Suppose I want use jupiter notebook/ipython as an development environment and copy everything to a python scripts afterwards. In ipython we have commands like
In [1]: cd ..
/Users/myname/Desktop/software
In [2]: ls
blah_blah_blah/
Suppose I finish my ipython notebook and want to copy everything (suppose I have 1000 lines and I could not edit them 1 by 1) to create my python script. Is it possible to enable my python script to understand such lines like "cd .." etc.
Any method for running your IPython code using a standard Python interpreter is going to be a little complicated. For example, see this question, with one of the answers illustrating the call to IPython's 'magic' methods to do a shell command:
from IPython.terminal.embed import InteractiveShellEmbed
ipshell = InteractiveShellEmbed()
ipshell.dummy_mode = True
ipshell.magic("%timeit abs(-42)")
The much easier option would be to simply use the IPython interpreter to run your saved script. You need to make sure that each shell command is preceded by a %, as this indicates a 'magic' command. Should be a simple find-and-replace task, as I doubt you are using too many shell commands. If there are a lot of different shell commands to prefix with % you could also write a short script to do this work for you. You also need to ensure that your script has the extension .ipy.
script.ipy:
%cd ..
%ls
x = "My script!"
print(x)
To run script from terminal:
>>> ipython script.ipy
As said above, we can run ipython scripts using ipython from command line:
ipython my_script.ipy
However, magic cells are currently not working well. For instance, if you have the following script:
%%writefile hello.txt
this is a line
another line
print ('hello')
my_path = '/path/to/folder'
list_of_txt = !ls {my_path}/*.txt
print (list_of_txt)
It will run the initial cell magic %%writefile as if the whole file was a single cell. In order to avoid that, you have to use get_ipython().run_cell_magic ()
Since each ipython command needs a different function, I find it helpful to write ipython commands you want in a notebook, and then in the notebook's menu do File > "Save and Export Notebook As" > "Executable Script" (in jupyter lab, but probably something similar can be done in plain jupyter notebook). It will convert all the ipython-specific code using appropriate functions from the IPython library.
For example, if we copy and paste the previous script in different cells of a notebook and export it as a .py file, it gives us the following python code:
code written in exported script:
get_ipython().run_cell_magic('writefile', 'hello.txt', 'this is a line\nanother line\n')
my_path = '/path/to/folder'
list_of_txt = get_ipython().getoutput('ls {my_path}/*.txt')
print (list_of_txt)
You can run the exported script directly using ipython from command line:
ipython my_exported_script.py
Or, inside an ipython session, using
%run my_exported_script.py
This works for IPython 1.x
osandos.pathmodules that can do things likecd.