bug#11899: 24.1.50; Weird names and unhelpful docstrings for some python function

[Fabián Ezequiel Gallina]

On 07/11/2012 09:42 AM, Ivan Andrus wrote:
> On Jul 11, 2012, at 3:14 AM, Fabián Ezequiel Gallina wrote:
> > On 07/10/2012 07:23 PM, Stefan Monnier wrote:
> > > >    python-nav-sentence-start
> > > >    python-nav-sentence-end
> > > >    python-nav-forward-sentence
> > > >    python-nav-backward-sentence
> > > > have weird names in the sense that they don't move by sentences.
> > Can you give me an example of what do you expect and the current behavior so I 
> > can have a more specific example to work on? A simplified python file and short 
> > explanation would be enough.
> I'm not complaining about the functionality here.  I just don't know what a 
> "sentence" means in a python file.  I think what is meant is a python 
> statement, but the docstrings should reflect that, or define what is meant by a sentence.
I agree here, the reason why I abused the term sentence for moving 
between statements is because there's no Emacs standard binding (nor 
concept) of moving between statements. Ideally 
python-nav-{backward,forward}-sentence must navigate blocks python code. 
Everything with an extra level of indentation after the block start must 
be considered part of the block (sentence). So given the following example:

    If something:
        do_this()
        do_that()

If the pointer is at the start of the if statement, and the user issues 
python-nav-forward-sentence he must be taken after the do_that() statement.

So what I'm going to do is to rename these existing functions to 
python-nav-{backward,forward}-statement and implement proper 
python-nav-{backward,forward}-sentence that would navigate blocks of code.

> That said, I did just notice that M-a (python-nav-backward-sentence) gets 
> "stuck" on the comment in the test below.  If I put a blank line after the 
> comment then it doesn't.
>
> print libgap('if 4>3 then\nprint("hi");\nfi')
> # Abort/segfault
> print libgap('SmallGroup(12,3)')
>
> I can create a new bug for that if desired.
>
> -Ivan
I noticed that too, there's no need to open a ticket for it, I'll work 
on that too while solving this ticket.


Thanks!

--
Fabián E. Gallina
Co-Founder of Anue
Skype: caffeineGalli
http://www.anue.biz/