On Oct 29 2020, at 1:53 pm, Gavin Scott <ga...@learn.bio> wrote:
> If I execute a cell containing only a semicolon in 2.2.6 JupyterLab and 3.8.5
> Anaconda Python, I get back an empty string which seems odd. If I try two
> semicolons I get an error that suggests each one is getting replaced by ("")
>
> [snip]
>
> The python REPL rejects these as SyntaxErrors. Bug? Or some obscure Jupyter
> thing?
This behavior is being caused by the IPython kernel. (I can see the same
behavior in the IPython repl.) IPython contains several features for
automatically calling functions and quoting arguments [1]. Specifically, if a
line begins with a semicolon, the first word is treated as a callable, and the
rest of the line is quoted as a single argument. That is, the entry
> ;func some args
gets transformed to
> func('some args')
In a little testing, it appears that if no string follows the callable name,
the empty string is used as a default. Thus
> ;func
becomes
> func('')
Moreover, if there is whitespace following the semicolon, an identity function
is assumed. Thus,
> ; a b
becomes
> (lambda x: x)('a b')
which returns
> 'a b'
My guess, then, is that
> ;
becomes
> (lambda x: x)('')
returning
> ''
However, the entry
> ;;
doesn't have whitespace following the first semicolon, so the identity function
behavior is not invoked. The second semicolon marks the end of the statement,
so IPython tries to divide the text between there ('') into a callable name and
an argument. It uses the empty string for the function name, and then fills in
the empty string default argument, and tries to execute
> ('')('')
giving the error you note. Arguably, this is a bug, but it's hard to say what
the "correct" behavior should be.
Hope that helps,
Robert
[1]
https://ipython.readthedocs.io/en/stable/interactive/reference.html?highlight=automatic#automatic-parentheses-and-quotes
--
You received this message because you are subscribed to the Google Groups
"Project Jupyter" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to jupyter+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/jupyter/47B670EA-5632-4887-B9EE-67C9FA89D091%40getmailspring.com.
