To learn about ptpython and why it's cool and useful, check out its git homepage.

Sometimes you just can't beat manually testing something out in a REPL environment. When I do that with Python, I really like using ptpython. The Python REPL is quite nice (at least when full readline support is active), but ptpython is much nicer.

I also make a lot of use of Python virual environments. Unfortunately, to use ptpython with a Python virtual environment, you normally need to either install ptpython in the Python virtual environment you're working in, or install it in your global Python site-packages (which you then set the virtual environment to use). Either approach can be inconvenient. Since ptpython itself depends on several other packages, it's possible they will conflict with versions your project uses. Also, installing ptpython in a project's virtual environment likely means your add libraries that aren't necessary to the project you are working on, which is not ideal for things like pip freeze.

These are the main problem these tools solve. They provide the following features.

• The ability to start ptpython without installing it in the active virtual environment or the Python install it's based on. There are some caveats on this - see below.
• The ability to work with both Python 2.7 and 3.x virtual environments.
• A distinct REPL history file per virtual environment.
• An optional message reminding you what virtual environment (if any) you have active when you start a a REPL (plain python or ptpython).

To work in this way, ptpython is installed as a stand-alone library directory and made accessible by adding its location to the PYTHONPATH environment variable.

Grab the latest install_ptp_tools.bin file from the github release page. Alternatively, clone the repo and create the installer yourself using the build_sfx.sh script.

Run install_ptp_tools.bin script. It will ask you where you want to install the ptpython libraries. The default location is ~/opt/ptpython/site-lib. It doesn't matter where it's installed, as long as you have write access to that location.

The installer will also prompt you for the location of script files needed to launch ptpython through this tool. The default location is ~/bin/ptpython.

When everything else is done, the script will give you instructions on how to update your shell profile script to add the tools to you PATH, based on where you installed it. By default, that would be the following line. (The example assumes you use bash as a shell.)

export PATH=~/bin/ptpython:$PATH

If you also want to use the virtual environment message for REPLs, add this as well.

export PYTHONSTARTUP=~/bin/ptpython/.pythonrc.py

Once all the above steps are done, you should be able to start an interactive ptpython session by running the command ptp. Why not use ptpython? Well, for one thing, ptp is shorter, but mainly I didn't want to mask any actual installs of ptpython on the PATH. If you don't like the name ptp (I can easily imagine that name colliding with other tools), feel free to rename or symlink it.

As mentioned above, ptpython depends on a number of other packages. When you run ptpython with another Python virtual environment active, ptpython's virtual environment's package paths will come first (have higher priority) when Python loads packages. If your virtual environment uses any of the same libraries as ptpython, and you use different versions, then ptpython may cause your code to run with different versions of packages than you expect. However, this only affects code you run within the ptpython REPL.

Similarly, if you have set a PYTHONPATH of your own, your settings will have a higher priority than the library path for ptpython. If your PYTHONPATH contains conflicting versions of libraries used by ptpython, then ptpython might not work correctly.

When it starts, the ptp command tries to determine if ptpython is already installed in your current (virtual) environment. If it is, ptp will defer to your environment for ptpython's libraries rather than using its own installed version.

History files are stored in ~/.ptpython. The normal history file is simply named history. The virtual-environment-specific history files are named history-$VIRTUAL_ENV. There is no explicit cleanup of these files, so if you create and delete virtual environments often, you may end up with a lot of "orphaned" history files.

This also means that the history will persist even if you delete and recreate a virtual environment with the same name. This may or may not be what you want. Putting them somewhere else, like in the virtual environment directory, would not be very hard. It can be changed by editing the ptp.py script installed alongside ptp.

If you want to upgrade the version of ptpython used by the ptp script, just run the following command: ptp --upgrade. This will use the pip currently on your PATH but will use it to explicitly reinstall/upgrade ptpython in the path selected for its libraries during the installation of these tools. It won't change your current Python environment's installed site libraries, even if ptpython is installed there.

If you find any bugs or have suggestions, feel free to open an issue on GitHub.

This bundle has been tested on Ubuntu 16.04 and CentOS 6/7.

It's possible some of the switches I've used with certain commands won't work on macOS systems. If so, let me know and I'll look for more compatible ways to achieve the same effect.