Module: terminal.embed

An embedded IPython shell.

3 Classes

class IPython.terminal.embed.KillEmbeded

Bases: Exception

class IPython.terminal.embed.EmbeddedMagics(shell=None, **kwargs)

Bases: IPython.core.magic.Magics

exit_raise(parameter_s='')

%exit_raise Make the current embedded kernel exit and raise and exception.

This function sets an internal flag so that an embedded IPython will raise a IPython.terminal.embed.KillEmbeded Exception on exit, and then exit the current I. This is useful to permanently exit a loop that create IPython embed instance.

kill_embedded(parameter_s='')
%kill_embedded [-i] [-x] [-y]

%kill_embedded : deactivate for good the current embedded IPython

This function (after asking for confirmation) sets an internal flag so that an embedded IPython will never activate again for the given call location. This is useful to permanently disable a shell that is being called inside a loop: once you’ve figured out what you needed from it, you may then kill it and the program will then continue to run without the interactive shell interfering again.

If for some reasons you need to kill the location where the instance is created and not called, for example if you create a single instance in one place and debug in many locations, you can use the --instance option to kill this specific instance. Like for the call location killing an “instance” should work even if it is recreated within a loop.

Note

This was the default behavior before IPython 5.2

optional arguments:
-i, --instance Kill instance instead of call location
-x, --exit Also exit the current session
-y, --yes Do not ask confirmation
class IPython.terminal.embed.InteractiveShellEmbed(**kw)

Bases: IPython.terminal.interactiveshell.TerminalInteractiveShell

__init__(**kw)
__call__(header='', local_ns=None, module=None, dummy=None, stack_depth=1, global_ns=None, compile_flags=None, **kw)

Activate the interactive interpreter.

__call__(self,header=’‘,local_ns=None,module=None,dummy=None) -> Start the interpreter shell with the given local and global namespaces, and optionally print a header string at startup.

The shell can be globally activated/deactivated using the dummy_mode attribute. This allows you to turn off a shell used for debugging globally.

However, each time you call the shell you can override the current state of dummy_mode with the optional keyword parameter ‘dummy’. For example, if you set dummy mode on with IPShell.dummy_mode = True, you can still have a specific call work by making it as IPShell(dummy=False).

init_sys_modules()

Explicitly overwrite Module: core.interactiveshell to do nothing.

mainloop(local_ns=None, module=None, stack_depth=0, display_banner=None, global_ns=None, compile_flags=None)

Embeds IPython into a running python program.

Parameters:
  • module (local_ns,) – Working local namespace (a dict) and module (a module or similar object). If given as None, they are automatically taken from the scope where the shell was called, so that program variables become visible.
  • stack_depth (int) – How many levels in the stack to go to looking for namespaces (when local_ns or module is None). This allows an intermediate caller to make sure that this function gets the namespace from the intended level in the stack. By default (0) it will get its locals and globals from the immediate caller.
  • compile_flags – A bit field identifying the __future__ features that are enabled, as passed to the builtin compile() function. If given as None, they are automatically taken from the scope where the shell was called.

1 Function

IPython.terminal.embed.embed(**kwargs)

Call this to embed IPython at the current point in your program.

The first invocation of this will create an InteractiveShellEmbed instance and then call it. Consecutive calls just call the already created instance.

If you don’t want the kernel to initialize the namespace from the scope of the surrounding function, and/or you want to load full IPython configuration, you probably want IPython.start_ipython() instead.

Here is a simple example:

from IPython import embed
a = 10
b = 20
embed(header='First time')
c = 30
d = 40
embed()

Full customization can be done by passing a Config in as the config argument.