Re: fill-paragraph vs. financial institutions

Re: fill-paragraph vs. financial institutions

[Richard Stallman]

    Now you've really done it. M-x fill-paragraph on
    487.2875,
    487.3375,
    487.4000,
    ...

    gives
    487.2875, 3375, 4000, 4500, 4750, 5250, 5500, 6000, 7250, 8000, 8500,
    487.8750,


adaptive-fill-regexp has an explicit alternative to match prefixes
such as `NNN.'  This would only apply in the case where the first two
lines of a paragraph both start that way.

Maybe we should delete that case.  Is that really a case that people
use?  In some kinds of outlines, the first line of a paragraph would
start that way, but you wouldn't put the prefix on every line.
Can anyone think of a reason why we should keep this case?

Re: fill-paragraph vs. financial institutions

[Juri Linkov]

> adaptive-fill-regexp has an explicit alternative to match prefixes
> such as `NNN.'  This would only apply in the case where the first two
> lines of a paragraph both start that way.
>
> Maybe we should delete that case.  Is that really a case that people
> use?  In some kinds of outlines, the first line of a paragraph would
> start that way, but you wouldn't put the prefix on every line.
> Can anyone think of a reason why we should keep this case?

Sometimes I get bitten by this too when Emacs convert a list of floating
point numbers e.g.

0.1 0.2
0.3 0.4
0.5 0.6
0.7 0.8

into

0.1 0.2 3 0.4 5 0.6 7 0.8

Note that adaptive-fill-regexp also supports prefixes of numbers
in parentheses like "(1)".  I never encountered a situation
where such prefixes cause a problem, so I suggest only to remove
a point (underlined below) from this part of adaptive-fill-regexp:

  "\\((?[0-9]+[.)][ 	]*\\)*"
               =

-- 
Juri Linkov
http://www.jurta.org/emacs/

Re: fill-paragraph vs. financial institutions

[Richard Stallman]

    Now you've really done it. M-x fill-paragraph on
    487.2875,
    487.3375,
    487.4000,
    ...

    gives
    487.2875, 3375, 4000, 4500, 4750, 5250, 5500, 6000, 7250, 8000, 8500,
    487.8750,


adaptive-fill-regexp has an explicit alternative to match prefixes
such as `NNN.'  This would only apply in the case where the first two
lines of a paragraph both start that way.

Maybe we should delete that case.  Is that really a case that people
use?  In some kinds of outlines, the first line of a paragraph would
start that way, but you wouldn't put the prefix on every line.
Can anyone think of a reason why we should keep this case?

I propose therefore the following change.  Does anyone see a problem in it?


*** fill.el	09 Jul 2006 12:00:39 -0400	1.189
--- fill.el	03 Sep 2006 06:51:42 -0400	
***************
*** 89,95 ****
  (defcustom adaptive-fill-regexp
    ;; Added `!' for doxygen comments starting with `//!' or `/*!'.
    ;; Added `%' for TeX comments.
!   (purecopy "[ \t]*\\([-!|#%;>*·•‣⁃◊]+[ \t]*\\|(?[0-9]+[.)][ \t]*\\)*")
    "*Regexp to match text at start of line that constitutes indentation.
  If Adaptive Fill mode is enabled, a prefix matching this pattern
  on the first and second lines of a paragraph is used as the
--- 89,96 ----
  (defcustom adaptive-fill-regexp
    ;; Added `!' for doxygen comments starting with `//!' or `/*!'.
    ;; Added `%' for TeX comments.
!   ;; RMS: deleted the code to match `1.' and `(1)'.
!   "[ \t]*\\([-!|#%;>*·•‣⁃◊]+[ \t]*\\)*"
    "*Regexp to match text at start of line that constitutes indentation.
  If Adaptive Fill mode is enabled, a prefix matching this pattern
  on the first and second lines of a paragraph is used as the

RE: fill-paragraph vs. financial institutions

[Drew Adams]

        Now you've really done it. M-x fill-paragraph on
        487.2875,
        487.3375,
        487.4000,
        ...

        gives
        487.2875, 3375, 4000, 4500, 4750, 5250, 5500, 6000, 7250,
        8000, 8500, 487.8750,

    adaptive-fill-regexp has an explicit alternative to match prefixes
    such as `NNN.'  This would only apply in the case where the first two
    lines of a paragraph both start that way.
    Maybe we should delete that case.

Sorry I didn't reply earlier. I just got bit by this. Yes, please get rid of
this feature. The traditional behavior is fine: fill and indent properly
when only the first line starts that way.

IOW, this:

1. Some text here in a long line that gets filled.

should not turn into this:

1. Some text here in a long
1. line that gets filled.

It should instead turn into this:

1. Some text here in a long
   line that gets filled.


    Is that really a case that people
    use?  In some kinds of outlines, the first line of a paragraph would
    start that way, but you wouldn't put the prefix on every line.
    Can anyone think of a reason why we should keep this case?

Not I.

    I propose therefore the following change.  Does anyone see a
    problem in it?

If that puts things back the way they were in previous versions, then that's
good.

Re: fill-paragraph vs. financial institutions

[Richard Stallman]

	I propose therefore the following change.  Does anyone see a
	problem in it?

    If that puts things back the way they were in previous versions, then that's
    good.

Could you explain what you mean by that?  I don't know of any change
that was made in this since Emacs 21.

RE: fill-paragraph vs. financial institutions

[Drew Adams]

    	I propose therefore the following change.  Does anyone see a
    	problem in it?

        If that puts things back the way they were in previous
        versions, then that's good.

    Could you explain what you mean by that?  I don't know of any change
    that was made in this since Emacs 21.

You're right; I was mistaken. The same behavior is even in Emacs 20 (I
haven't used Emacs 21). I coincidentally stumbled upon it for the first time
recently, and I thought it was something new.

Since I only hit it once over several years, it might not be a common case.
In any case, I don't see the utility of the double-line-repeats-number
behavior, so add my vote to your change, unless someone points out a good
reason for this case.