFault
• ARROW-407 - BitVector.copyFromSafe() should re-allocate if necessary instead of returning false
• ARROW-801 - [JAVA] Provide direct access to underlying buffer memory addresses in consistent way without generating garbage or large amount indirections

Improvement

• ARROW-1307 - [Python] Add pandas serialization section + Feather API to Sphinx docs
• ARROW-1317 - [Python] Add function to set Hadoop CLASSPATH
• ARROW-1331 - [Java] Refactor tests
• ARROW-1339 - [C++] Use boost::filesystem for handling of platform-specific file path encodings
• ARROW-1344 - [C++] Calling BufferOutputStream::Write after calling Finish crashes
• ARROW-1348 - [C++/Python] Add release verification script for Windows
• ARROW-1351 - Automate updating CHANGELOG.md as part of release scripts
• ARROW-1352 - [Integration] Improve print formatting for producer, consumer line
• ARROW-1355 - Make arrow buildable with java9
• ARROW-1356 - [Website] Add new committers
• ARROW-1358 - Update source release scripts to account for new SHA checksum policy
• ARROW-1359 - [Python] Add Parquet writer option to normalize field names for use in Spark
• ARROW-1366 - [Python] Add instructions for starting the Plasma store when installing pyarrow from wheels
• ARROW-1372 - [Plasma] Support for storing data in huge pages
• ARROW-1376 - [C++] RecordBatchStreamReader::Open API is inconsistent with writer
• ARROW-1381 - [Python] Improve performance of SerializedPyObject.to_buffer
• ARROW-1383 - [C++] Support std::vector in builder vector appends
• ARROW-1384 - [C++] Add convenience function for serializing a record batch to an IPC message
• ARROW-1386 - [C++] Unpin CMake version in MSVC build toolchain
• ARROW-1395 - [C++] Remove APIs deprecated as of 0.5.0 and later versions
• ARROW-1397 - [Packaging] Use Docker instead of Vagrant
• ARROW-1401 - [C++] Add extra debugging context to failures in RETURN_NOT_OK in debug builds
• ARROW-1402 - [C++] Possibly deprecate public APIs that use MutableBuffer
• ARROW-1404 - [Packaging] Build .deb and .rpm on Travis CI
• ARROW-1405 - [Python] Add logging option for verbose memory allocations
• ARROW-1406 - [Python] Harden user API for generating serialized schema and record batch messages as memoryview-compatible objects
• ARROW-1408 - [C++] Refactor and make IPC read / write APIs more consistent, add appropriate deprecations
• ARROW-1410 - Plasma object store occasionally pauses for a long time
• ARROW-1412 - [Plasma] Add higher level API for putting and getting Python objects
• ARROW-1413 - [C++] Add include-what-you-use configuration
• ARROW-1416 - [Format] Clarify example array in memory layout documentation
• ARROW-1418 - [Python] Introduce SerializationContext to register custom serialization callbacks
• ARROW-1419 - [GLib] Suppress sign-conversion warning on Clang
• ARROW-1427 - [GLib] Add a link to readme of Arrow GLib
• ARROW-1428 - [C++] Append steps to clone source code to README.mb
• ARROW-1432 - [C++] Build bundled jemalloc functions with private prefix
• ARROW-1433 - [C++] Simplify implementation of Array::Slice
• ARROW-1438 - [Plasma] Pull SerializationContext through PlasmaClient put and get
• ARROW-1441 - [Site] Add Ruby to Flexible section
• ARROW-1442 - [Website] Add pointer to nightly conda packages on /install
• ARROW-1447 - [C++] Round of include-what-you-use include cleanups
• ARROW-1448 - [Packaging] Support uploading built .deb and .rpm to Bintray
• ARROW-1449 - Implement Decimal using only Int128
• ARROW-1451 - [C++] Create arrow/io/api.h
• ARROW-1460 - [C++] Upgrade clang-format used to LLVM 4.0
• ARROW-1466 - [C++] Support DecimalArray in arrow::PrettyPrint
• ARROW-1468 - [C++] Append to PrimitiveBuilder from std::vector
• ARROW-1480 - [Python] Improve performance of serializing sets
• ARROW-1494 - [C++] Document that shared_ptr returned by RecordBatch::column needs to be retained
• ARROW-1499 - [Python] Consider adding option to parquet.write_table that sets options for maximum Spark compatibility
• ARROW-1505 - [GLib] Simplify arguments check
• ARROW-1506 - [C++] Support pkg-config for compute modules
• ARROW-1508 - C++: Add support for FixedSizeBinaryType in DictionaryBuilder
• ARROW-1511 - [C++] Deprecate arrow::MakePrimitiveArray
• ARROW-1513 - C++: Add cast from Dictionary to plain arrays
• ARROW-1515 - [GLib] Detect version directly
• ARROW-1516 - [GLib] Update document
• ARROW-1517 - Remove unnecessary temporary in DecimalUtil::ToString function
• ARROW-1519 - [C++] Move DecimalUtil functions to methods on the Int128 class
• ARROW-1528 - [GLib] Resolve include dependency
• ARROW-1530 - [C++] Install arrow/util/parallel.h
• ARROW-594 - [Python] Provide interface to write pyarrow.Table to a stream
• ARROW-786 - [Format] In-memory format for 128-bit Decimals, handling of sign bit
• ARROW-837 - [Python] Expose buffer allocation, FixedSizeBufferWriter
• ARROW-941 - [Docs] Improve "cold start" integration testing instructions

New Feature

• ARROW-1156 - [Python] pyarrow.Array.from_pandas should take a type parameter
• ARROW-1238 - [Java] Add JSON read/write support for decimals for integration tests
• ARROW-1364 - [C++] IPC reader and writer specialized for GPU device memory
• ARROW-1377 - [Python] Add function to assist with benchmarking Parquet scan performance
• ARROW-1387 - [C++] Set up GPU leaf library build toolchain
• ARROW-1392 - [C++] Implement reader and writer IO interfaces for GPU buffers
• ARROW-1396 - [C++] Add PrettyPrint function for Schemas, which also outputs any dictionaries
• ARROW-1399 - [C++] Add CUDA build version in a public header to help prevent ABI conflicts
• ARROW-1400 - [Python] Ability to create partitions when writing to Parquet
• ARROW-1415 - [GLib] Support date32 and date64
• ARROW-1417 - [Python] Allow more generic filesystem objects to be passed to ParquetDataset
• ARROW-1462 - [GLib] Support time array
• ARROW-1479 - [JS] Expand JavaScript implementation
• ARROW-1481 - [C++] Expose type casts as generic callable object that can write into pre-allocated memory
• ARROW-1504 - [GLib] Support timestamp
• ARROW-1510 - [C++] Support cast
• ARROW-229 - [C++] Implement safe casts for primitive types
• ARROW-592 - [C++] Provide .deb and .rpm packages
• ARROW-695 - Integration tests for Decimal types
• ARROW-696 - [C++] Add JSON read/write support for decimals for integration tests
• ARROW-759 - [Python] Implement a transient list serialization function that can handle a mix of scalars, lists, ndarrays, dicts
• ARROW-989 - [Python] Write pyarrow.Table to FileWriter or StreamWriter

Test

• ARROW-1390 - [Python] Extend tests for python serialization

Apache Arrow 0.6.0 (14 August 2017)

Bug

• ARROW-1192 - [JAVA] Improve splitAndTransfer performance for List and Union vectors
• ARROW-1195 - [C++] CpuInfo doesn't get cache size on Windows
• ARROW-1204 - [C++] lz4 ExternalProject fails in Visual Studio 2015
• ARROW-1225 - [Python] pyarrow.array does not attempt to convert bytes to UTF8 when passed a StringType
• ARROW-1237 - [JAVA] Expose the ability to set lastSet
• ARROW-1239 - issue with current version of git-commit-id-plugin
• ARROW-1240 - security: upgrade logback to address CVE-2017-5929
• ARROW-1242 - [Java] security - upgrade Jackson to mitigate 3 CVE vulnerabilities
• ARROW-1245 - [Integration] Java Integration Tests Disabled
• ARROW-1248 - [Python] C linkage warnings in Clang with public Cython API
• ARROW-1249 - [JAVA] Expose the fillEmpties function from NullableVector.mutator
• ARROW-1263 - [C++] CpuInfo should be able to get CPU features on Windows
• ARROW-1265 - [Plasma] Plasma store memory leak warnings in Python test suite
• ARROW-1267 - [Java] Handle zero length case in BitVector.splitAndTransfer
• ARROW-1269 - [Packaging] Add Windows wheel build scripts from ARROW-1068 to arrow-dist
• ARROW-1275 - [C++] Default static library prefix for Snappy should be "_static"
• ARROW-1276 - Cannot serializer empty DataFrame to parquet
• ARROW-1283 - [Java] VectorSchemaRoot should be able to be closed() more than once
• ARROW-1285 - PYTHON: NotImplemented exception creates empty parquet file
• ARROW-1287 - [Python] Emulate "whence" argument of seek in NativeFile
• ARROW-1290 - [C++] Use array capacity doubling in arrow::BufferBuilder
• ARROW-1291 - [Python] pa.RecordBatch.from_pandas doesn't accept DataFrame with numeric column names
• ARROW-1294 - [C++] New Appveyor build failures
• ARROW-1296 - [Java] templates/FixValueVectors reset() method doesn't set allocationSizeInBytes correctly
• ARROW-1300 - [JAVA] Fix ListVector Tests
• ARROW-1306 - [Python] Encoding? issue with error reporting for parquet.read_table
• ARROW-1308 - [C++] ld tries to link arrow_static even when -DARROW_BUILD_STATIC=off
• ARROW-1309 - [Python] Error inferring List type in Array.from_pandas when inner values are all None
• ARROW-1310 - [JAVA] Revert ARROW-886
• ARROW-1312 - [C++] Set default value to ARROW_JEMALLOC to OFF until ARROW-1282 is resolved
• ARROW-1326 - [Python] Fix Sphinx build in Travis CI
• ARROW-1327 - [Python] Failing to release GIL in MemoryMappedFile._open causes deadlock
• ARROW-1328 - [Python] pyarrow.Table.from_pandas option timestamps_to_ms changes column values
• ARROW-1330 - [Plasma] Turn on plasma tests on manylinux1
• ARROW-1335 - [C++] PrimitiveArray::raw_values has inconsistent semantics re: offsets compared with subclasses
• ARROW-1338 - [Python] Investigate non-deterministic core dump on Python 2.7, Travis CI builds
• ARROW-1340 - [Java] NullableMapVector field doesn't maintain metadata
• ARROW-1342 - [Python] Support strided array of lists
• ARROW-1343 - [Format/Java/C++] Ensuring encapsulated stream / IPC message sizes are always a multiple of 8
• ARROW-1350 - [C++] Include Plasma source tree in source distribution
• ARROW-187 - [C++] Decide on how pedantic we want to be about exceptions
• ARROW-276 - [JAVA] Nullable Value Vectors should extend BaseValueVector instead of BaseDataValueVector
• ARROW-573 - [Python/C++] Support ordered dictionaries data, pandas Categorical
• ARROW-884 - [C++] Exclude internal classes from documentation
• ARROW-932 - [Python] Fix compiler warnings on MSVC
• ARROW-968 - [Python] RecordBatch [i:j] syntax is incomplete

Improvement

• ARROW-1093 - [Python] Fail Python builds if flake8 yields warnings
• ARROW-1121 - [C++] Improve error message when opening OS file fails
• ARROW-1140 - [C++] Allow optional build of plasma
• ARROW-1149 - [Plasma] Create Cython client library for Plasma
• ARROW-1173 - [Plasma] Blog post for Plasma
• ARROW-1211 - [C++] Consider making default_memory_pool() the default for builder classes
• ARROW-1213 - [Python] Enable s3fs to be used with ParquetDataset and reader/writer functions
• ARROW-1219 - [C++] Use more vanilla Google C++ formatting
• ARROW-1224 - [Format] Clarify language around buffer padding and alignment in IPC
• ARROW-1230 - [Plasma] Install libraries and headers
• ARROW-1243 - [Java] security: upgrade all libraries to latest stable versions
• ARROW-1251 - [Python/C++] Revise build documentation to account for latest build toolchain
• ARROW-1253 - [C++] Use pre-built toolchain libraries where prudent to speed up CI builds
• ARROW-1255 - [Plasma] Check plasma flatbuffer messages with the flatbuffer verifier
• ARROW-1257 - [Plasma] Plasma documentation
• ARROW-1258 - [C++] Suppress dlmalloc warnings on Clang
• ARROW-1259 - [Plasma] Speed up Plasma tests
• ARROW-1260 - [Plasma] Use factory method to create Python PlasmaClient
• ARROW-1264 - [Plasma] Don't exit the Python interpreter if the plasma client can't connect to the store
• ARROW-1274 - [C++] add_compiler_export_flags() throws warning with CMake >= 3.3
• ARROW-1288 - Clean up many ASF license headers
• ARROW-1289 - [Python] Add PYARROW_BUILD_PLASMA option like Parquet
• ARROW-1301 - [C++/Python] Add remaining supported libhdfs UNIX-like filesystem APIs
• ARROW-1303 - [C++] Support downloading Boost
• ARROW-1315 - [GLib] Status check of arrow::ArrayBuilder::Finish() is missing
• ARROW-1323 - [GLib] Add garrow_boolean_array_get_values()
• ARROW-1333 - [Plasma] Sorting example for DataFrames in plasma
• ARROW-1334 - [C++] Instantiate arrow::Table from vector of Array objects (instead of Columns)

New Feature

• ARROW-1076 - [Python] Handle nanosecond timestamps more gracefully when writing to Parquet format
• ARROW-1104 - Integrate in-memory object store from Ray
• ARROW-1246 - [Format] Add Map logical type to metadata
• ARROW-1268 - [Website] Blog post on Arrow integration with Spark
• ARROW-1281 - [C++/Python] Add Docker setup for running HDFS tests and other tests we may not run in Travis CI
• ARROW-1305 - [GLib] Add GArrowIntArrayBuilder
• ARROW-1336 - [C++] Add arrow::schema factory function
• ARROW-439 - [Python] Add option in to_pandas conversions to yield Categorical from String/Binary arrays
• ARROW-622 - [Python] Investigate alternatives to timestamps_to_ms argument in pandas conversion

Task

• ARROW-1270 - [Packaging] Add Python wheel build scripts for macOS to arrow-dist
• ARROW-1272 - [Python] Add script to arrow-dist to generate and upload manylinux1 Python wheels
• ARROW-1273 - [Python] Add convenience functions for reading only Parquet metadata or effective Arrow schema from a particular Parquet file
• ARROW-1297 - 0.6.0 Release
• ARROW-1304 - [Java] Fix checkstyle checks warning

Test

• ARROW-1241 - [C++] Visual Studio 2017 Appveyor build job

Apache Arrow 0.5.0 (23 July 2017)

Bug

• ARROW-1074 - from_pandas doesnt convert ndarray to list
• ARROW-1079 - [Python] Empty "private" directories should be ignored by Parquet interface
• ARROW-1081 - C++: arrow::test::TestBase::MakePrimitive doesn't fill null_bitmap
• ARROW-1096 - [C++] Memory mapping file over 4GB fails on Windows
• ARROW-1097 - Reading tensor needs file to be opened in writeable mode
• ARROW-1098 - Document Error?
• ARROW-1101 - UnionListWriter is not implementing all methods on interface ScalarWriter
• ARROW-1103 - [Python] Utilize pandas metadata from common _metadata Parquet file if it exists
• ARROW-1107 - [JAVA] NullableMapVector getField() should return nullable type
• ARROW-1108 - Check if ArrowBuf is empty buffer in getActualConsumedMemory() and getPossibleConsumedMemory()
• ARROW-1109 - [JAVA] transferOwnership fails when readerIndex is not 0
• ARROW-1110 - [JAVA] make union vector naming consistent
• ARROW-1111 - [JAVA] Make aligning buffers optional, and allow -1 for unknown null count
• ARROW-1112 - [JAVA] Set lastSet for VarLength and List vectors when loading
• ARROW-1113 - [C++] gflags EP build gets triggered (as a no-op) on subsequent calls to make or ninja build
• ARROW-1115 - [C++] Use absolute path for ccache
• ARROW-1117 - [Docs] Minor issues in GLib README
• ARROW-1124 - [Python] pyarrow needs to depend on numpy>=1.10 (not 1.9)
• ARROW-1125 - Python: Table.from_pandas doesn't work anymore on partial schemas
• ARROW-1128 - [Docs] command to build a wheel is not properly rendered
• ARROW-1129 - [C++] Fix Linux toolchain build regression from ARROW-742
• ARROW-1131 - Python: Parquet unit tests are always skipped
• ARROW-1132 - [Python] Unable to write pandas DataFrame w/MultiIndex containing duplicate values to parquet
• ARROW-1136 - [C++/Python] Segfault on empty stream
• ARROW-1138 - Travis: Use OpenJDK7 instead of OracleJDK7
• ARROW-1139 - [C++] dlmalloc doesn't allow arrow to be built with clang 4 or gcc 7.1.1
• ARROW-1141 - on import get libjemalloc.so.2: cannot allocate memory in static TLS block
• ARROW-1143 - C++: Fix comparison of NullArray
• ARROW-1144 - [C++] Remove unused variable
• ARROW-1150 - [C++] AdaptiveIntBuilder compiler warning on MSVC
• ARROW-1152 - [Cython] read_tensor should work with a readable file
• ARROW-1155 - segmentation fault when run pa.Int16Value()
• ARROW-1157 - C++/Python: Decimal templates are not correctly exported on OSX
• ARROW-1159 - [C++] Static data members cannot be accessed from inline functions in Arrow headers by thirdparty users
• ARROW-1162 - Transfer Between Empty Lists Should Not Invoke Callback
• ARROW-1166 - Errors in Struct type's example and missing reference in Layout.md
• ARROW-1167 - [Python] Create chunked BinaryArray in Table.from_pandas when a column's data exceeds 2GB
• ARROW-1168 - [Python] pandas metadata may contain "mixed" data types
• ARROW-1169 - C++: jemalloc externalproject doesn't build with CMake's ninja generator
• ARROW-1170 - C++: ARROW_JEMALLOC=OFF breaks linking on unittest
• ARROW-1174 - [GLib] Investigate root cause of ListArray glib test failure
• ARROW-1177 - [C++] Detect int32 overflow in ListBuilder::Append
• ARROW-1179 - C++: Add missing virtual destructors
• ARROW-1180 - [GLib] garrow_tensor_get_dimension_name() returns invalid address
• ARROW-1181 - [Python] Parquet test fail if not enabled
• ARROW-1182 - C++: Specify BUILD_BYPRODUCTS for zlib and zstd
• ARROW-1186 - [C++] Enable option to build arrow with minimal dependencies needed to build Parquet library
• ARROW-1188 - Segfault when trying to serialize a DataFrame with Null-only Categorical Column
• ARROW-1190 - VectorLoader corrupts vectors with duplicate names
• ARROW-1191 - [JAVA] Implement getField() method for the complex readers
• ARROW-1194 - Getting record batch size with pa.get_record_batch_size returns a size that is too small for pandas DataFrame.
• ARROW-1197 - [GLib] record_batch.hpp Inclusion is missing
• ARROW-1200 - [C++] DictionaryBuilder should use signed integers for indices
• ARROW-1201 - [Python] Incomplete Python types cause a core dump when repr-ing
• ARROW-1203 - [C++] Disallow BinaryBuilder to append byte strings larger than the maximum value of int32_t
• ARROW-1205 - C++: Reference to type objects in ArrayLoader may cause segmentation faults.
• ARROW-1206 - [C++] Enable MSVC builds to work with some compression library support disabled
• ARROW-1208 - [C++] Toolchain build with ZSTD library from conda-forge failure
• ARROW-1215 - [Python] Class methods in API reference
• ARROW-1216 - Numpy arrays cannot be created from Arrow Buffers on Python 2
• ARROW-1218 - Arrow doesn't compile if all compression libraries are deactivated
• ARROW-1222 - [Python] pyarrow.array returns NullArray for array of unsupported Python objects
• ARROW-1223 - [GLib] Fix function name that returns wrapped object
• ARROW-1235 - [C++] macOS linker failure with operator<< and std::ostream
• ARROW-1236 - Library paths in exported pkg-config file are incorrect
• ARROW-601 - Some logical types not supported when loading Parquet
• ARROW-784 - Cleaning up thirdparty toolchain support in Arrow on Windows
• ARROW-992 - [Python] In place development builds do not have a __version__

Improvement

• ARROW-1041 - [Python] Support read_pandas on a directory of Parquet files
• ARROW-1100 - [Python] Add "mode" property to NativeFile instances
• ARROW-1102 - Make MessageSerializer.serializeMessage() public
• ARROW-1120 - [Python] Write support for int96
• ARROW-1137 - Python: Ensure Pandas roundtrip of all-None column
• ARROW-1148 - [C++] Raise minimum CMake version to 3.2
• ARROW-1151 - [C++] Add gcc branch prediction to status check macro
• ARROW-1160 - C++: Implement DictionaryBuilder
• ARROW-1165 - [C++] Refactor PythonDecimalToArrowDecimal to not use templates
• ARROW-1185 - [C++] Clean up arrow::Status implementation, add warn_unused_result attribute for clang
• ARROW-1187 - Serialize a DataFrame with None column
• ARROW-1193 - [C++] Support pkg-config for arrow_python.so
• ARROW-1196 - [C++] Appveyor separate jobs for Debug/Release builds from sources; Build with conda toolchain; Build with NMake Makefiles Generator
• ARROW-1199 - [C++] Introduce mutable POD struct for generic array data
• ARROW-1202 - Remove semicolons from status macros
• ARROW-1217 - [GLib] Add GInputStream based arrow::io::RandomAccessFile
• ARROW-1220 - [C++] Standartize usage of *_HOME cmake script variables for 3rd party libs
• ARROW-1221 - [C++] Pin clang-format version
• ARROW-1229 - [GLib] Follow Reader API change (get -> read)
• ARROW-742 - Handling exceptions during execution of std::wstring_convert
• ARROW-834 - [Python] Support creating Arrow arrays from Python iterables
• ARROW-915 - Struct Array reads limited support
• ARROW-935 - [Java] Build Javadoc in Travis CI
• ARROW-960 - [Python] Add source build guide for macOS + Homebrew
• ARROW-962 - [Python] Add schema attribute to FileReader
• ARROW-966 - [Python] pyarrow.list_ should also accept Field instance
• ARROW-978 - [Python] Use sphinx-bootstrap-theme for Sphinx documentation

New Feature

• ARROW-1048 - Allow user LD_LIBRARY_PATH to be used with source release script
• ARROW-1073 - C++: Adapative integer builder
• ARROW-1095 - [Website] Add Arrow icon asset
• ARROW-111 - [C++] Add static analyzer to tool chain to verify checking of Status returns
• ARROW-1122 - [Website] Guest blog post on Arrow + ODBC from turbodbc
• ARROW-1123 - C++: Make jemalloc the default allocator
• ARROW-1135 - Upgrade Travis CI clang builds to use LLVM 4.0
• ARROW-1142 - [C++] Move over compression library toolchain from parquet-cpp
• ARROW-1145 - [GLib] Add get_values()
• ARROW-1154 - [C++] Migrate more computational utility code from parquet-cpp
• ARROW-1183 - [Python] Implement time type conversions in to_pandas
• ARROW-1198 - Python: Add public C++ API to unwrap PyArrow object
• ARROW-1212 - [GLib] Add garrow_binary_array_get_offsets_buffer()
• ARROW-1214 - [Python] Add classes / functions to enable stream message components to be handled outside of the stream reader class
• ARROW-1227 - [GLib] Support GOutputStream
• ARROW-460 - [C++] Implement JSON round trip for DictionaryArray
• ARROW-462 - [C++] Implement in-memory conversions between non-nested primitive types and DictionaryArray equivalent
• ARROW-575 - Python: Auto-detect nested lists and nested numpy arrays in Pandas
• ARROW-597 - [Python] Add convenience function to yield DataFrame from any object that a StreamReader or FileReader can read from
• ARROW-599 - [C++] Add LZ4 codec to 3rd-party toolchain
• ARROW-600 - [C++] Add ZSTD codec to 3rd-party toolchain
• ARROW-692 - Java<->C++ Integration tests for dictionary-encoded vectors
• ARROW-693 - [Java] Add JSON support for dictionary vectors

Task

• ARROW-1052 - Arrow 0.5.0 release

Test

• ARROW-1228 - [GLib] Test file name should be the same name as target class
• ARROW-1233 - [C++] Validate cmake script resolving of 3rd party linked libs from correct location in toolchain build

Apache Arrow 0.4.1 (9 June 2017)

Bug

• ARROW-1039 - Python: pyarrow.Filesystem.read_parquet causing error if nthreads>1
• ARROW-1050 - [C++] Export arrow::ValidateArray
• ARROW-1051 - [Python] If pyarrow.parquet fails to import due to a shared library ABI conflict, the test_parquet.py tests silently do not run
• ARROW-1056 - [Python] Parquet+HDFS test failure due to writing pandas index
• ARROW-1057 - Fix cmake warning and msvc debug asserts
• ARROW-1062 - [GLib] Examples use old API
• ARROW-1066 - remove warning on feather for pandas >= 0.20.1
• ARROW-1070 - [C++] Feather files for date/time types should be written with the physical types
• ARROW-1075 - [GLib] Build error on macOS
• ARROW-1085 - [java] Follow up on template cleanup. Missing method for IntervalYear
• ARROW-1086 - [Python] pyarrow 0.4.0 on pypi is missing pxd files
• ARROW-1088 - [Python] test_unicode_filename test fails when unicode filenames aren't supported by system
• ARROW-1090 - [Python] build_ext usability
• ARROW-1091 - Decimal scale and precision are flipped
• ARROW-1092 - More Decimal and scale flipped follow-up
• ARROW-1094 - [C++] Incomplete buffer reads in arrow::io::ReadableFile should exactly truncate returned buffer
• ARROW-424 - [C++] Threadsafety in arrow/io/hdfs.h

Improvement

• ARROW-1020 - [Format] Add additional language to Schema.fbs to clarify naive vs. localized Timestamp values
• ARROW-1034 - [Python] Enable creation of binary wheels on Windows / MSVC
• ARROW-1049 - [java] vector template cleanup
• ARROW-1063 - [Website] Blog post and website updates for 0.4.0 release
• ARROW-1078 - [Python] Account for PARQUET-967
• ARROW-1080 - C++: Add tutorial about converting to/from row-wise representation
• ARROW-897 - [GLib] Build arrow-glib as a separate build in the Travis CI build matrix
• ARROW-986 - [Format] Update IPC.md to account for dictionary batches
• ARROW-990 - [JS] Add tslint support for linting TypeScript

Task

• ARROW-1068 - [Python] Create external repo with appveyor.yml configured for building Python wheel installers
• ARROW-1069 - Add instructions for publishing maven artifacts
• ARROW-1084 - Implementations of BufferAllocator should handle Netty's OutOfDirectMemoryError

Test

• ARROW-1060 - [Python] Add unit test for ARROW-1053
• ARROW-1082 - [GLib] Add CI on macOS

Apache Arrow 0.4.0 (22 May 2017)

Bug

• ARROW-1003 - [C++] Hdfs and java dlls fail to load when built for Windows with MSVC
• ARROW-1004 - ArrowInvalid: Invalid: Python object of type float is not None and is not a string, bool, or date object
• ARROW-1017 - Python: Table.to_pandas leaks memory
• ARROW-1023 - Python: Fix bundling of arrow-cpp for macOS
• ARROW-1033 - [Python] pytest discovers scripts/test_leak.py
• ARROW-1046 - [Python] Conform DataFrame metadata to pandas spec
• ARROW-1053 - [Python] Memory leak with RecordBatchFileReader
• ARROW-1054 - [Python] Test suite fails on pandas 0.19.2
• ARROW-1061 - [C++] Harden decimal parsing against invalid strings
• ARROW-1064 - ModuleNotFoundError: No module named 'pyarrow._parquet'
• ARROW-813 - [Python] setup.py sdist must also bundle dependent cmake modules
• ARROW-824 - Date and Time Vectors should reflect timezone-less semantics
• ARROW-856 - CmakeError by Unknown compiler.
• ARROW-881 - [Python] Reconstruct Pandas DataFrame indexes using custom_metadata
• ARROW-909 - libjemalloc.so.2: cannot open shared object file:
• ARROW-939 - Fix division by zero for zero-dimensional Tensors
• ARROW-940 - [JS] Generate multiple sets of artifacts
• ARROW-944 - Python: Compat broken for pandas==0.18.1
• ARROW-948 - [GLib] Update C++ header file list
• ARROW-952 - Compilation error on macOS with clang-802.0.42
• ARROW-958 - [Python] Conda build guide still needs ARROW_HOME, PARQUET_HOME
• ARROW-979 - [Python] Fix setuptools_scm version when release tag is not in the master timeline
• ARROW-991 - [Python] PyArray_SimpleNew should not be used with NPY_DATETIME
• ARROW-995 - [Website] 0.3 release announce has a typo in reference
• ARROW-998 - [Doc] File format documents incorrect schema location

Improvement

• ARROW-1000 - [GLib] Move install document to Website
• ARROW-1001 - [GLib] Unify writer files
• ARROW-1002 - [C++] It is not necessary to add padding after the magic header in the FileWriter implementation
• ARROW-1010 - [Website] Only show English posts in /blog/
• ARROW-1016 - Python: Include C++ headers (optionally) in wheels
• ARROW-1022 - [Python] Add nthreads option to Feather read method
• ARROW-1024 - Python: Update build time numpy version to 1.10.1
• ARROW-1025 - [Website] Improve changelog on website
• ARROW-1027 - [Python] Allow negative indexing in fields/columns on pyarrow Table and Schema objects
• ARROW-1028 - [Python] Documentation updates after ARROW-1008
• ARROW-1029 - [Python] Fix --with-parquet build on Windows, add unit tests to Appveyor
• ARROW-1030 - Python: Account for library versioning in parquet-cpp
• ARROW-1037 - [GLib] Follow reader name change
• ARROW-1038 - [GLib] Follow writer name change
• ARROW-1040 - [GLib] Follow tensor IO
• ARROW-182 - [C++] Remove Array::Validate virtual function and make a separate method
• ARROW-376 - Python: Convert non-range Pandas indices (optionally) to Arrow
• ARROW-532 - [Python] Expand pyarrow.parquet documentation for 0.3 release
• ARROW-579 - Python: Provide redistributable pyarrow wheels on OSX
• ARROW-891 - [Python] Expand Windows build instructions to not require looking at separate C++ docs
• ARROW-899 - [Docs] Add CHANGELOG for 0.3.0
• ARROW-901 - [Python] Write FixedSizeBinary to Parquet
• ARROW-913 - [Python] Only link jemalloc to the Cython extension where it's needed
• ARROW-923 - [Docs] Generate Changelog for website with JIRA links
• ARROW-929 - Move KEYS file to SVN, remove from git
• ARROW-943 - [GLib] Support running unit tests with source archive
• ARROW-945 - [GLib] Add a Lua example to show Torch integration
• ARROW-946 - [GLib] Use "new" instead of "open" for constructor name
• ARROW-947 - [Python] Improve execution time of manylinux1 build
• ARROW-953 - Use cmake / curl from conda-forge in CI builds
• ARROW-954 - Make it possible to compile Arrow with header-only boost
• ARROW-961 - [Python] Rename InMemoryOutputStream to BufferOutputStream
• ARROW-970 - [Python] Accidentally calling pyarrow.Table() should not segfault process
• ARROW-982 - [Website] Improve website front copy to highlight serialization efficiency benefits
• ARROW-984 - [GLib] Add Go examples
• ARROW-985 - [GLib] Update package information
• ARROW-988 - [JS] Add entry to Travis CI matrix
• ARROW-993 - [GLib] Add missing error checks in Go examples
• ARROW-996 - [Website] Add 0.3 release announce in Japanese

New Feature

• ARROW-1008 - [C++] Define abstract interface for stream iteration
• ARROW-1011 - [Format] Clarify requirements around buffer padding in validity bitmaps
• ARROW-1014 - 0.4.0 release
• ARROW-1031 - [GLib] Support pretty print
• ARROW-1044 - [GLib] Support Feather
• ARROW-29 - C++: Add re2 as optional 3rd-party toolchain dependency
• ARROW-446 - [Python] Document NativeFile interfaces, HDFS client in Sphinx
• ARROW-482 - [Java] Provide API access to custom_metadata Field attribute in IPC setting
• ARROW-596 - [Python] Add convenience function to convert pandas.DataFrame to pyarrow.Buffer containing a file or stream representation
• ARROW-714 - [C++] Add import_pyarrow C API in the style of NumPy for thirdparty C++ users
• ARROW-819 - [Python] Define public Cython API
• ARROW-872 - [JS] Read streaming format
• ARROW-873 - [JS] Implement fixed width list type
• ARROW-874 - [JS] Read dictionary-encoded vectors
• ARROW-963 - [GLib] Add equal
• ARROW-967 - [GLib] Support initializing array with buffer
• ARROW-977 - [java] Add Timezone aware timestamp vectors

Task

• ARROW-1015 - [Java] Implement schema-level metadata
• ARROW-629 - [JS] Add unit test suite
• ARROW-956 - remove pandas pre-0.20.0 compat
• ARROW-957 - [Doc] Add HDFS and Windows documents to doxygen output
• ARROW-997 - [Java] Implement transfer in FixedSizeListVector

Apache Arrow 0.3.0 (5 May 2017)

Bug

• ARROW-109 - [C++] Investigate recursive data types limit in flatbuffers
• ARROW-208 - Add checkstyle policy to java project
• ARROW-347 - Add method to pass CallBack when creating a transfer pair
• ARROW-413 - DATE type is not specified clearly
• ARROW-431 - [Python] Review GIL release and acquisition in to_pandas conversion
• ARROW-443 - [Python] Support for converting from strided pandas data in Table.from_pandas
• ARROW-451 - [C++] Override DataType::Equals for other types with additional metadata
• ARROW-454 - pojo.Field doesn't implement hashCode()
• ARROW-526 - [Format] Update IPC.md to account for File format changes and Streaming format
• ARROW-565 - [C++] Examine "Field::dictionary" member
• ARROW-570 - Determine Java tools JAR location from project metadata
• ARROW-584 - [C++] Fix compiler warnings exposed with -Wconversion
• ARROW-588 - [C++] Fix compiler warnings on 32-bit platforms
• ARROW-595 - [Python] StreamReader.schema returns None
• ARROW-604 - Python: boxed Field instances are missing the reference to DataType
• ARROW-613 - [JS] Implement random-access file format
• ARROW-617 - Time type is not specified clearly
• ARROW-619 - Python: Fix typos in setup.py args and LD_LIBRARY_PATH
• ARROW-623 - segfault with __repr__ of empty Field
• ARROW-624 - [C++] Restore MakePrimitiveArray function
• ARROW-627 - [C++] Compatibility macros for exported extern template class declarations
• ARROW-628 - [Python] Install nomkl metapackage when building parquet-cpp for faster Travis builds
• ARROW-630 - [C++] IPC unloading for BooleanArray does not account for offset
• ARROW-636 - [C++] Add Boost / other system requirements to C++ README
• ARROW-639 - [C++] Invalid offset in slices
• ARROW-642 - [Java] Remove temporary file in java/tools
• ARROW-644 - Python: Cython should be a setup-only requirement
• ARROW-652 - Remove trailing f in merge script output
• ARROW-654 - [C++] Support timezone metadata in file/stream formats
• ARROW-668 - [Python] Convert nanosecond timestamps to pandas.Timestamp when converting from TimestampValue
• ARROW-671 - [GLib] License file isn't installed
• ARROW-673 - [Java] Support additional Time metadata
• ARROW-677 - [java] Fix checkstyle jcl-over-slf4j conflict issue
• ARROW-678 - [GLib] Fix dependenciesfff
• ARROW-680 - [C++] Multiarch support impacts user-supplied install prefix
• ARROW-682 - Add self-validation checks in integration tests
• ARROW-683 - [C++] Support date32 (DateUnit::DAY) in IPC metadata, rename date to date64
• ARROW-686 - [C++] Account for time metadata changes, add time32 and time64 types
• ARROW-689 - [GLib] Install header files and documents to wrong directories
• ARROW-691 - [Java] Encode dictionary Int type in message format
• ARROW-697 - [Java] Raise appropriate exceptions when encountering large (> INT32_MAX) record batches
• ARROW-699 - [C++] Arrow dynamic libraries are missed on run of unit tests on Windows
• ARROW-702 - Fix BitVector.copyFromSafe to reAllocate instead of returning false
• ARROW-703 - Fix issue where setValueCount(0) doesn’t work in the case that we’ve shipped vectors across the wire
• ARROW-704 - Fix bad import caused by conflicting changes
• ARROW-709 - [C++] Restore type comparator for DecimalType
• ARROW-713 - [C++] Fix linking issue with ipc benchmark
• ARROW-715 - Python: Explicit pandas import makes it a hard requirement
• ARROW-716 - error building arrow/python
• ARROW-720 - [java] arrow should not have a dependency on slf4j bridges in compile
• ARROW-723 - Arrow freezes on write if chunk_size=0
• ARROW-726 - [C++] PyBuffer dtor may segfault if constructor passed an object not exporting buffer protocol
• ARROW-732 - Schema comparison bugs in struct and union types
• ARROW-736 - [Python] Mixed-type object DataFrame columns should not silently coerce to an Arrow type by default
• ARROW-738 - [Python] Fix manylinux1 packaging
• ARROW-739 - Parallel build fails non-deterministically.
• ARROW-740 - FileReader fails for large objects
• ARROW-747 - [C++] Fix spurious warning caused by passing dl to add_dependencies
• ARROW-749 - [Python] Delete incomplete binary files when writing fails
• ARROW-753 - [Python] Unit tests in arrow/python fail to link on some OS X platforms
• ARROW-756 - [C++] Do not pass -fPIC when compiling with MSVC
• ARROW-757 - [C++] MSVC build fails on googletest when using NMake
• ARROW-762 - Kerberos Problem with PyArrow
• ARROW-776 - [GLib] Cast type is wrong
• ARROW-777 - [Java] Resolve getObject behavior per changes / discussion in ARROW-729
• ARROW-778 - Modify merge tool to work on Windows
• ARROW-781 - [Python/C++] Increase reference count for base object?
• ARROW-783 - Integration tests fail for length-0 record batch
• ARROW-787 - [GLib] Fix compilation errors caused by ARROW-758
• ARROW-793 - [GLib] Wrong indent
• ARROW-794 - [C++] Check whether data is contiguous in ipc::WriteTensor
• ARROW-797 - [Python] Add updated pyarrow. public API listing in Sphinx docs
• ARROW-800 - [C++] Boost headers being transitively included in pyarrow
• ARROW-805 - listing empty HDFS directory returns an error instead of returning empty list
• ARROW-809 - C++: Writing sliced record batch to IPC writes the entire array
• ARROW-812 - Pip install pyarrow on mac failed.
• ARROW-817 - [C++] Fix incorrect code comment from ARROW-722
• ARROW-821 - [Python] Extra file _table_api.h generated during Python build process
• ARROW-822 - [Python] StreamWriter fails to open with socket as sink
• ARROW-826 - Compilation error on Mac with -DARROW_PYTHON=on
• ARROW-829 - Python: Parquet: Dictionary encoding is deactivated if column-wise compression was selected
• ARROW-830 - Python: jemalloc is not anymore publicly exposed
• ARROW-839 - [C++] Portable alternative to PyDate_to_ms function
• ARROW-847 - C++: BUILD_BYPRODUCTS not specified anymore for gtest
• ARROW-852 - Python: Also set Arrow Library PATHS when detection was done through pkg-config
• ARROW-853 - [Python] It is no longer necessary to modify the RPATH of the Cython extensions on many environments
• ARROW-858 - Remove dependency on boost regex
• ARROW-866 - [Python] Error from file object destructor
• ARROW-867 - [Python] Miscellaneous pyarrow MSVC fixes
• ARROW-875 - Nullable variable length vector fillEmpties() fills an extra value
• ARROW-879 - compat with pandas 0.20.0
• ARROW-882 - [C++] On Windows statically built lib file overwrites lib file of shared build
• ARROW-886 - VariableLengthVectors don't reAlloc offsets
• ARROW-887 - [format] For backward compatibility, new unit fields must have default values matching previous implied unit
• ARROW-888 - BitVector transfer() does not transfer ownership
• ARROW-895 - Nullable variable length vector lastSet not set correctly
• ARROW-900 - [Python] UnboundLocalError in ParquetDatasetPiece
• ARROW-903 - [GLib] Remove a needless "."
• ARROW-914 - [C++/Python] Fix Decimal ToBytes
• ARROW-922 - Allow Flatbuffers and RapidJSON to be used locally on Windows
• ARROW-928 - Update CMAKE script to detect unsupported msvc compilers versions
• ARROW-933 - [Python] arrow_python bindings have debug print statement
• ARROW-934 - [GLib] Glib sources missing from result of 02-source.sh
• ARROW-936 - Fix release README
• ARROW-938 - Fix Apache Rat errors from source release build

Improvement

• ARROW-316 - Finalize Date type
• ARROW-542 - [Java] Implement dictionaries in stream/file encoding
• ARROW-563 - C++: Support non-standard gcc version strings
• ARROW-566 - Python: Deterministic position of libarrow in manylinux1 wheels
• ARROW-569 - [C++] Set version for .pc
• ARROW-577 - [C++] Refactor StreamWriter and FileWriter to have private implementations
• ARROW-580 - C++: Also provide jemalloc_X targets if only a static or shared version is found
• ARROW-582 - [Java] Add Date/Time Support to JSON File
• ARROW-589 - C++: Use system provided shared jemalloc if static is unavailable
• ARROW-593 - [C++] Rename ReadableFileInterface to RandomAccessFile
• ARROW-612 - [Java] Field toString should show nullable flag status
• ARROW-615 - Move ByteArrayReadableSeekableByteChannel to vector.util package
• ARROW-631 - [GLib] Import C API (C++ API wrapper) based on GLib from https://github.com/kou/arrow-glib
• ARROW-646 - Cache miniconda packages
• ARROW-647 - [C++] Don't require Boost static libraries to support CentOS 7
• ARROW-648 - [C++] Support multiarch on Debian
• ARROW-650 - [GLib] Follow eadableFileInterface -> RnadomAccessFile change
• ARROW-651 - [C++] Set shared library version for .deb packages
• ARROW-655 - Implement DecimalArray
• ARROW-662 - [Format] Factor Flatbuffer schema metadata into a Schema.fbs
• ARROW-664 - Make C++ Arrow serialization deterministic
• ARROW-674 - [Java] Support additional Timestamp timezone metadata
• ARROW-675 - [GLib] Update package metadata
• ARROW-676 - [java] move from MinorType to FieldType in ValueVectors to carry all the relevant type bits
• ARROW-679 - [Format] Change RecordBatch and Field length members from int to long
• ARROW-681 - [C++] Build Arrow on Windows with dynamically linked boost
• ARROW-684 - Python: More informative message when parquet-cpp but not parquet-arrow is available
• ARROW-688 - [C++] Use CMAKE_INSTALL_INCLUDEDIR for consistency
• ARROW-690 - Only send JIRA updates to issues@arrow.apache.org
• ARROW-700 - Add headroom interface for allocator.
• ARROW-706 - [GLib] Add package install document
• ARROW-707 - Python: All none-Pandas column should be converted to NullArray
• ARROW-708 - [C++] Some IPC code simplification, perf analysis
• ARROW-712 - [C++] Implement Array::Accept as inline visitor
• ARROW-719 - [GLib] Support prepared source archive release
• ARROW-724 - Add "How to Contribute" section to README
• ARROW-725 - [Format] Constant length list type
• ARROW-727 - [Python] Write memoryview-compatible objects in NativeFile.write with zero copy
• ARROW-728 - [C++/Python] Add arrow::Table function for removing a column
• ARROW-731 - [C++] Add shared library related versions to .pc
• ARROW-741 - [Python] Add Python 3.6 to Travis CI
• ARROW-743 - [C++] Consolidate unit tests for code in array.h
• ARROW-744 - [GLib] Re-add an assertion to garrow_table_new() test
• ARROW-745 - [C++] Allow use of system cpplint
• ARROW-746 - [GLib] Add garrow_array_get_data_type()
• ARROW-751 - [Python] Rename all Cython extensions to "private" status with leading underscore
• ARROW-752 - [Python] Construct pyarrow.DictionaryArray from boxed pyarrow array objects
• ARROW-754 - [GLib] Add garrow_array_is_null()
• ARROW-755 - [GLib] Add garrow_array_get_value_type()
• ARROW-758 - [C++] Fix compiler warnings on MSVC x64
• ARROW-761 - [Python] Add function to compute the total size of tensor payloads, including metadata and padding
• ARROW-763 - C++: Use python-config to find libpythonX.X.dylib
• ARROW-765 - [Python] Make generic ArrowException subclass value error
• ARROW-769 - [GLib] Support building without installed Arrow C++
• ARROW-770 - [C++] Move clang-tidy/format config files back to C++ source tree
• ARROW-774 - [GLib] Remove needless LICENSE.txt copy
• ARROW-775 - [Java] add simple constructors to value vectors
• ARROW-779 - [C++/Python] Raise exception if old metadata encountered
• ARROW-782 - [C++] Change struct to class for objects that meet the criteria in the Google style guide
• ARROW-788 - Possible nondeterminism in Tensor serialization code
• ARROW-795 - [C++] Combine libarrow/libarrow_io/libarrow_ipc
• ARROW-802 - [GLib] Add read examples
• ARROW-803 - [GLib] Update package repository URL
• ARROW-804 - [GLib] Update build document
• ARROW-806 - [GLib] Support add/remove a column from table
• ARROW-807 - [GLib] Update "Since" tag
• ARROW-808 - [GLib] Remove needless ignore entries
• ARROW-810 - [GLib] Remove io/ipc prefix
• ARROW-811 - [GLib] Add GArrowBuffer
• ARROW-815 - [Java] Allow for expanding underlying buffer size after allocation
• ARROW-816 - [C++] Use conda packages for RapidJSON, Flatbuffers to speed up builds
• ARROW-818 - [Python] Review public pyarrow. API completeness and update docs
• ARROW-820 - [C++] Build dependencies for Parquet library without arrow support
• ARROW-825 - [Python] Generalize pyarrow.from_pylist to accept any object implementing the PySequence protocol
• ARROW-827 - [Python] Variety of Parquet improvements to support Dask integration
• ARROW-828 - [CPP] Document new requirement (libboost-regex-dev) in README.md
• ARROW-832 - [C++] Upgrade thirdparty gtest to 1.8.0
• ARROW-833 - [Python] "Quickstart" build / environment setup guide for Python developers
• ARROW-841 - [Python] Add pyarrow build to Appveyor
• ARROW-844 - [Format] Revise format/README.md to reflect progress reaching a more complete specification
• ARROW-845 - [Python] Sync FindArrow.cmake changes from parquet-cpp
• ARROW-846 - [GLib] Add GArrowTensor, GArrowInt8Tensor and GArrowUInt8Tensor
• ARROW-848 - [Python] Improvements / fixes to conda quickstart guide
• ARROW-849 - [C++] Add optional $ARROW_BUILD_TOOLCHAIN environment variable option for configuring build environment
• ARROW-857 - [Python] Automate publishing Python documentation to arrow-site
• ARROW-860 - [C++] Decide if typed Tensor subclasses are worthwhile
• ARROW-861 - [Python] Move DEVELOPMENT.md to Sphinx docs
• ARROW-862 - [Python] Improve source build instructions in README
• ARROW-863 - [GLib] Use GBytes to implement zero-copy
• ARROW-864 - [GLib] Unify Array files
• ARROW-868 - [GLib] Use GBytes to reduce copy
• ARROW-871 - [GLib] Unify DataType files
• ARROW-876 - [GLib] Unify ArrayBuffer files
• ARROW-877 - [GLib] Add garrow_array_get_null_bitmap()
• ARROW-878 - [GLib] Add garrow_binary_array_get_buffer()
• ARROW-892 - [GLib] Fix GArrowTensor document
• ARROW-893 - Add GLib document to Web site
• ARROW-894 - [GLib] Add GArrowPoolBuffer
• ARROW-896 - [Docs] Add Jekyll plugin for including rendered Jupyter notebooks on website
• ARROW-898 - [C++] Expand metadata support to field level, provide for sharing instances of KeyValueMetadata
• ARROW-904 - [GLib] Simplify error check codes
• ARROW-907 - C++: Convenience construct Table from schema and arrays
• ARROW-908 - [GLib] Unify OutputStream files
• ARROW-910 - [C++] Write 0-length EOS indicator at end of stream
• ARROW-916 - [GLib] Add GArrowBufferOutputStream
• ARROW-917 - [GLib] Add GArrowBufferReader
• ARROW-918 - [GLib] Use GArrowBuffer for read
• ARROW-919 - [GLib] Use "id" to get type enum value from GArrowDataType
• ARROW-920 - [GLib] Add Lua examples
• ARROW-925 - [GLib] Fix GArrowBufferReader test
• ARROW-930 - javadoc generation fails with java 8
• ARROW-931 - [GLib] Reconstruct input stream

New Feature

• ARROW-231 - C++: Add typed Resize to PoolBuffer
• ARROW-281 - [C++] IPC/RPC support on Win32 platforms
• ARROW-341 - [Python] Making libpyarrow available to third parties
• ARROW-452 - [C++/Python] Merge "Feather" file format implementation
• ARROW-459 - [C++] Implement IPC round trip for DictionaryArray, dictionaries shared across record batches
• ARROW-483 - [C++/Python] Provide access to custom_metadata Field attribute in IPC setting
• ARROW-491 - [C++] Add FixedWidthBinary type
• ARROW-493 - [C++] Allow in-memory array over 2^31 -1 elements but require splitting at IPC / RPC boundaries
• ARROW-502 - [C++/Python] Add MemoryPool implementation that logs allocation activity to std::cout
• ARROW-510 - Add integration tests for date and time types
• ARROW-520 - [C++] Add STL-compliant allocator that hooks into an arrow::MemoryPool
• ARROW-528 - [Python] Support _metadata or _common_metadata files when reading Parquet directories
• ARROW-534 - [C++] Add IPC tests for date/time types
• ARROW-539 - [Python] Support reading Parquet datasets with standard partition directory schemes
• ARROW-550 - [Format] Add a TensorMessage type
• ARROW-552 - [Python] Add scalar value support for Dictionary type
• ARROW-557 - [Python] Explicitly opt in to HDFS unit tests
• ARROW-568 - [C++] Add default implementations for TypeVisitor, ArrayVisitor methods that return NotImplemented
• ARROW-574 - Python: Add support for nested Python lists in Pandas conversion
• ARROW-576 - [C++] Complete round trip Union file/stream IPC tests
• ARROW-578 - [C++] Add CMake option to add custom $CXXFLAGS
• ARROW-598 - [Python] Add support for converting pyarrow.Buffer to a memoryview with zero copy
• ARROW-603 - [C++] Add RecordBatch::Validate method that at least checks that schema matches the array metadata
• ARROW-605 - [C++] Refactor generic ArrayLoader class, support work for Feather merge
• ARROW-606 - [C++] Upgrade to flatbuffers 1.6.0
• ARROW-608 - [Format] Days since epoch date type
• ARROW-610 - [C++] Win32 compatibility in file.cc
• ARROW-616 - [C++] Remove -g flag in release builds
• ARROW-618 - [Python] Implement support for DatetimeTZ custom type from pandas
• ARROW-620 - [C++] Add date/time support to JSON reader/writer for integration testing
• ARROW-621 - [C++] Implement an "inline visitor" template that enables visitor-pattern-like code without virtual function dispatch
• ARROW-625 - [C++] Add time unit to TimeType::ToString
• ARROW-626 - [Python] Enable pyarrow.BufferReader to read from any Python object implementing the buffer/memoryview protocol
• ARROW-632 - [Python] Add support for FixedWidthBinary type
• ARROW-635 - [C++] Add JSON read/write support for FixedWidthBinary
• ARROW-637 - [Format] Add time zone metadata to Timestamp type
• ARROW-656 - [C++] Implement IO interface that can read and write to a fixed-size mutable buffer
• ARROW-657 - [Python] Write and read tensors (with zero copy) into shared memory
• ARROW-658 - [C++] Implement in-memory arrow::Tensor objects
• ARROW-659 - [C++] Add multithreaded memcpy implementation (for hardware where it helps)
• ARROW-660 - [C++] Restore function that can read a complete encapsulated record batch message
• ARROW-661 - [C++] Add a Flatbuffer metadata type that supports array data over 2^31 - 1 elements
• ARROW-663 - [Java] Support additional Time metadata + vector value accessors
• ARROW-669 - [Python] Attach proper tzinfo when computing boxed scalars for TimestampArray
• ARROW-687 - [C++] Build and run full test suite in Appveyor
• ARROW-698 - [C++] Add options to StreamWriter/FileWriter to permit large record batches
• ARROW-701 - [Java] Support additional Date metadata
• ARROW-710 - [Python] Enable Feather APIs to read and write using Python file-like objects
• ARROW-717 - [C++] IPC zero-copy round trips for arrow::Tensor
• ARROW-718 - [Python] Expose arrow::Tensor with conversions to/from NumPy arrays
• ARROW-722 - [Python] pandas conversions for new date and time types/metadata
• ARROW-729 - [Java] Add vector type for 32-bit date as days since UNIX epoch
• ARROW-733 - [C++/Format] Change name of Fixed Width Binary to Fixed Size Binary for consistency
• ARROW-734 - [Python] Support for pyarrow on Windows / MSVC
• ARROW-735 - [C++] Developer instruction document for MSVC on Windows
• ARROW-737 - [C++] Support obtaining mutable slices of mutable buffers
• ARROW-768 - [Java] Change the "boxed" object representation of date and time types
• ARROW-771 - [Python] Add APIs for reading individual Parquet row groups
• ARROW-773 - [C++] Add function to create arrow::Table with column appended to existing table
• ARROW-865 - [Python] Verify Parquet roundtrips for new date/time types
• ARROW-880 - [GLib] Add garrow_primitive_array_get_buffer()
• ARROW-890 - [GLib] Add GArrowMutableBuffer
• ARROW-926 - Update KEYS to include wesm

Task

• ARROW-52 - Set up project blog
• ARROW-670 - Arrow 0.3 release
• ARROW-672 - [Format] Bump metadata version for 0.3 release
• ARROW-748 - [Python] Pin runtime library versions in conda-forge packages to force upgrades
• ARROW-798 - [Docs] Publish Format Markdown documents somehow on arrow.apache.org
• ARROW-869 - [JS] Rename directory to js/
• ARROW-95 - Scaffold Main Documentation using asciidoc
• ARROW-98 - Java: API documentation

Test

• ARROW-836 - Test for timedelta compat with pandas
• ARROW-927 - C++/Python: Add manylinux1 builds to Travis matrix

Apache Arrow 0.2.0 (15 February 2017)

Bug

• ARROW-112 - [C++] Style fix for constants/enums
• ARROW-202 - [C++] Integrate with appveyor ci for windows support and get arrow building on windows
• ARROW-220 - [C++] Build conda artifacts in a build environment with better cross-linux ABI compatibility
• ARROW-224 - [C++] Address static linking of boost dependencies
• ARROW-230 - Python: Do not name modules like native ones (i.e. rename pyarrow.io)
• ARROW-239 - [Python] HdfsFile.read called with no arguments should read remainder of file
• ARROW-261 - [C++] Refactor BinaryArray/StringArray classes to not inherit from ListArray
• ARROW-275 - Add tests for UnionVector in Arrow File
• ARROW-294 - [C++] Do not use fopen / fclose / etc. methods for memory mapped file implementation
• ARROW-322 - [C++] Do not build HDFS IO interface optionally
• ARROW-323 - [Python] Opt-in to PyArrow parquet build rather than skipping silently on failure
• ARROW-334 - [Python] OS X rpath issues on some configurations
• ARROW-337 - UnionListWriter.list() is doing more than it should, this can cause data corruption
• ARROW-339 - Make merge_arrow_pr script work with Python 3
• ARROW-340 - [C++] Opening a writeable file on disk that already exists does not truncate to zero
• ARROW-342 - Set Python version on release
• ARROW-345 - libhdfs integration doesn't work for Mac
• ARROW-346 - Python API Documentation
• ARROW-348 - [Python] CMake build type should be configurable on the command line
• ARROW-349 - Six is missing as a requirement in the python setup.py
• ARROW-351 - Time type has no unit
• ARROW-354 - Connot compare an array of empty strings to another
• ARROW-357 - Default Parquet chunk_size of 64k is too small
• ARROW-358 - [C++] libhdfs can be in non-standard locations in some Hadoop distributions
• ARROW-362 - Python: Calling to_pandas on a table read from Parquet leaks memory
• ARROW-371 - Python: Table with null timestamp becomes float in pandas
• ARROW-375 - columns parameter in parquet.read_table() raises KeyError for valid column
• ARROW-384 - Align Java and C++ RecordBatch data and metadata layout
• ARROW-386 - [Java] Respect case of struct / map field names
• ARROW-387 - [C++] arrow::io::BufferReader does not permit shared memory ownership in zero-copy reads
• ARROW-390 - C++: CMake fails on json-integration-test with ARROW_BUILD_TESTS=OFF
• ARROW-392 - Fix string/binary integration tests
• ARROW-393 - [JAVA] JSON file reader fails to set the buffer size on String data vector
• ARROW-395 - Arrow file format writes record batches in reverse order.
• ARROW-398 - [Java] Java file format requires bitmaps of all 1's to be written when there are no nulls
• ARROW-399 - [Java] ListVector.loadFieldBuffers ignores the ArrowFieldNode length metadata
• ARROW-400 - [Java] ArrowWriter writes length 0 for Struct types
• ARROW-401 - [Java] Floating point vectors should do an approximate comparison in integration tests
• ARROW-402 - [Java] "refCnt gone negative" error in integration tests
• ARROW-403 - [JAVA] UnionVector: Creating a transfer pair doesn't transfer the schema to destination vector
• ARROW-404 - [Python] Closing an HdfsClient while there are still open file handles results in a crash
• ARROW-405 - [C++] Be less stringent about finding include/hdfs.h in HADOOP_HOME
• ARROW-406 - [C++] Large HDFS reads must utilize the set file buffer size when making RPCs
• ARROW-408 - [C++/Python] Remove defunct conda recipes
• ARROW-414 - [Java] "Buffer too large to resize to ..." error
• ARROW-420 - Align Date implementation between Java and C++
• ARROW-421 - [Python] Zero-copy buffers read by pyarrow::PyBytesReader must retain a reference to the parent PyBytes to avoid premature garbage collection issues
• ARROW-422 - C++: IPC should depend on rapidjson_ep if RapidJSON is vendored
• ARROW-429 - git-archive SHA-256 checksums are changing
• ARROW-433 - [Python] Date conversion is locale-dependent
• ARROW-434 - Segfaults and encoding issues in Python Parquet reads
• ARROW-435 - C++: Spelling mistake in if(RAPIDJSON_VENDORED)
• ARROW-437 - [C++] clang compiler warnings from overridden virtual functions
• ARROW-445 - C++: arrow_ipc is built before arrow/ipc/Message_generated.h was generated
• ARROW-447 - Python: Align scalar/pylist string encoding with pandas' one.
• ARROW-455 - [C++] BufferOutputStream dtor does not call Close()
• ARROW-469 - C++: Add option so that resize doesn't decrease the capacity
• ARROW-481 - [Python] Fix Python 2.7 regression in patch for PARQUET-472
• ARROW-486 - [C++] arrow::io::MemoryMappedFile can't be casted to arrow::io::FileInterface
• ARROW-487 - Python: ConvertTableToPandas segfaults if ObjectBlock::Write fails
• ARROW-494 - [C++] When MemoryMappedFile is destructed, memory is unmapped even if buffer referecnes still exist
• ARROW-499 - Update file serialization to use streaming serialization format
• ARROW-505 - [C++] Fix compiler warnings in release mode
• ARROW-511 - [Python] List[T] conversions not implemented for single arrays
• ARROW-513 - [C++] Fix Appveyor build
• ARROW-519 - [C++] Missing vtable in libarrow.dylib on Xcode 6.4
• ARROW-523 - Python: Account for changes in PARQUET-834
• ARROW-533 - [C++] arrow::TimestampArray / TimeArray has a broken constructor
• ARROW-535 - [Python] Add type mapping for NPY_LONGLONG
• ARROW-537 - [C++] StringArray/BinaryArray comparisons may be incorrect when values with non-zero length are null
• ARROW-540 - [C++] Fix build in aftermath of ARROW-33
• ARROW-543 - C++: Lazily computed null_counts counts number of non-null entries
• ARROW-544 - [C++] ArrayLoader::LoadBinary fails for length-0 arrays
• ARROW-545 - [Python] Ignore files without .parq or .parquet prefix when reading directory of files
• ARROW-548 - [Python] Add nthreads option to pyarrow.Filesystem.read_parquet
• ARROW-551 - C++: Construction of Column with nullptr Array segfaults
• ARROW-556 - [Integration] Can not run Integration tests if different cpp build path
• ARROW-561 - Update java & python dependencies to improve downstream packaging experience

Improvement

• ARROW-189 - C++: Use ExternalProject to build thirdparty dependencies
• ARROW-191 - Python: Provide infrastructure for manylinux1 wheels
• ARROW-328 - [C++] Return shared_ptr by value instead of const-ref?
• ARROW-330 - [C++] CMake functions to simplify shared / static library configuration
• ARROW-333 - Make writers update their internal schema even when no data is written.
• ARROW-335 - Improve Type apis and toString() by encapsulating flatbuffers better
• ARROW-336 - Run Apache Rat in Travis builds
• ARROW-338 - [C++] Refactor IPC vector "loading" and "unloading" to be based on cleaner visitor pattern
• ARROW-350 - Add Kerberos support to HDFS shim
• ARROW-355 - Add tests for serialising arrays of empty strings to Parquet
• ARROW-356 - Add documentation about reading Parquet
• ARROW-360 - C++: Add method to shrink PoolBuffer using realloc
• ARROW-361 - Python: Support reading a column-selection from Parquet files
• ARROW-365 - Python: Provide Array.to_pandas()
• ARROW-366 - [java] implement Dictionary vector
• ARROW-374 - Python: clarify unicode vs. binary in API
• ARROW-379 - Python: Use setuptools_scm/setuptools_scm_git_archive to provide the version number
• ARROW-380 - [Java] optimize null count when serializing vectors.
• ARROW-382 - Python: Extend API documentation
• ARROW-396 - Python: Add pyarrow.schema.Schema.equals
• ARROW-409 - Python: Change pyarrow.Table.dataframe_from_batches API to create Table instead
• ARROW-411 - [Java] Move Intergration.compare and Intergration.compareSchemas to a public utils class
• ARROW-423 - C++: Define BUILD_BYPRODUCTS in external project to support non-make CMake generators
• ARROW-425 - Python: Expose a C function to convert arrow::Table to pyarrow.Table
• ARROW-426 - Python: Conversion from pyarrow.Array to a Python list
• ARROW-430 - Python: Better version handling
• ARROW-432 - [Python] Avoid unnecessary memory copy in to_pandas conversion by using low-level pandas internals APIs
• ARROW-450 - Python: Fixes for PARQUET-818
• ARROW-457 - Python: Better control over memory pool
• ARROW-458 - Python: Expose jemalloc MemoryPool
• ARROW-463 - C++: Support jemalloc 4.x
• ARROW-466 - C++: ExternalProject for jemalloc
• ARROW-468 - Python: Conversion of nested data in pd.DataFrames to/from Arrow structures
• ARROW-474 - Create an Arrow streaming file fomat
• ARROW-479 - Python: Test for expected schema in Pandas conversion
• ARROW-485 - [Java] Users are required to initialize VariableLengthVectors.offsetVector before calling VariableLengthVectors.mutator.getSafe
• ARROW-490 - Python: Update manylinux1 build scripts
• ARROW-524 - [java] provide apis to access nested vectors and buffers
• ARROW-525 - Python: Add more documentation to the package
• ARROW-529 - Python: Add jemalloc and Python 3.6 to manylinux1 build
• ARROW-546 - Python: Account for changes in PARQUET-867
• ARROW-553 - C++: Faster valid bitmap building

New Feature

• ARROW-108 - [C++] Add IPC round trip for union types
• ARROW-221 - Add switch for writing Parquet 1.0 compatible logical types
• ARROW-227 - [C++/Python] Hook arrow_io generic reader / writer interface into arrow_parquet
• ARROW-228 - [Python] Create an Arrow-cpp-compatible interface for reading bytes from Python file-like objects
• ARROW-243 - [C++] Add "driver" option to HdfsClient to choose between libhdfs and libhdfs3 at runtime
• ARROW-303 - [C++] Also build static libraries for leaf libraries
• ARROW-312 - [Python] Provide Python API to read/write the Arrow IPC file format
• ARROW-317 - [C++] Implement zero-copy Slice method on arrow::Buffer that retains reference to parent
• ARROW-33 - C++: Implement zero-copy array slicing
• ARROW-332 - [Python] Add helper function to convert RecordBatch to pandas.DataFrame
• ARROW-363 - Set up Java/C++ integration test harness
• ARROW-369 - [Python] Add ability to convert multiple record batches at once to pandas
• ARROW-373 - [C++] Implement C++ version of JSON file format for testing
• ARROW-377 - Python: Add support for conversion of Pandas.Categorical
• ARROW-381 - [C++] Simplify primitive array type builders to use a default type singleton
• ARROW-383 - [C++] Implement C++ version of ARROW-367 integration test validator
• ARROW-389 - Python: Write Parquet files to pyarrow.io.NativeFile objects
• ARROW-394 - Add integration tests for boolean, list, struct, and other basic types
• ARROW-410 - [C++] Add Flush method to arrow::io::OutputStream
• ARROW-415 - C++: Add Equals implementation to compare Tables
• ARROW-416 - C++: Add Equals implementation to compare Columns
• ARROW-417 - C++: Add Equals implementation to compare ChunkedArrays
• ARROW-418 - [C++] Consolidate array container and builder code, remove arrow/types
• ARROW-419 - [C++] Promote util/{status.h, buffer.h, memory-pool.h} to top level of arrow/ source directory
• ARROW-427 - [C++] Implement dictionary-encoded array container
• ARROW-428 - [Python] Deserialize from Arrow record batches to pandas in parallel using a thread pool
• ARROW-438 - [Python] Concatenate Table instances with equal schemas
• ARROW-440 - [C++] Support pkg-config
• ARROW-441 - [Python] Expose Arrow's file and memory map classes as NativeFile subclasses
• ARROW-442 - [Python] Add public Python API to inspect Parquet file metadata
• ARROW-444 - [Python] Avoid unnecessary memory copies from use of PyBytes_* C APIs
• ARROW-449 - Python: Conversion from pyarrow.{Table,RecordBatch} to a Python dict
• ARROW-456 - C++: Add jemalloc based MemoryPool
• ARROW-461 - [Python] Implement conversion between arrow::DictionaryArray and pandas.Categorical
• ARROW-467 - [Python] Run parquet-cpp unit tests in Travis CI
• ARROW-470 - [Python] Add "FileSystem" abstraction to access directories of files in a uniform way
• ARROW-471 - [Python] Enable ParquetFile to pass down separately-obtained file metadata
• ARROW-472 - [Python] Expose parquet::{SchemaDescriptor, ColumnDescriptor}::Equals
• ARROW-475 - [Python] High level support for reading directories of Parquet files (as a single Arrow table) from supported file system interfaces
• ARROW-476 - [Integration] Add integration tests for Binary / Varbytes type
• ARROW-477 - [Java] Add support for second/microsecond/nanosecond timestamps in-memory and in IPC/JSON layer
• ARROW-478 - [Python] Accept a PyBytes object in the pyarrow.io.BufferReader ctor
• ARROW-484 - Add more detail about what of technology can be found in the Arrow implementations to README
• ARROW-495 - [C++] Add C++ implementation of streaming serialized format
• ARROW-497 - [Java] Integration test harness for streaming format
• ARROW-498 - [C++] Integration test harness for streaming format
• ARROW-503 - [Python] Interface to streaming binary format
• ARROW-508 - [C++] Make file/memory-mapped file interfaces threadsafe
• ARROW-509 - [Python] Add support for PARQUET-835 (parallel column reads)
• ARROW-512 - C++: Add method to check for primitive types
• ARROW-514 - [Python] Accept pyarrow.io.Buffer as input to StreamReader, FileReader classes
• ARROW-515 - [Python] Add StreamReader/FileReader methods that read all record batches as a Table
• ARROW-521 - [C++/Python] Track peak memory use in default MemoryPool
• ARROW-531 - Python: Document jemalloc, extend Pandas section, add Getting Involved
• ARROW-538 - [C++] Set up AddressSanitizer (ASAN) builds
• ARROW-547 - [Python] Expose Array::Slice and RecordBatch::Slice
• ARROW-81 - [Format] Add a Category logical type (distinct from dictionary-encoding)

Task

• ARROW-268 - [C++] Flesh out union implementation to have all required methods for IPC
• ARROW-327 - [Python] Remove conda builds from Travis CI processes
• ARROW-353 - Arrow release 0.2
• ARROW-359 - Need to document ARROW_LIBHDFS_DIR
• ARROW-367 - [java] converter csv/json <=> Arrow file format for Integration tests
• ARROW-368 - Document use of LD_LIBRARY_PATH when using Python
• ARROW-372 - Create JSON arrow file format for integration tests
• ARROW-506 - Implement Arrow Echo server for integration testing
• ARROW-527 - clean drill-module.conf file
• ARROW-558 - Add KEYS files
• ARROW-96 - C++: API documentation using Doxygen
• ARROW-97 - Python: API documentation via sphinx-apidoc

Apache Arrow 0.1.0 (7 October 2016)

Bug

• ARROW-103 - Missing patterns from .gitignore
• ARROW-104 - Update Layout.md based on discussion on the mailing list
• ARROW-105 - Unit tests fail if assertions are disabled
• ARROW-113 - TestValueVector test fails if cannot allocate 2GB of memory
• ARROW-16 - Building cpp issues on XCode 7.2.1
• ARROW-17 - Set some vector fields to default access level for Drill compatibility
• ARROW-18 - Fix bug with decimal precision and scale
• ARROW-185 - [C++] Make sure alignment and memory padding conform to spec
• ARROW-188 - Python: Add numpy as install requirement
• ARROW-193 - For the instruction, typos "int his" should be "in this"
• ARROW-194 - C++: Allow read-only memory mapped source
• ARROW-200 - [Python] Convert Values String looks like it has incorrect error handling
• ARROW-209 - [C++] Broken builds: llvm.org apt repos are unavailable
• ARROW-210 - [C++] Tidy up the type system a little bit
• ARROW-211 - Several typos/errors in Layout.md examples
• ARROW-217 - Fix Travis w.r.t conda 4.1.0 changes
• ARROW-219 - [C++] Passed CMAKE_CXX_FLAGS are being dropped, fix compiler warnings
• ARROW-223 - Do not link against libpython
• ARROW-225 - [C++/Python] master Travis CI build is broken
• ARROW-244 - [C++] Some global APIs of IPC module should be visible to the outside
• ARROW-246 - [Java] UnionVector doesn't call allocateNew() when creating it's vectorType
• ARROW-247 - [C++] Missing explicit destructor in RowBatchReader causes an incomplete type error
• ARROW-250 - Fix for ARROW-246 may cause memory leaks
• ARROW-259 - Use flatbuffer fields in java implementation
• ARROW-265 - Negative decimal values have wrong padding
• ARROW-266 - [C++] Fix the broken build
• ARROW-274 - Make the MapVector nullable
• ARROW-278 - [Format] Struct type name consistency in implementations and metadata
• ARROW-283 - [C++] Update arrow_parquet to account for API changes in PARQUET-573
• ARROW-284 - [C++] Triage builds by disabling Arrow-Parquet module
• ARROW-287 - [java] Make nullable vectors use a BitVecor instead of UInt1Vector for bits
• ARROW-297 - Fix Arrow pom for release
• ARROW-304 - NullableMapReaderImpl.isSet() always returns true
• ARROW-308 - UnionListWriter.setPosition() should not call startList()
• ARROW-309 - Types.getMinorTypeForArrowType() does not work for Union type
• ARROW-313 - XCode 8.0 breaks builds
• ARROW-314 - JSONScalar is unnecessary and unused.
• ARROW-320 - ComplexCopier.copy(FieldReader, FieldWriter) should not start a list if reader is not set
• ARROW-321 - Fix Arrow licences
• ARROW-36 - Remove fixVersions from patch tool (until we have them)
• ARROW-46 - Port DRILL-4410 to Arrow
• ARROW-5 - Error when run maven install
• ARROW-51 - Move ValueVector test from Drill project
• ARROW-55 - Python: fix legacy Python (2.7) tests and add to Travis CI
• ARROW-62 - Format: Are the nulls bits 0 or 1 for null values?
• ARROW-63 - C++: ctest fails if Python 3 is the active Python interpreter
• ARROW-65 - Python: FindPythonLibsNew does not work in a virtualenv
• ARROW-69 - Change permissions for assignable users
• ARROW-72 - FindParquet searches for non-existent header
• ARROW-75 - C++: Fix handling of empty strings
• ARROW-77 - C++: conform null bit interpretation to match ARROW-62
• ARROW-80 - Segmentation fault on len(Array) for empty arrays
• ARROW-88 - C++: Refactor given PARQUET-572
• ARROW-93 - XCode 7.3 breaks builds
• ARROW-94 - Expand list example to clarify null vs empty list

Improvement

• ARROW-10 - Fix mismatch of javadoc names and method parameters
• ARROW-15 - Fix a naming typo for memory.AllocationManager.AllocationOutcome
• ARROW-190 - Python: Provide installable sdist builds
• ARROW-199 - [C++] Refine third party dependency
• ARROW-206 - [C++] Expose an equality API for arrays that compares a range of slots on two arrays
• ARROW-212 - [C++] Clarify the fact that PrimitiveArray is now abstract class
• ARROW-213 - Exposing static arrow build
• ARROW-218 - Add option to use GitHub API token via environment variable when merging PRs
• ARROW-234 - [C++] Build with libhdfs support in arrow_io in conda builds
• ARROW-238 - C++: InternalMemoryPool::Free() should throw an error when there is insufficient allocated memory
• ARROW-245 - [Format] Clarify Arrow's relationship with big endian platforms
• ARROW-252 - Add implementation guidelines to the documentation
• ARROW-253 - Int types should only have width of 8*2^n (8, 16, 32, 64)
• ARROW-254 - Remove Bit type as it is redundant with boolean
• ARROW-255 - Finalize Dictionary representation
• ARROW-256 - Add versioning to the arrow spec.
• ARROW-257 - Add a typeids Vector to Union type
• ARROW-264 - Create an Arrow File format
• ARROW-270 - [Format] Define more generic Interval logical type
• ARROW-271 - Update Field structure to be more explicit
• ARROW-279 - rename vector module to arrow-vector for consistency
• ARROW-280 - [C++] Consolidate file and shared memory IO interfaces
• ARROW-285 - Allow for custom flatc compiler
• ARROW-286 - Build thirdparty dependencies in parallel
• ARROW-289 - Install test-util.h
• ARROW-290 - Specialize alloc() in ArrowBuf
• ARROW-292 - [Java] Upgrade Netty to 4.041
• ARROW-299 - Use absolute namespace in macros
• ARROW-305 - Add compression and use_dictionary options to Parquet interface
• ARROW-306 - Add option to pass cmake arguments via environment variable
• ARROW-315 - Finalize timestamp type
• ARROW-319 - Add canonical Arrow Schema json representation
• ARROW-324 - Update arrow metadata diagram
• ARROW-325 - make TestArrowFile not dependent on timezone
• ARROW-50 - C++: Enable library builds for 3rd-party users without having to build thirdparty googletest
• ARROW-54 - Python: rename package to "pyarrow"
• ARROW-64 - Add zsh support to C++ build scripts
• ARROW-66 - Maybe some missing steps in installation guide
• ARROW-68 - Update setup_build_env and third-party script to be more userfriendly
• ARROW-71 - C++: Add script to run clang-tidy on codebase
• ARROW-73 - Support CMake 2.8
• ARROW-78 - C++: Add constructor for DecimalType
• ARROW-79 - Python: Add benchmarks
• ARROW-8 - Set up Travis CI
• ARROW-85 - C++: memcmp can be avoided in Equal when comparing with the same Buffer
• ARROW-86 - Python: Implement zero-copy Arrow-to-Pandas conversion
• ARROW-87 - Implement Decimal schema conversion for all ways supported in Parquet
• ARROW-89 - Python: Add benchmarks for Arrow<->Pandas conversion
• ARROW-9 - Rename some unchanged "Drill" to "Arrow"
• ARROW-91 - C++: First draft of an adapter class for parquet-cpp's ParquetFileReader that produces Arrow table/row batch objects

New Feature

• ARROW-100 - [C++] Computing RowBatch size
• ARROW-106 - Add IPC round trip for string types (string, char, varchar, binary)
• ARROW-107 - [C++] add ipc round trip for struct types
• ARROW-13 - Add PR merge tool similar to that used in Parquet
• ARROW-19 - C++: Externalize memory allocations and add a MemoryPool abstract interface to builder classes
• ARROW-197 - [Python] Add conda dev recipe for pyarrow
• ARROW-2 - Post Simple Website
• ARROW-20 - C++: Add null count member to Array containers, remove nullable member
• ARROW-201 - C++: Initial ParquetWriter implementation
• ARROW-203 - Python: Basic filename based Parquet read/write
• ARROW-204 - [Python] Automate uploading conda build artifacts for libarrow and pyarrow
• ARROW-21 - C++: Add in-memory schema metadata container
• ARROW-214 - C++: Add String support to Parquet I/O
• ARROW-215 - C++: Support other integer types in Parquet I/O
• ARROW-22 - C++: Add schema adapter routines for converting flat Parquet schemas to in-memory Arrow schemas
• ARROW-222 - [C++] Create prototype file-like interface to HDFS (via libhdfs) and begin defining more general IO interface for Arrow data adapters
• ARROW-23 - C++: Add logical "Column" container for chunked data
• ARROW-233 - [C++] Add visibility defines for limiting shared library symbol visibility
• ARROW-236 - [Python] Enable Parquet read/write to work with HDFS file objects
• ARROW-237 - [C++] Create Arrow specializations of Parquet allocator and read interfaces
• ARROW-24 - C++: Add logical "Table" container
• ARROW-242 - C++/Python: Support Timestamp Data Type
• ARROW-26 - C++: Add developer instructions for building parquet-cpp integration
• ARROW-262 - [Format] Add a new format document for metadata and logical types for messaging and IPC / on-wire/file representations
• ARROW-267 - [C++] C++ implementation of file-like layout for RPC / IPC
• ARROW-28 - C++: Add google/benchmark to the 3rd-party build toolchain
• ARROW-293 - [C++] Implementations of IO interfaces for operating system files
• ARROW-296 - [C++] Remove arrow_parquet C++ module and related parts of build system
• ARROW-3 - Post Initial Arrow Format Spec
• ARROW-30 - Python: pandas/NumPy to/from Arrow conversion routines
• ARROW-301 - [Format] Add some form of user field metadata to IPC schemas
• ARROW-302 - [Python] Add support to use the Arrow file format with file-like objects
• ARROW-31 - Python: basic PyList <-> Arrow marshaling code
• ARROW-318 - [Python] Revise README to reflect current state of project
• ARROW-37 - C++: Represent boolean array data in bit-packed form
• ARROW-4 - Initial Arrow CPP Implementation
• ARROW-42 - Python: Add to Travis CI build
• ARROW-43 - Python: Add rudimentary console __repr__ for array types
• ARROW-44 - Python: Implement basic object model for scalar values (i.e. results of arrow_arr[i])
• ARROW-48 - Python: Add Schema object wrapper
• ARROW-49 - Python: Add Column and Table wrapper interface
• ARROW-53 - Python: Fix RPATH and add source installation instructions
• ARROW-56 - Format: Specify LSB bit ordering in bit arrays
• ARROW-57 - Format: Draft data headers IDL for data interchange
• ARROW-58 - Format: Draft type metadata ("schemas") IDL
• ARROW-59 - Python: Boolean data support for builtin data structures
• ARROW-60 - C++: Struct type builder API
• ARROW-67 - C++: Draft type metadata conversion to/from IPC representation
• ARROW-7 - Add Python library build toolchain
• ARROW-70 - C++: Add "lite" DCHECK macros used in parquet-cpp
• ARROW-76 - Revise format document to include null count, defer non-nullable arrays to the domain of metadata
• ARROW-82 - C++: Implement IPC exchange for List types
• ARROW-90 - Apache Arrow cpp code does not support power architecture
• ARROW-92 - C++: Arrow to Parquet Schema conversion

Task

• ARROW-1 - Import Initial Codebase
• ARROW-101 - Fix java warnings emitted by java compiler
• ARROW-102 - travis-ci support for java project
• ARROW-11 - Mirror JIRA activity to dev@arrow.apache.org
• ARROW-14 - Add JIRA components
• ARROW-251 - [C++] Expose APIs for getting code and message of the status
• ARROW-272 - Arrow release 0.1
• ARROW-298 - create release scripts
• ARROW-35 - Add a short call-to-action / how-to-get-involved to the main README.md

Test

• ARROW-260 - TestValueVector.testFixedVectorReallocation and testVariableVectorReallocation are flaky
• ARROW-83 - Add basic test infrastructure for DecimalType