Why not push?

Why not push?

[Marcin Borkowski]

Hi there,

I don't want to be nitpicking, but I'm just curious.  I'm looking at the
function `org-split-string'.  It uses (two times) the following
construction:

(setq list (cons (something) list))

Is there any particular reason for not using `push' there?

Best,

-- 
Marcin Borkowski
http://octd.wmi.amu.edu.pl/en/Marcin_Borkowski
Faculty of Mathematics and Computer Science
Adam Mickiewicz University

Re: Why not push?

[Nicolas Goaziou]

Hello,

Marcin Borkowski <mbork@wmi.amu.edu.pl> writes:

> I don't want to be nitpicking, but I'm just curious.  I'm looking at the
> function `org-split-string'.  It uses (two times) the following
> construction:
>
> (setq list (cons (something) list))
>
> Is there any particular reason for not using `push' there?

No. Good luck with refactoring "org.el". ;)


Regards,

-- 
Nicolas Goaziou

Re: Why not push?

[Eric Abrahamsen]

Nicolas Goaziou <mail@nicolasgoaziou.fr> writes:

> Hello,
>
> Marcin Borkowski <mbork@wmi.amu.edu.pl> writes:
>
>> I don't want to be nitpicking, but I'm just curious.  I'm looking at the
>> function `org-split-string'.  It uses (two times) the following
>> construction:
>>
>> (setq list (cons (something) list))
>>
>> Is there any particular reason for not using `push' there?
>
> No. Good luck with refactoring "org.el". ;)

Shhh... if you don't say anything, he might go and do it!

Re: Why not push?

[Marcin Borkowski]

On 2015-02-15, at 11:42, Eric Abrahamsen <eric@ericabrahamsen.net> wrote:

> Nicolas Goaziou <mail@nicolasgoaziou.fr> writes:
>
>> Hello,
>>
>> Marcin Borkowski <mbork@wmi.amu.edu.pl> writes:
>>
>>> I don't want to be nitpicking, but I'm just curious.  I'm looking at the
>>> function `org-split-string'.  It uses (two times) the following
>>> construction:
>>>
>>> (setq list (cons (something) list))
>>>
>>> Is there any particular reason for not using `push' there?
>>
>> No. Good luck with refactoring "org.el". ;)
>
> Shhh... if you don't say anything, he might go and do it!

;-)

Actually, not.  (Though it might make a nice student project.  And it
would be a good idea to do it.)  But I haven't signed the FSF papers.

OTOH, I have a personal project where I devote some amount of time every
day to read someone else's code.  I started with parts of simple.el, and
now the time has come for ox-latex (and its dependencies).

Best,

-- 
Marcin Borkowski
http://octd.wmi.amu.edu.pl/en/Marcin_Borkowski
Faculty of Mathematics and Computer Science
Adam Mickiewicz University

Re: Why not push?

[Rasmus]

Marcin Borkowski <mbork@wmi.amu.edu.pl> writes:

> Hi there,
>
> I don't want to be nitpicking, but I'm just curious.  I'm looking at the
> function `org-split-string'.  It uses (two times) the following
> construction:
>
> (setq list (cons (something) list))
>
> Is there any particular reason for not using `push' there?

These days you even have split-string in subr which you can make behave
like org-split-string.

Oh, and there's all the cl re-implementations...  I like org-some better
than cl-some just cause it sounds nice when you say it (try): 

     org (tiny pause) some.

—Rasmus

-- 
When in doubt, do it!

Re: Why not push?

[Eric Abrahamsen]

Marcin Borkowski <mbork@wmi.amu.edu.pl> writes:

> On 2015-02-15, at 11:42, Eric Abrahamsen <eric@ericabrahamsen.net> wrote:
>
>> Nicolas Goaziou <mail@nicolasgoaziou.fr> writes:
>>
>>> Hello,
>>>
>>> Marcin Borkowski <mbork@wmi.amu.edu.pl> writes:
>>>
>>>> I don't want to be nitpicking, but I'm just curious.  I'm looking at the
>>>> function `org-split-string'.  It uses (two times) the following
>>>> construction:
>>>>
>>>> (setq list (cons (something) list))
>>>>
>>>> Is there any particular reason for not using `push' there?
>>>
>>> No. Good luck with refactoring "org.el". ;)
>>
>> Shhh... if you don't say anything, he might go and do it!
>
> ;-)
>
> Actually, not.  (Though it might make a nice student project.  And it
> would be a good idea to do it.)  But I haven't signed the FSF papers.
>
> OTOH, I have a personal project where I devote some amount of time every
> day to read someone else's code.  I started with parts of simple.el, and
> now the time has come for ox-latex (and its dependencies).

Ah well, maybe we'll trick someone else into it :)

My own personal project is to spend a bit of every day working on the
emacs packages that make my computing life possible: Org, Gnus, and
BBDB. I'd prefer to spend most of it on Org, but the fact is that when
your email setup is broken, that always takes priority. Org and Gnus
both suffer (and both benefit) from wild and messy codebases. Org has
Nicolas to help whip it into shape; Gnus doesn't.

If you're just reading code, ox-latex is a great place to start. If
you're fixing code, org-agenda could use some love...

Eric

Re: Why not push?

[Yuri Niyazov]

On Sun, Feb 15, 2015 at 2:38 AM, Nicolas Goaziou <mail@nicolasgoaziou.fr> wrote:
> No. Good luck with refactoring "org.el". ;)

Could you clarify what you think are the biggest issues with
refactoring org.el ?

Re: Why not push?

[Nicolas Goaziou]

Yuri Niyazov <yuri.niyazov@gmail.com> writes:

> Could you clarify what you think are the biggest issues with
> refactoring org.el ?

Its size.

Regards,

Re: Why not push?

[Marcin Borkowski]

On 2015-02-15, at 22:22, Yuri Niyazov <yuri.niyazov@gmail.com> wrote:

> On Sun, Feb 15, 2015 at 2:38 AM, Nicolas Goaziou <mail@nicolasgoaziou.fr> wrote:
>> No. Good luck with refactoring "org.el". ;)
>
> Could you clarify what you think are the biggest issues with
> refactoring org.el ?

IMHO, the biggest and most useful thing would be to divide the humongous
functions into smaller, reusable parts.

Best,

-- 
Marcin Borkowski
http://octd.wmi.amu.edu.pl/en/Marcin_Borkowski
Faculty of Mathematics and Computer Science
Adam Mickiewicz University

Re: Why not push?

[Marcin Borkowski]

On 2015-02-15, at 13:26, Eric Abrahamsen <eric@ericabrahamsen.net> wrote:

> Ah well, maybe we'll trick someone else into it :)

;-)

> My own personal project is to spend a bit of every day working on the
> emacs packages that make my computing life possible: Org, Gnus, and
> BBDB. I'd prefer to spend most of it on Org, but the fact is that when
> your email setup is broken, that always takes priority. Org and Gnus
> both suffer (and both benefit) from wild and messy codebases. Org has
> Nicolas to help whip it into shape; Gnus doesn't.

That is a good one.  It wouldn't work for me, though, for reason I've
explained earlier.

> If you're just reading code, ox-latex is a great place to start. If
> you're fixing code, org-agenda could use some love...

Well, since I want to write a modification of the LaTeX exporter,
ox-latex seems a natural thing to look at.  OTOH, org-agenda looks scary
;-).

> Eric

Best,

-- 
Marcin Borkowski
http://octd.wmi.amu.edu.pl/en/Marcin_Borkowski
Faculty of Mathematics and Computer Science
Adam Mickiewicz University

Re: Why not push?

[Marcin Borkowski]

On 2015-02-15, at 12:35, Rasmus <rasmus@gmx.us> wrote:

> Marcin Borkowski <mbork@wmi.amu.edu.pl> writes:
>
>> Hi there,
>>
>> I don't want to be nitpicking, but I'm just curious.  I'm looking at the
>> function `org-split-string'.  It uses (two times) the following
>> construction:
>>
>> (setq list (cons (something) list))
>>
>> Is there any particular reason for not using `push' there?
>
> These days you even have split-string in subr which you can make behave
> like org-split-string.

Interesting.  I'll look into it.

> Oh, and there's all the cl re-implementations...  I like org-some better
> than cl-some just cause it sounds nice when you say it (try): 

Yes, it seems to me that reinventing the wheel is quite common in Emacs
libraries.

>      org (tiny pause) some.

At least one lame pun comes to mind...

> —Rasmus

Best,

-- 
Marcin Borkowski
http://octd.wmi.amu.edu.pl/en/Marcin_Borkowski
Faculty of Mathematics and Computer Science
Adam Mickiewicz University

Re: Why not push?

[Sebastien Vauban]

Hi Marcin,

Marcin Borkowski wrote:
> I don't want to be nitpicking, but I'm just curious.  I'm looking at the
> function `org-split-string'.  It uses (two times) the following
> construction:
>
> (setq list (cons (something) list))
>
> Is there any particular reason for not using `push' there?

IIUC, Emacs developers did not like needing

  (require 'cl)

in the packages' code base.  Not sure why (they did not like it, or why
they did not integrate such handy functions in the Emacs core).

And it seems to have changed.  They even have a lot of `cl-*' functions
now.

Though, for whatever reason, they don't have `cl-push', but well
`cl-pushnew'...

Not all that clear to me.

Best regards,
  Seb

-- 
Sebastien Vauban

Re: Why not push?

[Bastien Guerry]

Hi Marcin,

Marcin Borkowski <mbork@wmi.amu.edu.pl> writes:

> I don't want to be nitpicking, but I'm just curious.  I'm looking at the
> function `org-split-string'.  It uses (two times) the following
> construction:
>
> (setq list (cons (something) list))
>
> Is there any particular reason for not using `push' there?

I pushed a tiny clean up here.  Next question is: why not using

(split-string STRING SEPARATORS t)

?

-- 
 Bastien

Re: Why not push?

[Nicolas Goaziou]

Hello,

Bastien Guerry <bzg@gnu.org> writes:

> I pushed a tiny clean up here.  Next question is: why not using
>
> (split-string STRING SEPARATORS t)
>
> ?

It's a first step. It would be nice to get rid of `org-split-string'
altogether in the long run. We may make `org-split-string' an obsolete
alias for `split-string', and update code base accordingly.

Regards,

-- 
Nicolas Goaziou

Re: Why not push?

[Bastien Guerry]

Nicolas Goaziou <mail@nicolasgoaziou.fr> writes:

> It's a first step. It would be nice to get rid of `org-split-string'
> altogether in the long run. We may make `org-split-string' an obsolete
> alias for `split-string', and update code base accordingly.

Agreed.  I was unclear, I meant: why not using split-string instead of
org-split-string. 

-- 
 Bastien