[PATCH] Replace latin-1-specific case-manipulation

[PATCH] Replace latin-1-specific case-manipulation

[Mattias Engdegård]

There is some broken case-manipulation code from the latin-1 era in cmpl-coerce-string-case. OK to apply?

---
Replace latin-1-specific case-manipulation

* lisp/completion.el (cmpl-coerce-string-case):
Rewrite case-changing code written for latin-1 but not valid today.

diff --git a/lisp/completion.el b/lisp/completion.el
index 14ecc9b83d..1073ae8915 100644
--- a/lisp/completion.el
+++ b/lisp/completion.el
@@ -409,10 +409,7 @@ Used to decide whether to save completions.")
 (defun cmpl-coerce-string-case (string case-type)
   (cond ((eq case-type :down) (downcase string))
        ((eq case-type :up) (upcase string))
-       ((eq case-type :capitalized)
-        (setq string (downcase string))
-        (aset string 0 (logand ?\337 (aref string 0)))
-        string)
+       ((eq case-type :capitalized) (capitalize string))
        (t string)))
 
 (defun cmpl-merge-string-cases (string-to-coerce given-string)

Re: [PATCH] Replace latin-1-specific case-manipulation

[Eli Zaretskii]

> From: Mattias Engdegård <mattiase@acm.org>
> Date: Wed, 20 Feb 2019 15:26:45 +0100
> 
> There is some broken case-manipulation code from the latin-1 era in cmpl-coerce-string-case. OK to apply?

How did you deduce it was for Latin-1?

Re: [PATCH] Replace latin-1-specific case-manipulation

[Mattias Engdegård]

20 feb. 2019 kl. 18.24 skrev Eli Zaretskii <eliz@gnu.org>:
> 
>> From: Mattias Engdegård <mattiase@acm.org>
>> Date: Wed, 20 Feb 2019 15:26:45 +0100
>> 
>> There is some broken case-manipulation code from the latin-1 era in cmpl-coerce-string-case. OK to apply?
> 
> How did you deduce it was for Latin-1?

Just a guess, from the bitwise AND with octal 337 which corresponds to the positions of upper/lower case letters in Latin-1.
But I suppose it could work for other in the ISO 8859-n series as well.

Re: [PATCH] Replace latin-1-specific case-manipulation

[Eli Zaretskii]

> From: Mattias Engdegård <mattiase@acm.org>
> Date: Wed, 20 Feb 2019 21:47:17 +0100
> Cc: emacs-devel@gnu.org
> 
> > How did you deduce it was for Latin-1?
> 
> Just a guess, from the bitwise AND with octal 337 which corresponds to the positions of upper/lower case letters in Latin-1.
> But I suppose it could work for other in the ISO 8859-n series as well.

I think it's simply because Emacs didn't have 'capitalize' back then.
That code survived since 1990.

So of course it's OK to use capitalize now.

Thanks.

RE: [PATCH] Replace latin-1-specific case-manipulation

[Drew Adams]

> I think it's simply because Emacs didn't have
> 'capitalize' back then.  That code survived
> since 1990.  So of course it's OK to use
> capitalize now.

FWIW, I think I used `capitalize', `capitalize-word',
and `capitalize-region' long ago, in the mid 1980s.

I could be mistaken, but my recollection is that
at least commands like `capitalize-word' and
`upcase-word' were used back then.

Re: [PATCH] Replace latin-1-specific case-manipulation

[Eli Zaretskii]

> Date: Wed, 20 Feb 2019 21:25:45 -0800 (PST)
> From: Drew Adams <drew.adams@oracle.com>
> Cc: emacs-devel@gnu.org
> 
> > I think it's simply because Emacs didn't have
> > 'capitalize' back then.  That code survived
> > since 1990.  So of course it's OK to use
> > capitalize now.
> 
> FWIW, I think I used `capitalize', `capitalize-word',
> and `capitalize-region' long ago, in the mid 1980s.

In what program?  Emacs was first released "in mid 1980s", did you use
Emacs before version 18.

> I could be mistaken, but my recollection is that
> at least commands like `capitalize-word' and
> `upcase-word' were used back then.

For ASCII, most probably.

Re: [PATCH] Replace latin-1-specific case-manipulation

[Mattias Engdegård]

21 feb. 2019 kl. 04.37 skrev Eli Zaretskii <eliz@gnu.org>:
> 
> I think it's simply because Emacs didn't have 'capitalize' back then.
> That code survived since 1990.
> 
> So of course it's OK to use capitalize now.

Thank you, pushed.

RE: [PATCH] Replace latin-1-specific case-manipulation

[Drew Adams]

> > > I think it's simply because Emacs didn't have
> > > 'capitalize' back then.  That code survived
> > > since 1990.  So of course it's OK to use
> > > capitalize now.
> >
> > FWIW, I think I used `capitalize', `capitalize-word',
> > and `capitalize-region' long ago, in the mid 1980s.
> 
> In what program?  Emacs was first released "in mid 1980s", did you use
> Emacs before version 18.

GNU Emacs, from the outset of my use of it.

But I don't know if it was before v18.  (I
definitely used v18.)

And I could be mistaken.  My recollection is
using capitalization commands as soon as GNU
Emacs was available (to me), in the 1980s.

My doubt is that this was introduced only in
1990 - but I could be wrong.

In any case, if there's no reason for this
code not to use `capitalize' then use
`capitalize'.  I wouldn't think that that
decision should depend on when `capitalize'
was first introduced.

> > I could be mistaken, but my recollection is that
> > at least commands like `capitalize-word' and
> > `upcase-word' were used back then.
> 
> For ASCII, most probably.

Yes, probably.  Some non-ASCII latin-1 too,
but I may have used such commands only with
ASCII.