Gustaf Waldemarson wrote:
> One noticeable caveat is that **any** parenthesis can now be additionally
> indented, e.g., the follow is now also possible:
>
>     this_is_a_tuple = (long_variable_name_here,
>                                       also_a_long_variable_name)
>
> Although, given that this can be cycled at will by the user, I'm not sure if it
> is a bad additional feature or not.
>
> Ideally, I suppose that `python-indent-context` could be modified to add a
> `:inside-cond-paren` symbol that signals that the parenthesis is for a
> conditional expression and thus the extra indentation should be applied, and not
> in any other case. That does seem a bit harder for me to fix at a cursory glance
> however, so maybe this fix is enough?

Hi Gustaf,

I agree with you in that it's better to have a new indent context, and
I tried to implement it.

At first, I thought that it would be enough to add a counterpart of
the user option `python-indent-def-block-scale' and corresponding
`:inside-paren-newline-start-from-block' context.
`python-indent-def-block-scale' can be used to customize the following
code

#+begin_src python
if (
        "VALUE" in my_unnecessarily_long_dictionary
        and some_other_long_condition_case
        ):
    do_something()
#+end_src

to be indented as follows (with a TAB at "):" line):

#+begin_src python
if (
    "VALUE" in my_unnecessarily_long_dictionary
    and some_other_long_condition_case
):
    do_something()
#+end_src

This is the style used by the popular formatter "black".

From the name `python-indent-def-block-scale' and its docstring, it is
easy to assume that it only works for def block, but in fact it works
for every blocks.  As `python-indent-def-block-scale' works only when
there is no item on the same line following the opening paren, I tried
to add a similar user option and an indent context for the opening
paren followed by some items on the same line.  It could indent as
follows:

#+begin_src python
if ("VALUE" in my_unnecessarily_long_dictionary
        and some_other_long_condition_case):
    do_something()
#+end_src

However, it could not handle correctly the following example:

#+begin_src python
elif (some_case or
          another_case):
    do_another()
#+end_src

The extra indentation is not needed here.

So I think it is best to increase the indentation only if the
calculated indentation equals to the indentation of the contents of
the block ("do_something()" in the above example).  This is similar to
the way I fixed Bug#57262.

Unlike Bug#57262, the current indentation shown below is not a
violation of the latest PEP8:

#+begin_src python
if ("VALUE" in my_unnecessarily_long_dictionary
    and some_other_long_condition_case):
    do_something()
#+end_src

Although pycodestyle reports E129 "visually indented line with same
indent as next logical line," PEP8 was changed to allow this.  This is
explained in the following issue, for example:
https://github.com/PyCQA/pycodestyle/issues/474

So changing this indentation should be a user option.  Attached is my
implementation of this.  The user option
`python-indent-block-paren-deeper' is added to customize this
indentation.  I would be glad if you could try it.