Dockerfile to build latest pyarrow version

Dockerfile

FROM ubuntu:xenial

RUN apt-get update && apt-get install -y \
        automake \
        bison \
        cmake \
        curl \
        flex \
        g++ \
        git \
        libboost-all-dev \
        libevent-dev \
        libssl-dev \
        libtool \
        pkg-config \
        wget

WORKDIR /src

RUN wget https://bootstrap.pypa.io/get-pip.py && python get-pip.py
RUN wget http://apache.crihan.fr/dist/thrift/0.9.3/thrift-0.9.3.tar.gz && tar -xvzf thrift-0.9.3.tar.gz

RUN git clone https://github.com/apache/parquet-cpp.git && git clone git://git.apache.org/arrow.git

RUN cd thrift-0.9.3 && cmake . && make install
RUN cd parquet-cpp && \
    ./thirdparty/download_thirdparty.sh && \
    ./thirdparty/build_thirdparty.sh

RUN cd parquet-cpp && \
    /bin/bash -c ". ./thirdparty/set_thirdparty_env.sh && cmake . && make install"

RUN cd arrow/cpp && \
    ./thirdparty/download_thirdparty.sh && \
    ./thirdparty/build_thirdparty.sh

RUN cd arrow/cpp && \
    /bin/bash -c ". ./thirdparty/set_thirdparty_env.sh && cmake . && make install"

ENV ARROW_HOME /usr/local

RUN pip install pandas cython

RUN cd arrow/python && \
    python setup.py build_ext --inplace

pyarrow library does not support (yet) install target on its setup script. To successfully import and test pyarrow, you need to cd in /src/arrow/python.