在 ubuntu 上从源代码安装 pyV8 时出错

Question

尝试在 ubuntu 中安装 PyV8 时，输入命令：

python setup.py build

然后它显示此错误：

error: command 'c++' failed with exit status 1

有人对此有解决方案吗？

Answer 1

这是我的 Dockerfile 中的内容。以下内容在 Debian Stretch 上经过测试并在生产环境中运行。我建议完全使用我正在使用的 PyV8 / V8 设置——我至少花了一周时间来弄清楚哪种组合不会导致内存泄漏。我还建议通读讨论和 JSContext 修复 here 和 here。

简而言之，对 PyV8 的支持几乎不存在 - 要么将它用作玩具，要么完全遵循这个配方，要么花费大量时间和精力来分叉 repo 并使其变得更好。如果从头开始，我建议改用 Node-JS，并通过一些 IPC 方法与 Python 进行通信。

ENV MY_HOME /home/forge
ENV MY_LIB $FORGE_HOME/lib

# preparing dependencies for V8 and PyV8
ENV V8_HOME $MY_LIB/v8
RUN apt-get update && \
    apt-get install -y libboost-thread-dev \
        libboost-all-dev \
        libboost-dev \
        libboost-python-dev \
        autoconf \
        libtool \
        systemtap \
        scons

# compiling an older version of boost, required for this version of V8
RUN mkdir -p $MY_LIB/boost && cd $MY_LIB/boost && \
        wget http://sourceforge.net/projects/boost/files/boost/1.54.0/boost_1_54_0.tar.gz && tar -xvzf boost_1_54_0.tar.gz && cd $MY_LIB/boost/boost_1_54_0 && \
        ./bootstrap.sh && \
        ./b2 install --prefix=/usr/local --with-python --with-thread && \
        ldconfig && \
        ldconfig /usr/local/lib

# preparing gcc 4.9 - anything newer will lead to errors with the V8 codebase
ENV CC "gcc-4.9"
ENV CPP "gcc-4.9 -E"
ENV CXX "g++-4.9"
ENV PATH_BEFORE_V8 "${MY_HOME}/bin:${PATH}"
ENV PATH "${MY_HOME}/bin:${PATH}"
RUN echo "deb http://ftp.us.debian.org/debian/ jessie main contrib non-free" >> /etc/apt/sources.list && \
    echo "deb-src http://ftp.us.debian.org/debian/ jessie main contrib non-free" >> /etc/apt/sources.list && \
    apt-get update && \
    apt-get install -y gcc-4.9 g++-4.9 && \
    mkdir -p ${MY_HOME}/bin && cd ${MY_HOME}/bin && \
    ln -s /usr/bin/${CC} ${MY_HOME}/bin/gcc && \
    ln -s /usr/bin/${CC} ${MY_HOME}/bin/x86_64-linux-gnu-gcc && \
    ln -s /usr/bin/${CXX} ${MY_HOME}/bin/g++ && \
    ln -s /usr/bin/${CXX} ${MY_HOME}/bin/x86_64-linux-gnu-g++

# compiling a specific version of V8 and PyV8, since older combos lead to memory leaks
RUN git clone https://github.com/muellermichel/V8_r10452.git $V8_HOME && \
    git clone https://github.com/muellermichel/PyV8_r429.git $MY_LIB/pyv8 && \
    cd $MY_LIB/pyv8 && python setup.py build && python setup.py install

# cleaning up
RUN PATH=${PATH_BEFORE_V8} && \
    head -n -2 /etc/apt/sources.list > ${MY_HOME}/sources.list.temp && \
    mv ${MY_HOME}/sources.list.temp /etc/apt/sources.list && \
    apt-get update
ENV PATH "${PATH_BEFORE_V8}"
ENV CC ""
ENV CPP ""
ENV CXX ""

依赖于现已失效的 googlecode 并为 Ubuntu 12.04 制作的旧版本：

export MY_LIB_FOLDER=[PUT-YOUR-DESIRED-INSTALL-PATH-HERE]

apt-get install -y libboost-thread-dev
apt-get install -y libboost-all-dev
apt-get install -y libboost-dev
apt-get install -y libboost-python-dev
apt-get install -y git-core autoconf libtool systemtap
apt-get install -y subversion

apt-get install -y wget
mkdir -p $MY_LIB_FOLDER/boost && cd $MY_LIB_FOLDER/boost && wget http://sourceforge.net/projects/boost/files/boost/1.54.0/boost_1_54_0.tar.gz && tar -xvzf boost_1_54_0.tar.gz
cd $MY_LIB_FOLDER/boost/boost_1_54_0 && ./bootstrap.sh && ./b2 install --prefix=/usr/local --with-python --with-thread && ldconfig && ldconfig /usr/local/lib
svn checkout -r10452 http://v8.googlecode.com/svn/trunk/ $MY_LIB_FOLDER/v8
export V8_HOME=$MY_LIB_FOLDER/v8
svn checkout -r429 http://pyv8.googlecode.com/svn/trunk/ $MY_LIB_FOLDER/pyv8
git clone https://github.com/taguchimail/pyv8-linux-x64.git $MY_LIB_FOLDER/pyv8-taguchimail && cd $MY_LIB_FOLDER/pyv8-taguchimail && git checkout origin/stable
apt-get install -y scons
cd $MY_LIB_FOLDER/pyv8 && patch -p0 < $MY_LIB_FOLDER/pyv8-taguchimail/patches/pyv8.patch && python setup.py build && python setup.py install 

Answer 2

遇到了同样的问题，这对我有用：

export LIB=~
apt-get install -y curl libboost-thread-dev libboost-all-dev libboost-dev libboost-python-dev git-core autoconf libtool
svn checkout -r19632 http://v8.googlecode.com/svn/trunk/ $LIB/v8
export V8_HOME=$LIB/v8
svn checkout http://pyv8.googlecode.com/svn/trunk/ $LIB/pyv8 && cd $LIB/pyv8 && python setup.py build && python setup.py install

在 cmets 中找到解决方案 - https://code.google.com/p/pyv8/wiki/HowToBuild

Answer 3

我正在使用基于 Debian 的发行版。这是我安装 PyV8 的方法（你需要 已安装 git）：

cd /usr/share

sudo git clone https://github.com/emmetio/pyv8-binaries.git

cd pyv8-binaries/

sudo 解压 pyv8-linux64.zip

sudo cp -a PyV​​8.py _PyV8.so /usr/bin