---

layout: default title: Apache Arrow 0.15.0 Release permalink: /release/0.15.0.html

Apache Arrow 0.15.0 (5 October 2019)

This is a major release covering more than 3 months of development.

Download

• Source Artifacts
• Binary Artifacts
  • For CentOS
  • For Debian
  • For Python
  • For Ubuntu
• Git tag

Contributors

This release includes 672 commits from 80 distinct contributors.

$ git shortlog -sn apache-arrow-0.14.0..apache-arrow-0.15.0
    96	Wes McKinney
    63	Antoine Pitrou
    59	tianchen
    55	Sutou Kouhei
    46	liyafan82
    38	Neal Richardson
    34	Joris Van den Bossche
    29	Krisztián Szűcs
    24	Andy Grove
    20	Benjamin Kietzman
    18	Prudhvi Porandla
    17	Micah Kornfield
    15	François Saint-Jacques
    13	David Li
    12	Yosuke Shiro
     9	Pindikura Ravindra
     8	Romain Francois
     7	Omer Ozarslan
     7	Praveen
     6	Renjie Liu
     5	ptaylor
     5	Kenta Murata
     5	Hatem Helal
     5	Bryan Cutler
     4	Marco Neumann
     4	Uwe L. Korn
     4	Eric Erhardt
     3	ARF1
     3	Chao Sun
     3	Paddy Horan
     2	James Lamb
     2	andyscho
     2	Ryan Murray
     2	Martin Radev
     2	Sebastien Binet
     1	Zhuo Peng
     1	b-rms
     1	czxrrr
     1	emkornfield
     1	lihalite
     1	mmaclach
     1	psuman
     1	roshie548
     1	shengjun.li
     1	tianchen92
     1	Ádám Lippai
     1	Aaron Opfer
     1	Adam Lippai
     1	Artem Alekseev
     1	Chen Li
     1	Eric Liang
     1	Galuh Sahid
     1	Hengruo Zhang
     1	Ingo Mueller
     1	Ingvar-Y
     1	Itamar Turner-Trauring
     1	Jeka Pats
     1	Johan Peltenburg
     1	Kenneth Jung
     1	Liya Fan
     1	Marcin Juszkiewicz
     1	Marius Seritan
     1	Mark Harris
     1	Mark Mikofski
     1	Neville Dipale
     1	Paul Taylor
     1	Philipp Moritz
     1	Richard Liaw
     1	Rok
     1	Ruslan Kuprieiev
     1	TP Boudreau
     1	Takuya Kato
     1	Tao He
     1	Thomas Elvey
     1	Tobias Mayer
     1	Ulzii Otgonbaatar
     1	Yuan Zhou
     1	Yuqi Gu
     1	Zeyuan Shang
     1	Zherui Cao

Patch Committers

The following Apache committers merged contributed patches to the repository.

$ git shortlog -csn apache-arrow-0.14.0..apache-arrow-0.15.0
   214	Wes McKinney
    85	Sutou Kouhei
    82	Micah Kornfield
    70	Antoine Pitrou
    44	Pindikura Ravindra
    32	Krisztián Szűcs
    29	François Saint-Jacques
    25	Neal Richardson
    19	Andy Grove
    12	Yosuke Shiro
    10	Benjamin Kietzman
    10	Bryan Cutler
    10	Paddy Horan
     9	Praveen
     6	Neville Dipale
     4	Uwe L. Korn
     3	Philipp Moritz
     3	GitHub
     1	Romain Francois
     1	ptaylor
     1	Chao Sun
     1	emkornfield
     1	Kenta Murata

Changelog

New Features and Improvements

• ARROW-1324 - [C++] Support ARROW_BOOST_VENDORED on Windows / MSVC
• ARROW-1561 - [C++] Kernel implementations for “isin” (set containment)
• ARROW-1566 - [C++] Implement non-materializing sort kernels
• ARROW-1741 - [C++] Comparison function for DictionaryArray to determine if indices are “compatible”
• ARROW-1789 - [Format] Consolidate specification documents and improve clarity for new implementation authors
• ARROW-1875 - [Java] Write 64-bit ints as strings in integration test JSON files
• ARROW-2769 - [C++][Python] Deprecate and rename add_metadata methods
• ARROW-2931 - [Crossbow] Windows builds are attempting to run linux and osx packaging tasks
• ARROW-3032 - [Python] Clean up NumPy-related C++ headers
• ARROW-3204 - [R] Enable package to be made available on CRAN
• ARROW-3243 - [C++] Upgrade jemalloc to version 5
• ARROW-3246 - [Python][Parquet] direct reading/writing of pandas categoricals in parquet
• ARROW-3325 - [Python] Support reading Parquet binary/string columns directly as DictionaryArray
• ARROW-3531 - [Python] Deprecate Schema.field_by_name in favor of __getitem__
• ARROW-3538 - [Python] ability to override the automated assignment of uuid for filenames when writing datasets
• ARROW-3579 - [Crossbow] Unintuitive error message when remote branch has not been pushed
• ARROW-3643 - [Rust] Optimize `push_slice` of `BufferBuilder`
• ARROW-3710 - [Crossbow][Python] Run nightly tests against pandas master
• ARROW-3772 - [C++] Read Parquet dictionary encoded ColumnChunks directly into an Arrow DictionaryArray
• ARROW-3777 - [C++] Implement a mock “high latency” filesystem
• ARROW-3817 - [R] $ method for RecordBatch
• ARROW-3829 - [Python] Support protocols to extract Arrow objects from third-party classes
• ARROW-3943 - [R] Write vignette for R package
• ARROW-4036 - [C++] Make status codes pluggable
• ARROW-4095 - [C++] Implement optimizations for dictionary unification where dictionaries are prefixes of the unified dictionary
• ARROW-4111 - [Python] Create time types from Python sequences of integers
• ARROW-4218 - [Rust] [Parquet] Implement ColumnReader
• ARROW-4220 - [Python] Add buffered input and output stream ASV benchmarks with simulated high latency IO
• ARROW-4365 - [Rust] [Parquet] Implement RecordReader
• ARROW-4398 - [Python] Add benchmarks for Arrow<>Parquet BYTE_ARRAY serialization (read and write)
• ARROW-4473 - [Website] Add instructions to do a test-deploy of Arrow website and fix bugs
• ARROW-4507 - [Format] Create outline and introduction for new document.
• ARROW-4508 - [Format] Copy content from Layout.rst to new document.
• ARROW-4509 - [Format] Copy content from Metadata.rst to new document.
• ARROW-4510 - [Format] copy content from IPC.rst to new document.
• ARROW-4511 - [Format] remove individual documents in favor of new document once all content is moved
• ARROW-453 - [C++] Add filesystem implementation for Amazon S3
• ARROW-4648 - [C++/Question] Naming/organizational inconsistencies in cpp codebase
• ARROW-4649 - [C++/CI/R] Add (nightly) job that builds `brew install apache-arrow --HEAD`
• ARROW-4752 - [Rust] Add explicit SIMD vectorization for the divide kernel
• ARROW-4810 - [Format][C++] Add “LargeList” type with 64-bit offsets
• ARROW-4841 - [C++] Persist CMake options in generated CMake config
• ARROW-5134 - [R][CI] Run nightly tests against multiple R versions
• ARROW-517 - [C++] Verbose Array::Equals
• ARROW-5211 - [Format] Missing documentation under `Dictionary encoding` section on MetaData page
• ARROW-5216 - [CI] Add Appveyor badge to README
• ARROW-5307 - [CI][GLib] Enable GTK-Doc
• ARROW-5343 - [C++] Consider using Buffer for transpose maps in DictionaryType::Unify instead of std::vector
• ARROW-5344 - [C++] Use ArrayDataVisitor in implementation of dictionary unpacking in compute/kernels/cast.cc
• ARROW-5351 - [Rust] Add support for take kernel functions
• ARROW-5358 - [Rust] Implement equality check for ArrayData and Array
• ARROW-5380 - [C++] Fix and enable UBSan for unaligned accesses.
• ARROW-5439 - [Java] Utilize stream EOS in File format
• ARROW-5444 - [Release][Website] After 0.14 release, update what is an “official” release
• ARROW-5458 - [C++] ARMv8 parallel CRC32c computation optimization
• ARROW-5480 - [Python] Pandas categorical type doesn't survive a round-trip through parquet
• ARROW-5483 - [Java] add ValueVector constructors that take a Field object
• ARROW-5494 - [Python] Create FileSystem bindings
• ARROW-5505 - [R] Stop masking base R functions/rethink namespacing
• ARROW-5527 - [C++] HashTable/MemoTable should use Buffer(s)/Builder(s) for heap data
• ARROW-5558 - [C++] Support Array::View on arrays with non-zero offsets
• ARROW-5559 - [C++] Introduce IpcOptions struct object for better API-stability when adding new options
• ARROW-5564 - [C++] Add uriparser to conda-forge
• ARROW-5579 - [Java] shade flatbuffer dependency
• ARROW-5580 - [C++][Gandiva] Correct definitions of timestamp functions in Gandiva
• ARROW-5588 - [C++] Better support for building UnionArrays
• ARROW-5594 - [C++] add support for UnionArrays to Take and Filter
• ARROW-5610 - [Python] Define extension type API in Python to “receive” or “send” a foreign extension type
• ARROW-5646 - [Crossbow][Documentation] Move the user guide to the Sphinx documentation
• ARROW-5681 - [FlightRPC] Wrap gRPC exceptions/statuses
• ARROW-5686 - [R] Review R Windows CI build
• ARROW-5716 - [Developer] Improve merge PR script to acknowledge co-authors
• ARROW-5717 - [Python] Support dictionary unification when converting variable dictionaries to pandas
• ARROW-5719 - [Java] Support in-place vector sorting
• ARROW-5722 - [Rust] Implement std::fmt::Debug for ListArray, BinaryArray and StructArray
• ARROW-5734 - [Python] Dispatch to Table.from_arrays from pyarrow.table factory function
• ARROW-5736 - [Format][C++] Support small bit-width indices in sparse tensor
• ARROW-5741 - [JS] Make numeric vector from functions consistent with TypedArray.from
• ARROW-5743 - [C++] Add CMake option to enable “large memory” unit tests
• ARROW-5746 - [Website] Move website source out of apache/arrow
• ARROW-5747 - [C++] Better column name and header support in CSV reader
• ARROW-5758 - [C++][Gandiva] Support casting decimals to varchar and vice versa
• ARROW-5762 - [Integration][JS] Integration Tests for Map Type
• ARROW-5777 - [C++] BasicDecimal128 is a small object it doesn't always make sense to pass by const ref
• ARROW-5778 - [Java] Extract the logic for vector data copying to the super classes
• ARROW-5784 - [Release][GLib] Replace c_glib/ after running c_glib/autogen.sh in dev/release/02-source.sh
• ARROW-5786 - [Release] Use arrow-jni profile in dev/release/01-prepare.sh
• ARROW-5788 - [Rust] Use { version = “...”, path = “../...” } for arrow and parquet dependencies
• ARROW-5789 - [C++] Small Warning/Linkage cleanups
• ARROW-5792 - [Rust] [Parquet] A visitor trait for parquet types.
• ARROW-5798 - [Packaging][deb] Update doc architecture
• ARROW-5800 - [R] Dockerize R Travis CI tests so they can be run anywhere via docker-compose
• ARROW-5803 - [C++] Dockerize C++ with clang 7 Travis CI unit test logic
• ARROW-5812 - [Java] Refactor method name and param type in BaseIntVector
• ARROW-5813 - [C++] Support checking the equality of the different contiguous tensors
• ARROW-5814 - [Java] Implement a <Object, int> HashMap for DictionaryEncoder
• ARROW-5827 - [C++] Require c-ares CMake config
• ARROW-5828 - [C++] Add Protocol Buffers version check
• ARROW-5830 - [C++] Stop using memcmp in TensorEquals
• ARROW-5832 - [Java] Support search operations for vector data
• ARROW-5833 - [C++] Factor out status copying code from cast.cc
• ARROW-5834 - [Java] Apply new hash map in DictionaryEncoder
• ARROW-5835 - [Java] Support Dictionary Encoding for binary type
• ARROW-5841 - [Website] Add 0.14.0 release note
• ARROW-5842 - [Java] Revise the semantic of lastSet in ListVector
• ARROW-5843 - [Java] Improve the readability and performance of BitVectorHelper#getNullCount
• ARROW-5844 - [Java] Support comparison & sort for more numeric types
• ARROW-5846 - [Java] Create Avro adapter module and add dependencies
• ARROW-5853 - [Python] Expose boolean filter kernel on Array
• ARROW-5861 - [Java] Initial implement to convert Avro record with primitive types
• ARROW-5862 - [Java] Provide dictionary builder
• ARROW-5864 - [Python] simplify cython wrapping of Result
• ARROW-5865 - [Release] Helper script for rebasing open pull requests on master
• ARROW-5866 - [C++] Remove duplicate library in cpp/Brewfile
• ARROW-5867 - [C++][Gandiva] Add support for cast int to decimal
• ARROW-5872 - Support mod(double, double) method in Gandiva
• ARROW-5876 - [FlightRPC] Implement basic auth across all languages
• ARROW-5877 - [FlightRPC] Fix auth incompatibilities between Python/Java
• ARROW-5880 - [C++] Update arrow parquet writer to use TypedBufferBuilder
• ARROW-5881 - [Java] Provide functionalities to efficiently determine if a validity buffer has completely 1 bits/0 bits
• ARROW-5883 - [Java] Support dictionary encoding for List and Struct type
• ARROW-5888 - [Python][C++] Add metadata to store Arrow time zones in Parquet file metadata
• ARROW-5891 - [C++][Gandiva] Remove duplicates in function registries
• ARROW-5892 - [C++][Gandiva] Support function aliases
• ARROW-5893 - [C++] Remove arrow::Column class from C++ library
• ARROW-5897 - [Java] Remove duplicated logic in MapVector
• ARROW-5898 - [Java] Provide functionality to efficiently compute hash code for arbitrary memory segment
• ARROW-5900 - [Gandiva] [Java] Decimal precision,scale bounds check
• ARROW-5901 - [Rust] Implement PartialEq to compare array and json values
• ARROW-5902 - [Java] Implement hash table and equals & hashCode API for dictionary encoding
• ARROW-5903 - [Java] Set methods in DecimalVector are slow
• ARROW-5904 - [Java] [Plasma] Fix compilation of Plasma Java client
• ARROW-5906 - [CI] Set -DARROW_VERBOSE_THIRDPARTY_BUILD=OFF in builds running in Travis CI, maybe all docker-compose builds by default
• ARROW-5908 - [C#] ArrowStreamWriter doesn't align buffers to 8 bytes
• ARROW-5909 - [Java] Optimize ByteFunctionHelpers equals & compare logic
• ARROW-5911 - [Java] Make ListVector and MapVector create reader lazily
• ARROW-5917 - [Java] Redesign the dictionary encoder
• ARROW-5918 - [Java] Add get to BaseIntVector interface
• ARROW-5919 - [R] Add nightly tests for building r-arrow with dependencies from conda-forge
• ARROW-5920 - [Java] Support sort & compare for all variable width vectors
• ARROW-5924 - [C++][Plasma] It is not convenient to release a GPU object
• ARROW-5934 - [Python] Bundle arrow's LICENSE with the wheels
• ARROW-5937 - [Release] Stop parallel binary upload
• ARROW-5938 - [Release] Create branch for adding release note automatically
• ARROW-5939 - [Release] Add support for generating vote email template separately
• ARROW-5940 - [Release] Add support for re-uploading sign/checksum for binary artifacts
• ARROW-5941 - [Release] Avoid re-uploading already uploaded binary artifacts
• ARROW-5943 - [GLib][Gandiva] Add support for function aliases
• ARROW-5944 - [C++][Gandiva] Remove ‘div’ alias for ‘divide’
• ARROW-5945 - [Rust] [DataFusion] Table trait should support building complete queries
• ARROW-5947 - [Rust] [DataFusion] Remove serde_json dependency
• ARROW-5948 - [Rust] [DataFusion] create_logical_plan should not call optimizer
• ARROW-5955 - [Plasma] Support setting memory quotas per plasma client for better isolation
• ARROW-5957 - [C++][Gandiva] Implement div function in Gandiva
• ARROW-5958 - [Python] Link zlib statically in the wheels
• ARROW-5961 - [R] Be able to run R-only tests even without C++ library
• ARROW-5962 - [CI][Python] Do not test manylinux1 wheels in Travis CI
• ARROW-5967 - [Java] DateUtility#timeZoneList is not correct
• ARROW-5970 - [Java] Provide pointer to Arrow buffer
• ARROW-5974 - [Python][C++] Enable CSV reader to read from concatenated gzip stream
• ARROW-5975 - [C++][Gandiva] Add method to cast Date(in Milliseconds) to timestamp
• ARROW-5976 - [C++] RETURN_IF_ERROR(ctx) should be namespaced
• ARROW-5977 - [C++] [Python] Method for read_csv to limit which columns are read?
• ARROW-5979 - [FlightRPC] Expose (de)serialization of protocol types
• ARROW-5985 - [Developer] Do not suggest setting Fix Version for point releases in dev/merge_arrow_pr.py
• ARROW-5986 - [Java] Code cleanup for dictionary encoding
• ARROW-5988 - [Java] Avro adapter implement simple Record type
• ARROW-5997 - [Java] Support dictionary encoding for Union type
• ARROW-5998 - [Java] Open a document to track the API changes
• ARROW-6000 - [Python] Expose LargeBinaryType and LargeStringType
• ARROW-6008 - [Release] Don't parallelize the bintray upload script
• ARROW-6009 - [Release][JS] Ignore NPM errors in the javascript release script
• ARROW-6013 - [Java] Support range searcher
• ARROW-6017 - [FlightRPC] Allow creating Locations with unknown schemes
• ARROW-6020 - [Java] Refactor ByteFunctionHelper#hash with new added ArrowBufHasher
• ARROW-6021 - [Java] Extract copyFrom and copyFromSafe methods to ValueVector interface
• ARROW-6022 - [Java] Support equals API in ValueVector to compare two vectors equal
• ARROW-6023 - [C++][Gandiva] Add functions in Gandiva
• ARROW-6024 - [Java] Provide more hash algorithms
• ARROW-6026 - [Doc] Add CONTRIBUTING.md
• ARROW-6030 - [Java] Efficiently compute hash code for ArrowBufPointer
• ARROW-6031 - [Java] Support iterating a vector by ArrowBufPointer
• ARROW-6034 - [C++][Gandiva] Add string functions in Gandiva
• ARROW-6035 - [Java] Avro adapter support convert nullable value
• ARROW-6036 - [GLib] Add support for skip rows and column_names CSV read option
• ARROW-6037 - [GLib] Add a missing version macro
• ARROW-6039 - [GLib] Add garrow_array_filter()
• ARROW-6041 - [Website] Blog post announcing R package release
• ARROW-6042 - [C++] Implement alternative DictionaryBuilder that always yields int32 indices
• ARROW-6045 - [C++] Benchmark for Parquet float and NaN encoding/decoding
• ARROW-6048 - [C++] Add ChunkedArray::View which calls to Array::View
• ARROW-6049 - [C++] Support using Array::View from compatible dictionary type to another
• ARROW-6053 - [Python] RecordBatchStreamReader::Open2 cdef type signature doesn't match C++
• ARROW-6063 - [FlightRPC] Implement “half-closed” semantics for DoPut
• ARROW-6065 - [C++] Reorganize parquet/arrow/reader.cc, remove code duplication, improve readability
• ARROW-6069 - [Rust] [Parquet] Implement Converter to convert record reader to arrow primitive array.
• ARROW-6070 - [Java] Avoid creating new schema before IPC sending
• ARROW-6077 - [C++][Parquet] Build logical schema tree mapping Arrow fields to Parquet schema levels
• ARROW-6078 - [Java] Implement dictionary-encoded subfields for List type
• ARROW-6079 - [Java] Implement/test UnionFixedSizeListWriter for FixedSizeListVector
• ARROW-6080 - [Java] Support compare and search operation for BaseRepeatedValueVector
• ARROW-6083 - [Java] Refactor Jdbc adapter consume logic
• ARROW-6084 - [Python] Support LargeList
• ARROW-6085 - [Rust] [DataFusion] Create traits for phsyical query plan
• ARROW-6086 - [Rust] [DataFusion] Implement parallel execution for parquet scan
• ARROW-6087 - [Rust] [DataFusion] Implement parallel execution for CSV scan
• ARROW-6088 - [Rust] [DataFusion] Implement parallel execution for projection
• ARROW-6089 - [Rust] [DataFusion] Implement parallel execution for selection
• ARROW-6090 - [Rust] [DataFusion] Implement parallel execution for hash aggregate
• ARROW-6093 - [Java] reduce branches in algo for first match in VectorRangeSearcher
• ARROW-6094 - [Format][Flight] Add GetFlightSchema to Flight RPC
• ARROW-6096 - [C++] Conditionally depend on boost regex library
• ARROW-6097 - [Java] Avro adapter implement unions type
• ARROW-6100 - [Rust] Pin to specific Rust nightly release
• ARROW-6101 - [Rust] [DataFusion] Create physical plan from logical plan
• ARROW-6104 - [Rust] [DataFusion] Don't allow bare_trait_objects
• ARROW-6105 - [C++][Parquet][Python] Add test case showing dictionary-encoded subfields in nested type
• ARROW-6113 - [Java] Support vector deduplicate function
• ARROW-6115 - [Python] support LargeList, LargeString, LargeBinary in conversion to pandas
• ARROW-6118 - [Java] Replace google Preconditions with Arrow Preconditions
• ARROW-6121 - [Tools] Improve merge tool cli ergonomic
• ARROW-6125 - [Python] Remove any APIs deprecated prior to 0.14.x
• ARROW-6127 - [Website] Add favicons and meta tags
• ARROW-6128 - [C++] Can't build with g++ 8.3.0 by class-memaccess warning
• ARROW-6130 - [Release] Use 0.15.0 as the next release
• ARROW-6134 - [C++][Gandiva] Add concat function in Gandiva
• ARROW-6137 - [C++][Gandiva] Change output format of castVARCHAR(timestamp) in Gandiva
• ARROW-6138 - [C++] Add a basic (single RecordBatch) implementation of Dataset
• ARROW-6139 - [Documentation][R] Build R docs (pkgdown) site and add to arrow-site
• ARROW-6141 - [C++] Enable memory-mapping a file region that is offset from the beginning of the file
• ARROW-6142 - [R] Install instructions on linux could be clearer
• ARROW-6143 - [Java] Unify the copyFrom and copyFromSafe methods for all vectors
• ARROW-6144 - [C++][Gandiva] Implement random function in Gandiva
• ARROW-6155 - [Java] Extract a super interface for vectors whose elements reside in continuous memory segments
• ARROW-6156 - [Java] Support compare semantics for ArrowBufPointer
• ARROW-6161 - [C++] Implements dataset::ParquetFile and associated Scan structures
• ARROW-6162 - [C++][Gandiva] Do not truncate string in castVARCHAR_varchar when out_len parameter is zero
• ARROW-6172 - [Java] Provide benchmarks to set IntVector with different methods
• ARROW-6177 - [C++] Add Array::Validate()
• ARROW-6180 - [C++] Create InputStream that is an isolated reader of a segment of a RandomAccessFile
• ARROW-6181 - [R] Only allow R package to install without libarrow on linux
• ARROW-6183 - [R] Document that you don‘t have to use tidyselect if you don’t want
• ARROW-6185 - [Java] Provide hash table based dictionary builder
• ARROW-6187 - [C++] fallback to storage type when writing ExtensionType to Parquet
• ARROW-6188 - [GLib] Add garrow_array_is_in()
• ARROW-6192 - [GLib] Use the same SO version as C++
• ARROW-6194 - [Java] Add non-static approach in DictionaryEncoder making it easy to extend and reuse
• ARROW-6196 - [Ruby] Add support for building Arrow::TimeNNArray by .new
• ARROW-6197 - [GLib] Add garrow_decimal128_rescale()
• ARROW-6199 - [Java] Avro adapter avoid potential resource leak.
• ARROW-6203 - [GLib] Add garrow_array_sort_to_indices()
• ARROW-6204 - [GLib] Add garrow_array_is_in_chunked_array()
• ARROW-6206 - [Java][Docs] Document environment variables/java properties
• ARROW-6209 - [Java] Extract set null method to the base class for fixed width vectors
• ARROW-6212 - [Java] Support vector rank operation
• ARROW-6216 - [C++] Allow user to select the compression level
• ARROW-6217 - [Website] Remove needless _site/ directory
• ARROW-6219 - [Java] Add API for JDBC adapter that can convert less then the full result set at a time.
• ARROW-6220 - [Java] Add API to avro adapter to limit number of rows returned at a time.
• ARROW-6225 - [Website] Update arrow-site/README and any other places to point website contributors in right direction
• ARROW-6229 - [C++] Add a DataSource implementation which scans a directory
• ARROW-6230 - [R] Reading in Parquet files are 20x slower than reading fst files in R
• ARROW-6231 - [C++][Python] Consider assigning default column names when reading CSV file and header_rows=0
• ARROW-6232 - [C++] Rename Argsort kernel to SortToIndices
• ARROW-6237 - [R] Add option to set CXXFLAGS when compiling R package with $ARROW_R_CXXFLAGS
• ARROW-6238 - [C++] Implement SimpleDataSource/SimpleDataFragment
• ARROW-6240 - [Ruby] Arrow::Decimal128Array returns BigDecimal
• ARROW-6242 - [C++] Implements basic Dataset/Scanner/ScannerBuilder
• ARROW-6243 - [C++] Implement basic Filter expression classes
• ARROW-6244 - [C++] Implement Partition DataSource
• ARROW-6246 - [Website] Add link to R documentation site
• ARROW-6247 - [Java] Provide a common interface for float4 and float8 vectors
• ARROW-6249 - [Java] Remove useless class ByteArrayWrapper
• ARROW-6250 - [Java] Implement ApproxEqualsVisitor comparing approx for floating point
• ARROW-6252 - [Python] Add pyarrow.Array.diff method that exposes arrow::Diff
• ARROW-6253 - [Python] Expose “enable_buffered_stream” option from parquet::ReaderProperties in pyarrow.parquet.read_table
• ARROW-6258 - [R] Add macOS build scripts
• ARROW-6260 - [Website] Use deploy key on Travis to build and push to asf-site
• ARROW-6262 - [Developer] Show JIRA issue before merging
• ARROW-6264 - [Java] There is no need to consider byte order in ArrowBufHasher
• ARROW-6265 - [Java] Avro adapter implement Array/Map/Fixed type
• ARROW-6267 - [Ruby] Add Arrow::Time for Arrow::Time{32,64}DataType value
• ARROW-6271 - [Rust] [DataFusion] Add example for running SQL against Parquet
• ARROW-6272 - [Rust] [DataFusion] Add register_parquet convenience method to ExecutionContext
• ARROW-6278 - [R] Read parquet files from raw vector
• ARROW-6279 - [Python] Add Table.slice method or allow slices in __getitem__
• ARROW-6284 - [C++] Allow references in std::tuple when converting tuple to arrow array
• ARROW-6287 - [Rust] [DataFusion] Refactor TableProvider to return thread-safe BatchIterator
• ARROW-6288 - [Java] Implement TypeEqualsVisitor comparing vector type equals considering names and metadata
• ARROW-6289 - [Java] Add empty() in UnionVector to create instance
• ARROW-6292 - [C++] Add an option to build with mimalloc
• ARROW-6294 - [C++] Use hyphen for plasma-store-server executable
• ARROW-6296 - [Java] Cleanup JDBC interfaces and eliminate one memcopy for binary/varchar fields
• ARROW-6297 - [Java] Compare ArrowBufPointers by unsinged integers
• ARROW-6300 - [C++] Add io::OutputStream::Abort()
• ARROW-6303 - [Rust] Add a feature to disable SIMD
• ARROW-6304 - [Java] Add description to each maven artifact
• ARROW-6306 - [Java] Support stable sort by stable comparators
• ARROW-6310 - [C++] Write 64-bit integers as strings in JSON integration test files
• ARROW-6311 - [Java] Make ApproxEqualsVisitor accept DiffFunction to make it more flexible
• ARROW-6313 - [Format] Tracking for ensuring flatbuffer serialized values are aligned in stream/files.
• ARROW-6314 - [C++] Implement changes to ensure flatbuffer alignment.
• ARROW-6315 - [Java] Make change to ensure flatbuffer reads are aligned
• ARROW-6316 - [Go] Make change to ensure flatbuffer reads are aligned
• ARROW-6317 - [JS] Implement changes to ensure flatbuffer alignment
• ARROW-6318 - [Integration] Update integration test to use generated binaries to ensure backwards compatibility
• ARROW-6319 - [C++] Extract the core of NumericTensor::Value as Tensor::Value
• ARROW-6326 - [C++] Nullable fields when converting std::tuple to Table
• ARROW-6328 - Click.option-s should have help text
• ARROW-6329 - [Format] Add 4-byte “stream continuation” to IPC message format to align Flatbuffers
• ARROW-6331 - [Java] Incorporate ErrorProne into the java build
• ARROW-6334 - [Java] Improve the dictionary builder API to return the position of the value in the dictionary
• ARROW-6335 - [Java] Improve the performance of DictionaryHashTable
• ARROW-6336 - [Python] Clarify pyarrow.serialize/deserialize docstrings viz-a-viz relationship with Arrow IPC protocol
• ARROW-6337 - [R] as_tibble in R API is a misnomer
• ARROW-6338 - [R] Type function names don't match type names
• ARROW-6342 - [Python] Add pyarrow.record_batch factory function with same basic API / semantics as pyarrow.table
• ARROW-6346 - [GLib] Add garrow_array_view()
• ARROW-6347 - [GLib] Add garrow_array_diff_unified()
• ARROW-6350 - [Ruby] Remove Arrow::Struct and use Hash instead
• ARROW-6351 - [Ruby] Improve Arrow#values performance
• ARROW-6353 - [Python] Allow user to select compression level in pyarrow.parquet.write_table
• ARROW-6355 - [Java] Make range equal visitor reusable
• ARROW-6356 - [Java] Avro adapter implement Enum type and nested Record type
• ARROW-6357 - [C++] S3: allow for background writes
• ARROW-6358 - [C++] FileSystem::DeleteDir should make it optional to delete the directory itself
• ARROW-6360 - [R] Update support for compression
• ARROW-6362 - [C++] S3: more flexible credential options
• ARROW-6365 - [R] Should be able to coerce numeric to integer with schema
• ARROW-6366 - [Java] Make field vectors final explicitly
• ARROW-6368 - [C++] Add RecordBatch projection functionality
• ARROW-6373 - [C++] Make FixedWidthBinaryBuilder consistent with other primitive fixed width builders
• ARROW-6375 - [C++] Extend ConversionTraits to allow efficiently appending list values in STL API
• ARROW-6379 - [C++] Do not append any buffers when serializing NullType for IPC
• ARROW-6381 - [C++] BufferOutputStream::Write is slow for many small writes
• ARROW-6383 - [Java] report outstanding child allocators on parent allocator close
• ARROW-6384 - [C++] Bump dependencies
• ARROW-6385 - [C++] Investigate xxh3
• ARROW-6391 - [Python][Flight] Add built-in methods on FlightServerBase to start server and wait for it to be available
• ARROW-6397 - [C++][CI] Fix S3 minio failure
• ARROW-6401 - [Java] Implement dictionary-encoded subfields for Struct type
• ARROW-6402 - [C++] Suppress sign-compare warning with g++ 9.2.1
• ARROW-6403 - [Python] Expose FileReader::ReadRowGroups() to Python
• ARROW-6408 - [Rust] Use “if cfg!” pattern in SIMD kernel implementations
• ARROW-6413 - [R] Support autogenerating column names
• ARROW-6415 - [R] Remove usage of R CMD config CXXCPP
• ARROW-6416 - [Python] Confusing API & documentation regarding chunksizes
• ARROW-6419 - [Website] Blog post about Parquet dictionary performance work coming in 0.15.x release
• ARROW-6422 - [Gandiva] Fix double-conversion linker issue
• ARROW-6426 - [FlightRPC] Expose gRPC configuration knobs in Flight
• ARROW-6427 - [GLib] Add support for column names autogeneration CSV read option
• ARROW-6438 - [R] Add bindings for filesystem API
• ARROW-6447 - [C++] Builds with ARROW_JEMALLOC=ON wait until jemalloc_ep is complete before building any libarrow .cc files
• ARROW-6450 - [C++] Use 2x reallocation strategy in arrow::BufferBuilder instead of 1.5x
• ARROW-6451 - [Format] Add clarifications to Columnar.rst about the contents of “null” slots in Varbinary or List arrays
• ARROW-6453 - [C++] More informative error messages from S3
• ARROW-6454 - [Developer] Add LLVM license to LICENSE.txt due to binary redistribution in packages
• ARROW-6458 - [Java] Remove value boxing/unboxing for ApproxEqualsVisitor
• ARROW-6460 - [Java] Add benchmark and large fake data UT for avro adapter
• ARROW-6462 - [C++] Can't build with bundled double-conversion on CentOS 6 x86_64
• ARROW-6465 - [Python] Improve Windows build instructions
• ARROW-6474 - [Python] Provide mechanism for python to write out old format
• ARROW-6475 - [C++] Don't try to dictionary encode dictionary arrays
• ARROW-6477 - [Packaging][Crossbow] Use Azure Pipelines to build linux packages
• ARROW-6480 - [Developer] Add command to generate and send e-mail report for a Crossbow run
• ARROW-6484 - [Java] Enable create indexType for DictionaryEncoding according to dictionary value count
• ARROW-6487 - [Rust] [DataFusion] Create test utils module
• ARROW-6489 - [Developer][Documentation] Fix merge script and readme
• ARROW-6490 - [Java] log error for leak in allocator close
• ARROW-6491 - [Java] fix master build failure caused by ErrorProne
• ARROW-6494 - [C++][Dataset] Implement basic PartitionScheme
• ARROW-6504 - [Python][Packaging] Add mimalloc to conda packages for better performance
• ARROW-6505 - [Website] Add new committers
• ARROW-6518 - [Packaging][Python] Flight failing in OSX Python wheel builds
• ARROW-6519 - [Java] Use IPC continuation token to mark EOS
• ARROW-6524 - [Developer][Packaging] Nightly build report's subject should contain Arrow
• ARROW-6525 - [C++] CloseFromDestructor() should perhaps not crash
• ARROW-6526 - [C++] Poison data in PoolBuffer destructor
• ARROW-6527 - [C++] Add OutputStream::Write() variant taking an owned buffer
• ARROW-6531 - [Python] Add detach() method to buffered streams
• ARROW-6532 - [R] Write parquet files with compression
• ARROW-6533 - [R] Compression codec should take a “level”
• ARROW-6534 - [Java] Fix typos and spelling
• ARROW-6539 - [R] Provide mechanism to write out old format
• ARROW-6540 - [R] Add Validate() methods
• ARROW-6541 - [Format][C++] Use two-part EOS and amend Format documentation
• ARROW-6542 - [R] Add View() method to array types
• ARROW-6544 - [R] Documentation/polishing for 0.15 release
• ARROW-6545 - [Go] Update Go IPC writer to use two-part EOS per mailing list discussion
• ARROW-6546 - [C++] Add missing FlatBuffers source dependency
• ARROW-6549 - [C++] Switch back to latest jemalloc 5.x
• ARROW-6556 - [Python] Prepare for pandas release without SparseDataFrame
• ARROW-6557 - [Python] Always return pandas.Series from Array/ChunkedArray.to_pandas, propagate field names to Series from RecordBatch, Table
• ARROW-6558 - [C++] Refactor Iterator to a type erased handle
• ARROW-6559 - [Developer][C++] Add “archery” option to specify system toolchain for C++ builds
• ARROW-6563 - [Rust] [DataFusion] Create “merge” execution plan
• ARROW-6569 - [Website] Add support for auto deployment by GitHub Actions
• ARROW-6570 - [Python] Use MemoryPool to allocate memory for NumPy arrays in to_pandas calls
• ARROW-6580 - [Java] Support comparison for unsigned integers
• ARROW-6584 - [Python][Wheel] Bundle zlib again with the windows wheels
• ARROW-6588 - [C++] Suppress class-memaccess warning with g++ 9.2.1
• ARROW-6589 - [C++] Support BinaryType in MakeArrayOfNull
• ARROW-6590 - [C++] Do not require ARROW_JSON=ON when ARROW_IPC=ON
• ARROW-6591 - [R] Ignore .Rhistory files in source control
• ARROW-6599 - [Rust] [DataFusion] Implement SUM aggregate expression
• ARROW-6601 - [Java] Improve JDBC adapter performance & add benchmark
• ARROW-6605 - [C++] Add recursion depth control to fs::Selector
• ARROW-6606 - [C++] Construct tree structure from std::vectorfs::FileStats
• ARROW-6609 - [C++] Add minimal build Dockerfile example
• ARROW-6610 - [C++] Add ARROW_FILESYSTEM=ON/OFF CMake configuration flag
• ARROW-6613 - [C++] Remove dependency on boost::filesystem
• ARROW-6614 - [C++][Dataset] Implement FileSystemDataSourceDiscovery
• ARROW-6621 - [Rust][DataFusion] Examples for DataFusion are not executed in CI
• ARROW-6629 - [Doc][C++] Document the FileSystem API
• ARROW-6630 - [Doc][C++] Document the file readers (CSV, JSON, Parquet, etc.)
• ARROW-6644 - [JS] Amend NullType IPC protocol to append no buffers
• ARROW-6647 - [C++] Can't build with g++ 4.8.5 on CentOS 7 by member initializer for shared_ptr
• ARROW-6648 - [Go] Expose the bitutil package
• ARROW-6649 - [R] print() methods for Table, RecordBatch, etc.
• ARROW-6653 - [Developer] Add support for auto JIRA link on pull request
• ARROW-6655 - [Python] Filesystem bindings for S3
• ARROW-6664 - [C++] Add option to build without SSE4.2
• ARROW-6665 - [Rust] [DataFusion] Implement numeric literal expressions
• ARROW-6667 - [Python] Avoid Reference Cycles in pyarrow.parquet
• ARROW-6668 - [Rust] [DataFusion] Implement CAST expression
• ARROW-6669 - [Rust] [DataFusion] Implement physical expression for binary expressions
• ARROW-6675 - [JS] Add scanReverse function to dataFrame and filteredDataframe
• ARROW-6683 - [Python] Add unit tests that validate cross-compatibility with pyarrow.parquet when fastparquet is installed
• ARROW-6725 - [CI] Disable 3rdparty fuzzit nightly builds
• ARROW-6735 - [C++] Suppress sign-compare warning with g++ 9.2.1
• ARROW-6752 - [Go] implement Stringer for Null array
• ARROW-6755 - [Release] Improvements to Windows release verification script
• ARROW-6771 - [Packaging][Python] Missing pytest dependency from conda and wheel builds
• ARROW-750 - [Format] Add LargeBinary and LargeString types

Bug Fixes

• ARROW-1184 - [Java] Dictionary.equals is not working correctly
• ARROW-2317 - [Python] fix C linkage warning
• ARROW-2490 - [C++] input stream locking inconsistent
• ARROW-3176 - [Python] Overflow in Date32 column conversion to pandas
• ARROW-3203 - [C++] Build error on Debian Buster
• ARROW-3651 - [Python] Datetimes from non-DateTimeIndex cannot be deserialized
• ARROW-3652 - [Python] CategoricalIndex is lost after reading back
• ARROW-3762 - [C++] Parquet arrow::Table reads error when overflowing capacity of BinaryArray
• ARROW-3933 - [Python] Segfault reading Parquet files from GNOMAD
• ARROW-4187 - [C++] file-benchmark uses <poll.h>
• ARROW-4746 - [C++/Python] PyDataTime_Date wrongly casted to PyDataTime_DateTime
• ARROW-4836 - [Python] “Cannot tell() a compressed stream” when using RecordBatchStreamWriter
• ARROW-4848 - [C++] Static libparquet not compiled with -DARROW_STATIC on Windows
• ARROW-4880 - [Python] python/asv-build.sh is probably broken after CMake refactor
• ARROW-4883 - [Python] read_csv() returns garbage if given file object in text mode
• ARROW-5028 - [Python][C++] Creating list with pyarrow.array can overflow child builder
• ARROW-5085 - [Python/C++] Conversion of dict encoded null column fails in parquet writing when using RowGroups
• ARROW-5086 - [Python] Space leak in ParquetFile.read_row_group()
• ARROW-5089 - [C++/Python] Writing dictionary encoded columns to parquet is extremely slow when using chunk size
• ARROW-5125 - [Python] Cannot roundtrip extreme dates through pyarrow
• ARROW-5220 - [Python] index / unknown columns in specified schema in Table.from_pandas
• ARROW-5292 - [C++] Static libraries are built on AppVeyor
• ARROW-5300 - [C++] 0.13 FAILED to build with option -DARROW_NO_DEFAULT_MEMORY_POOL
• ARROW-5374 - [Python] Misleading error message when calling pyarrow.read_record_batch on a complete IPC stream
• ARROW-5414 - [C++] Using “Ninja” build system generator overrides default Release build type on Windows
• ARROW-5450 - [Python] TimestampArray.to_pylist() fails with OverflowError: Python int too large to convert to C long
• ARROW-5471 - [C++][Gandiva]Array offset is ignored in Gandiva projector
• ARROW-5522 - [Packaging][Documentation] Comments out of date in python/manylinux1/build_arrow.sh
• ARROW-5525 - [C++][CI] Enable continuous fuzzing
• ARROW-5560 - [C++][Plasma] Cannot create Plasma object after OutOfMemory error
• ARROW-5562 - [C++][Parquet] parquet writer does not handle negative zero correctly
• ARROW-5630 - [Python][Parquet] Table of nested arrays doesn't round trip
• ARROW-5638 - [C++] cmake fails to generate Xcode project when Gandiva JNI bindings are enabled
• ARROW-5651 - [Python] Incorrect conversion from strided Numpy array when other type is specified
• ARROW-5682 - [Python] from_pandas conversion casts values to string inconsistently
• ARROW-5731 - [CI] Turbodbc integration tests are failing
• ARROW-5753 - [Rust] Fix test failure in CI code coverage
• ARROW-5772 - [GLib][Plasma][CUDA] Plasma::Client#refer_object test is failed
• ARROW-5775 - [C++] StructArray : cached boxed fields not thread-safe
• ARROW-5776 - [Gandiva][Crossbow] Revert template to have commit ids.
• ARROW-5790 - [Python] Passing zero-dim numpy array to pa.array causes segfault
• ARROW-5817 - [Python] Use pytest marks for Flight test to avoid silently skipping unit tests due to import failures
• ARROW-5823 - [Rust] CI scripts miss --all-targets cargo argument
• ARROW-5824 - [Gandiva] [C++] Fix decimal null
• ARROW-5836 - [Java][OSX] Flight tests are failing: address already in use
• ARROW-5838 - [C++][Flight][OSX] Building 3rdparty grpc cannot find OpenSSL
• ARROW-5848 - [C++] SO versioning schema after release 1.0.0
• ARROW-5849 - [C++] Compiler warnings on mingw-w64
• ARROW-5851 - [C++] Compilation of reference benchmarks fails
• ARROW-5856 - [Python] linking 3rd party cython modules against pyarrow fails since 0.14.0
• ARROW-5860 - [Java] [Vector] Fix decimal byte setter
• ARROW-5863 - [Python] Segmentation Fault via pytest-runner
• ARROW-5868 - [Python] manylinux2010 wheels have shared library dependency on liblz4
• ARROW-5870 - [C++] Development compile instructions need to include “make”
• ARROW-5873 - [Python] Segmentation fault when comparing schema with None
• ARROW-5874 - [Python] pyarrow 0.14.0 macOS wheels depend on shared libs under /usr/local/opt
• ARROW-5878 - [Python][C++] Parquet reader not forward compatible for timestamps without timezone
• ARROW-5884 - [Java] Fix the get method of StructVector
• ARROW-5886 - [Python][Packaging] Manylinux1/2010 compliance issue with libz
• ARROW-5887 - [C#] ArrowStreamWriter writes FieldNodes in wrong order
• ARROW-5889 - [Python][C++] Parquet backwards compat for timestamps without timezone broken
• ARROW-5894 - [C++] libgandiva.so.14 is exporting libstdc++ symbols
• ARROW-5899 - [Python][Packaging] Bundle uriparser.dll in windows wheels
• ARROW-5910 - [Python] read_tensor() fails on non-seekable streams
• ARROW-5921 - [C++][Fuzzing] Missing nullptr checks in IPC
• ARROW-5923 - [C++] Fix int96 comment
• ARROW-5925 - [Gandiva][C++] cast decimal to int should round up
• ARROW-5930 - [FlightRPC] [Python] Flight CI tests are failing
• ARROW-5935 - [C++] ArrayBuilders with mutable type are not robustly supported
• ARROW-5946 - [Rust] [DataFusion] Projection push down with aggregate producing incorrect results
• ARROW-5952 - [Python] Segfault when reading empty table with category as pandas dataframe
• ARROW-5959 - [C++][CI] Fuzzit does not know about branch + commit hash
• ARROW-5960 - [C++] Boost dependencies are specified in wrong order
• ARROW-5963 - [R] R Appveyor job does not test changes in the C++ library
• ARROW-5964 - [C++][Gandiva] Cast double to decimal with rounding returns 0
• ARROW-5966 - [Python] Capacity error when converting large UTF32 numpy array to arrow array
• ARROW-5968 - [Java] Remove duplicate Preconditions check in JDBC adapter
• ARROW-5969 - [CI] [R] Lint failures
• ARROW-5973 - [Java] Variable width vectors' get methods should return null when the underlying data is null
• ARROW-5978 - [FlightRPC] [Java] Integration test client doesn't close buffers
• ARROW-5989 - [C++][Python] pyarrow.lib.ArrowIOError: Unable to load libjvm when using openjdk-8
• ARROW-5990 - [Python] RowGroupMetaData.column misses bounds check
• ARROW-5992 - [C++] Array::View fails for string/utf8 as binary
• ARROW-5996 - [Java] Avoid resource leak in flight service
• ARROW-5999 - [C++] Required header files missing when built with -DARROW_DATASET=OFF
• ARROW-6002 - [C++][Gandiva] TestCastFunctions does not test int64 casting`
• ARROW-6004 - [C++] CSV reader ignore_empty_lines option doesn't handle empty lines
• ARROW-6005 - [C++] parquet::arrow::FileReader::GetRecordBatchReader() does not behave as documented since ARROW-1012
• ARROW-6006 - [C++] Empty IPC streams containing a dictionary are corrupt
• ARROW-6012 - [C++] Fall back on known Apache mirror for Thrift downloads
• ARROW-6016 - [Python] pyarrow get_library_dirs assertion error
• ARROW-6029 - [R] Improve R docs on how to fix library version mismatch
• ARROW-6032 - [C++] CountSetBits doesn't ensure 64-bit aligned accesses
• ARROW-6038 - [Python] pyarrow.Table.from_batches produces corrupted table if any of the batches were empty
• ARROW-6040 - [Java] Dictionary entries are required in IPC streams even when empty
• ARROW-6046 - [C++] Slice RecordBatch of String array with offset 0 returns whole batch
• ARROW-6047 - [Rust] Rust nightly 1.38.0 builds failing
• ARROW-6050 - [Java] Update out-of-date java/flight/README.md
• ARROW-6054 - pyarrow.serialize should respect the value of structured dtype of numpy
• ARROW-6058 - [Python][Parquet] Failure when reading Parquet file from S3 with s3fs
• ARROW-6060 - [Python] too large memory cost using pyarrow.parquet.read_table with use_threads=True
• ARROW-6061 - [C++] Cannot build libarrow without rapidjson
• ARROW-6066 - [Website] Fix blog post author header
• ARROW-6067 - [Python] Large memory test failures
• ARROW-6068 - [Python] Hypothesis test failure, Add StructType::Make that accepts vector of fields
• ARROW-6073 - [C++] Decimal128Builder is not reset in Finish()
• ARROW-6082 - [Python] create pa.dictionary() type with non-integer indices type crashes
• ARROW-6092 - [C++] Python 2.7: arrow_python_test failure
• ARROW-6095 - [C++] Python subproject ignores ARROW_TEST_LINKAGE
• ARROW-6108 - [C++] Appveyor Build_Debug configuration is hanging in C++ unit tests
• ARROW-6116 - [C++][Gandiva] Fix bug in TimedTestFilterAdd2
• ARROW-6117 - [Java] Fix the set method of FixedSizeBinaryVector
• ARROW-6120 - [C++][Gandiva] including some headers causes decimal_test to fail
• ARROW-6126 - [C++] IPC stream reader handling of empty streams potentially not robust
• ARROW-6132 - [Python] ListArray.from_arrays does not check validity of input arrays
• ARROW-6135 - [C++] KeyValueMetadata::Equals should not be order-sensitive
• ARROW-6136 - [FlightRPC][Java] Don't double-close response stream
• ARROW-6145 - [Java] UnionVector created by MinorType#getNewVector could not keep field type info properly
• ARROW-6148 - [C++][Packaging] Improve aarch64 support
• ARROW-6152 - [C++][Parquet] Write arrow::Array directly into parquet::TypedColumnWriter
• ARROW-6153 - [R] Address parquet deprecation warning
• ARROW-6158 - [Python] possible to create StructArray with type that conflicts with child array's types
• ARROW-6159 - [C++] PrettyPrint of arrow::Schema missing identation for first line
• ARROW-6160 - [Java] AbstractStructVector#getPrimitiveVectors fails to work with complex child vectors
• ARROW-6166 - [Go] Slice of slice causes index out of range panic
• ARROW-6167 - [R] macOS binary R packages on CRAN don't have arrow_available
• ARROW-6170 - [R] “docker-compose build r” is slow
• ARROW-6171 - [R] “docker-compose run r” fails
• ARROW-6174 - [C++] Validate chunks in ChunkedArray::Validate
• ARROW-6175 - [Java] Fix MapVector#getMinorType and extend AbstractContainerVector addOrGet complex vector API
• ARROW-6178 - [Developer] Don't fail in merge script on bad primary author input in multi-author PRs
• ARROW-6182 - [R] Add note to README about r-arrow conda installation
• ARROW-6186 - [Packaging][C++] Plasma headers not included for ubuntu-xenial libplasma-dev debian package
• ARROW-6190 - [C++] Define and declare functions regardless of NDEBUG
• ARROW-6193 - [GLib] Add missing require in test
• ARROW-6200 - [Java] Method getBufferSizeFor in BaseRepeatedValueVector/ListVector not correct
• ARROW-6202 - [Java] Exception in thread “main” org.apache.arrow.memory.OutOfMemoryException: Unable to allocate buffer of size 4 due to memory limit. Current allocation: 2147483646
• ARROW-6205 - [C++] ARROW_DEPRECATED warning when including io/interfaces.h from CUDA (.cu) source
• ARROW-6208 - [Java] Correct byte order before comparing in ByteFunctionHelpers
• ARROW-6210 - [Java] remove equals API from ValueVector
• ARROW-6211 - [Java] Remove dependency on RangeEqualsVisitor from ValueVector interface
• ARROW-6214 - [R] Sanitizer errors triggered via R bindings
• ARROW-6215 - [Java] RangeEqualVisitor does not properly compare ZeroVector
• ARROW-6218 - [Java] Add UINT type test in integration to avoid potential overflow
• ARROW-6223 - [C++] Configuration error with Anaconda Python 3.7.4
• ARROW-6224 - [Python] remaining usages of the ‘data’ attribute (from previous Column) cause warnings
• ARROW-6227 - [Python] pyarrow.array() shouldn't coerce np.nan to string
• ARROW-6234 - [Java] ListVector hashCode() is not correct
• ARROW-6241 - [Java] Failures on master
• ARROW-6259 - [C++][CI] Flatbuffers-related failures in CI on macOS
• ARROW-6263 - [Python] RecordBatch.from_arrays does not check array types against a passed schema
• ARROW-6266 - [Java] Resolve the ambiguous method overload in RangeEqualsVisitor
• ARROW-6268 - Empty buffer should have a valid address
• ARROW-6269 - [C++][Fuzzing] IPC reads do not check decimal precision
• ARROW-6270 - [C++][Fuzzing] IPC reads do not check buffer indices
• ARROW-6290 - [Rust] [DataFusion] sql_csv example errors when running
• ARROW-6291 - [C++] CMake ignores ARROW_PARQUET
• ARROW-6301 - [Python] atexit: pyarrow.lib.ArrowKeyError: ‘No type extension with name arrow.py_extension_type found’
• ARROW-6302 - [Python][Parquet] Reading dictionary type with serialized Arrow schema does not restore “ordered” type property
• ARROW-6309 - [C++] Parquet tests and executables are linked statically
• ARROW-6323 - [R] Expand file paths when passing to readers
• ARROW-6325 - [Python] wrong conversion of DataFrame with boolean values
• ARROW-6330 - [C++] Include missing headers in api.h
• ARROW-6332 - [Java][C++][Gandiva] Handle size of varchar vectors correctly
• ARROW-6339 - [Python][C++] Rowgroup statistics for pd.NaT array ill defined
• ARROW-6343 - [Java] [Vector] Fix allocation helper
• ARROW-6344 - [C++][Gandiva] substring does not handle multibyte characters
• ARROW-6345 - [C++][Python] “ordered” flag seemingly not taken into account when comparing DictionaryType values for equality
• ARROW-6348 - [R] arrow::read_csv_arrow namespace error when package not loaded
• ARROW-6354 - [C++] Building without Parquet fails
• ARROW-6363 - [R] segfault in Table__from_dots with unexpected schema
• ARROW-6364 - [R] Handling unexpected input to time64() et al
• ARROW-6369 - [Python] Support list-of-boolean in Array.to_pandas conversion
• ARROW-6371 - [Doc] Row to columnar conversion example mentions arrow::Column in comments
• ARROW-6372 - [Rust][Datafusion] Casting from Un-signed to Signed Integers not supported
• ARROW-6376 - [Developer] PR merge script has “master” target ref hard-coded
• ARROW-6387 - [Archery] Errors with make
• ARROW-6392 - [Python][Flight] list_actions Server RPC is not tested in test_flight.py, nor is return value validated
• ARROW-6406 - [C++] jemalloc_ep fails for offline build
• ARROW-6411 - [C++][Parquet] DictEncoderImpl::PutIndicesTyped has bad performance on some systems
• ARROW-6412 - [C++] arrow-flight-test can crash because of port allocation
• ARROW-6418 - [C++] Plasma cmake targets are not exported
• ARROW-6423 - [Python] pyarrow.CompressedOutputStream() never completes with compression=‘snappy’
• ARROW-6424 - [C++][Fuzzing] Fuzzit nightly is broken
• ARROW-6428 - [CI][Crossbow] Nightly turbodbc job fails
• ARROW-6431 - [Python] Test suite fails without pandas installed
• ARROW-6432 - [CI][Crossbow] Remove alpine crossbow jobs
• ARROW-6433 - [CI][Crossbow] Nightly java docker job fails
• ARROW-6434 - [CI][Crossbow] Nightly HDFS integration job fails
• ARROW-6435 - [CI][Crossbow] Nightly dask integration job fails
• ARROW-6440 - [CI][Crossbow] Nightly ubuntu, debian, and centos package builds fail
• ARROW-6441 - [CI][Crossbow] Nightly Centos 6 job fails
• ARROW-6443 - [CI][Crossbow] Nightly conda osx builds fail
• ARROW-6445 - [CI][Crossbow] Nightly Gandiva jar trusty job fails
• ARROW-6446 - [OSX][Python][Wheel] Turn off ORC feature in the wheel building scripts
• ARROW-6449 - [R] io “tell()” methods are inconsistently named and untested
• ARROW-6457 - [C++] CMake build locally fails with MSVC 2015 build generator
• ARROW-6461 - [Java] EchoServer can close socket before client has finished reading
• ARROW-6472 - [Java] ValueVector#accept may has potential cast exception
• ARROW-6476 - [Java][CI] Travis java all-jdks job is broken
• ARROW-6478 - [C++] Roll back to jemalloc stable-4 branch until performance issues in 5.2.x addressed
• ARROW-6481 - [Python][C++] Bad performance of read_csv() with column_types
• ARROW-6488 - [Python] pyarrow.NULL equals to itself
• ARROW-6492 - [Python] file written with latest fastparquet cannot be read with latest pyarrow
• ARROW-6502 - [GLib][CI] MinGW failure in CI
• ARROW-6506 - [C++] Validation of ExtensionType with nested type fails
• ARROW-6509 - [C++][Gandiva] Re-enable Gandiva JNI tests and fix Travis CI failure
• ARROW-6520 - [Python] Segmentation fault on writing tables with fixed size binary fields
• ARROW-6522 - [Python] Test suite fails with pandas 0.23.4, pytest 3.8.1
• ARROW-6530 - [CI][Crossbow][R] Nightly R job doesn't install all dependencies
• ARROW-6550 - [C++] Filter expressions PR failing manylinux package builds
• ARROW-6552 - [C++] boost::optional in STL test fails compiling in gcc 4.8.2
• ARROW-6560 - [Python] Failures in *-nopandas integration tests
• ARROW-6561 - [Python] pandas-master integration test failure
• ARROW-6562 - [GLib] Fix wrong sliced data of GArrowBuffer
• ARROW-6564 - [Python] Do not require pandas for invoking Array.__array__
• ARROW-6565 - [Rust] [DataFusion] Intermittent test failure due to temp dir already existing
• ARROW-6568 - [C++][Python][Parquet] pyarrow.parquet crash writing zero-chunk dictionary-type column
• ARROW-6572 - [C++] Reading some Parquet data can return uninitialized memory
• ARROW-6573 - [Python] Segfault when writing to parquet
• ARROW-6576 - [R] Fix sparklyr integration tests
• ARROW-6597 - [Python] Segfault in test_pandas with Python 2.7
• ARROW-6618 - [Python] Reading a zero-size buffer can segfault
• ARROW-6622 - [C++][R] SubTreeFileSystem path error on Windows
• ARROW-6623 - [CI][Python] Dask docker integration test broken perhaps by statistics-related change
• ARROW-6639 - [Packaging][RPM] Add support for CentOS 7 on aarch64
• ARROW-6640 - [C++] Error when BufferedInputStream Peek more than bytes buffered
• ARROW-6642 - [Python] chained access of ParquetDataset's metadata segfaults
• ARROW-6651 - [R] Fix R conda job
• ARROW-6652 - [Python] to_pandas conversion removes timezone from type
• ARROW-6660 - [Rust] [DataFusion] Minor docs update for 0.15.0 release
• ARROW-6670 - [CI][R] Fix fix for R nightly jobs
• ARROW-6674 - [Python] Fix or ignore the test warnings
• ARROW-6677 - [FlightRPC][C++] Document using Flight in C++
• ARROW-6678 - [C++] Regression in Parquet file compatibility introduced by ARROW-3246
• ARROW-6679 - [RELEASE] autobrew license in LICENSE.txt is not acceptable
• ARROW-6682 - [C#] Arrow R/C++ hangs reading binary file generated by C#
• ARROW-6687 - [Rust] [DataFusion] Query returns incorrect row count
• ARROW-6701 - [C++][R] Lint failing on R cpp code
• ARROW-6703 - [Packaging][Linux] Restore ARROW_VERSION environment variable
• ARROW-6705 - [Rust] [DataFusion] README has invalid github URL
• ARROW-6709 - [JAVA] Jdbc adapter currentIndex should increment when value is null
• ARROW-6714 - [R] Fix untested RecordBatchWriter case
• ARROW-6716 - [CI] [Rust] New 1.40.0 nightly causing builds to fail
• ARROW-6751 - [CI] ccache doesn't cache on Travis-CI
• ARROW-6760 - [C++] JSON: improve error message when column changed type
• ARROW-6762 - [C++] JSON reader segfaults on newline
• ARROW-6773 - [C++] Filter kernel returns invalid data when filtering with an Array slice