How to install Python 2.7 and Python 3.3 on CentOS

In this guide I will show you how to install Python 2.7 and 3.3 on CentOS 6. The examples below are for Python 2.7.6 and Python 3.3.5, but the procedure is the same for any modern version of Python including the upcoming Python 3.4.0.

I make regular updates to this guide to track new versions. Please see the end of the document for a changelog.

CentOS 6 ships with Python 2.6.6 and several critical system utilities, for example yum, will break if the default Python interpreter is upgraded. The trick is to install new versions of Python in /usr/local (or some other non-standard location) so that they can live side-by-side with the system version.

This guide should work for all versions of CentOS 6, but I have only verified it on CentOS 6.5 64 bit. It will probably work for some versions of CentOS 5 also.

Execute all the commands below as root either by logging in as root or by using sudo.

Preparations – install prerequisites

In order to compile Python you must first install the development tools and a few extra libs. The extra libs are not strictly needed to compile Python but without them your new Python interpreter will be quite useless.

yum groupinstall "Development tools" yum install zlib-devel bzip2-devel openssl-devel ncurses-devel sqlite-devel readline-devel tk-devel gdbm-devel db4-devel libpcap-devel xz-devel

Things to consider

Before you compile and install Python there are a few things you should know and/or consider:

Unicode

Python has a long and complicated history when it comes to Unicode support. Unless you have very specific reasons you should configure Python 3.2 and earlier to enable UTF-32 support. This increases memory usage but improves compatibility. In Python 3.3 the Unicode support has been completely rewritten and strings are automatically stored using the most efficient encoding possible.

You enable UTF-32 in Python 2.7 by adding --enable-unicode=ucs4 to the configure command. In Python 3.2 the flag is called --with-wide-unicode.

Shared library

You should probably compile Python as a shared library. All modern Linux distros ship with Python compiled as a shared library, and there are third-party tools such as mod_wsgi and Blender that won’t work without it. If you compile Python as a shared library you must also tell it how to find the library. You have two options:

• Compile the path into the executable by adding this to the end of the configure command: LDFLAGS="-Wl,-rpath /usr/local/lib"
• Open the file /etc/ld.so.conf in a text editor and add the path /usr/local/lib to the end of it. After you have added the line you must run /sbin/ldconfig to make the dynamic linker aware of the change. This is how the file will look after adding the line on a clean install of CentOS 6.5:

  /etc/ld.so.conf

  include ld.so.conf.d/*.conf /usr/local/lib

  1 2	include ld.so.conf.d/*.conf /usr/local/lib

Use “make altinstall” to prevent problems

It is critical that you use make altinstall when you install your custom version of Python. If you use the normal make install you will end up with two different versions of Python in the filesystem both named python. This can lead to problems that are very hard to diagnose.

Download, compile and install Python

Here are the commands to download, compile and install Python. If you modify /etc/ld.so.conf as discussed above you can remove the LDFLAGS parameter below.

# Python 2.7.6: wget http://python.org/ftp/python/2.7.6/Python-2.7.6.tar.xz tar xf Python-2.7.6.tar.xz cd Python-2.7.6 ./configure --prefix=/usr/local --enable-unicode=ucs4 --enable-shared LDFLAGS="-Wl,-rpath /usr/local/lib" make && make altinstall # Python 3.3.5: wget http://python.org/ftp/python/3.3.5/Python-3.3.5.tar.xz tar xf Python-3.3.5.tar.xz cd Python-3.3.5 ./configure --prefix=/usr/local --enable-shared LDFLAGS="-Wl,-rpath /usr/local/lib" make && make altinstall

After running the commands above your newly installed Python interpreter will be available as /usr/local/bin/python2.7 or /usr/local/bin/python3.3. The system version of Python 2.6.6 will continue to be available as /usr/bin/python, /usr/bin/python2 and /usr/bin/python2.6.

Download and install Setuptools + pip

Setuptools has replaced Distribute as the official package manager used for installing packages from the Python Package Index. Each Python interpreter on your system needs its own install of Setuptools. I also suggest you install pip. It builds on top of Setuptools and provides a few extra functions that are useful when you manage your packages.

The instructions below will install the latest version of Setuptools and pip for you.

# First get the setup script for Setuptools: wget https://bitbucket.org/pypa/setuptools/raw/bootstrap/ez_setup.py # Then install it for Python 2.7 and/or Python 3.3: python2.7 ez_setup.py python3.3 ez_setup.py # Now install pip using the newly installed setuptools: easy_install-2.7 pip easy_install-3.3 pip # With pip installed you can now do things like this: pip2.7 install [packagename] pip2.7 install --upgrade [packagename] pip2.7 uninstall [packagename]

The packages will end up in /usr/local/lib/pythonX.Y/site-packages/ (where X.Y is the Python version).

What’s next?

If you are using Python 2.7 I strongly recommend that you install virtualenv and learn how to use it. Virtualenv makes it possible to create isolated Python environments. If you are using Python 3.3 then you don’t need virtualenv because that functionality is already built in.

Each isolated Python environment (also called sandbox) can have its own Python version and packages. This is very useful when you work on multiple projects or on different versions of the same project.

Create your first isolated Python environment

# Install virtualenv for Python 2.7 and create a sandbox called my27project: pip2.7 install virtualenv virtualenv-2.7 my27project # Use the built-in pyvenv program in Python 3.3 to create a sandbox called my33project: pyvenv-3.3 my33project # Check the system Python interpreter version: python --version # This will show Python 2.6.6 # Activate the my27project sandbox and check the version of the default Python interpreter in it: source my27project/bin/activate python --version # This will show Python 2.7.6 deactivate # Activate the my33project sandbox and check the version of the default Python interpreter in it: source my33project/bin/activate python --version # This will show Python 3.3.5 deactivate

When you use virtualenv to create a sandbox it will automatically install setuptools and pip for you inside the sandbox. If you use pyvenv then you must do it yourself. You can reuse the ez_setup.py file you downloaded earlier and just run it after you activate your new sandbox.

-----------------

error while loading shared libraries: xxx.so.0:cannot open shared object file: No such file or directory
                                            
那就表示系統不知道xxx.so 放在哪個目錄下。                   
這個時候就要在/etc/ld.so.conf中加入xxx.so所在的目錄。                    
一般而言，有很多so檔會在/usr/local/lib這個目錄下，所以在/etc/ld.so.conf中加入/usr/local/lib這一行，可以解決此問題。
將 /etc/ld.so.conf存檔後，還要執行「/sbin/ldconfig –v」來更新一下才會生效

/etc/ld.so.conf:
这个文件记录了编译时使用的动态链接库的路径。
默认情况下，编译器只会使用/lib和/usr/lib这两个目录下的库文件
如果你安装了某些库，没有指定 --prefix=/usr 这样lib库就装到了/usr/local下，而又没有在/etc/ld.so.conf中添加/usr/local/lib，就会报错了

ldconfig是个什么东东吧 ：
它是一个程序，通常它位于/sbin下，是root用户使用的东东。具体作用及用法可以man ldconfig查到
简单的说，它的作用就是将/etc/ld.so.conf列出的路径下的库文件 缓存到/etc/ld.so.cache 以供使用
因此当安装完一些库文件，(例如刚安装好glib)，或者修改ld.so.conf增加新的库路径后，需要运行一下/sbin/ldconfig
使所有的库文件都被缓存到ld.so.cache中，如果没做，即使库文件明明就在/usr/lib下的，也是不会被使用的，结果
编译过程中抱错，缺少xxx库。