Re: [PATCH 00/32] [emacs] Add outline headings and switch to lexical scope

[Tomi Ollila <tomi.ollila@iki.fi>]

On Sun, Dec 27 2020, Tomi Ollila wrote:

> On Mon, Dec 14 2020, Jonas Bernoulli wrote:
>
>> Hello,
>>
>> This patch serious changes a number of things, including boring
>> cleanup, but the big changes are the following.  The respective
>> commits are also marked in the list below.  See these commits
>> for longer descriptions.
>>
>> 1. Split all libraries into sections by adding new headings or
>>    formatting existing headings properly to be compatible with
>>    outline-minor-mode.
>>
>> 2. Use lexical-scope in all libraries.  This is potentially faster
>>    and is recommended for all new code (and old code someone cares
>>    about enough).  It also has the advantage that it can reveal
>>    subtle bugs.
>
> The patch series did not apply for me (fully) anymore. First had to

Ok, applied when using `git am -3`

Sending this email to remind everyone the --3way option of git-am,
as it is forgotten so often (or then it is just me).

Now I have this series in use, which my User-Agent: may also reveal
(Notmuch/0.31.3+70~g95b022c (https://notmuchmail.org) Emacs/27.1).

Tomi

> be skipped, then two of more of the last, therefore I did not put
> this into use for now (as, if it fails I don't know whether the
> reason is dropped changes or my way of building it...)
>
> I like the series (on paper), 2 comments
>
> In patch 23/32, line
>  +	    (y-or-n-p (format "Fcc header %s is an absolute path%s%s" subdir
> should be
>  +	    (y-or-n-p (format "Fcc header %s is an absolute path %s %s" subdir
>
> (I.e. spaces around %s's)
>
> Then, I personally don't see enough point for requiring subr-x
> just just for changing (string= foo "") to (string-empty-p). If
> there were more features picked from subr-x then that would be
> different...
>
>> Among the many more commits that do not concern these two big
>> changes, there are a few that might need special attention.
>> (Again see the commits for details.)
>>
>> 3. No longer use Ido in notmuch-mua-prompt-for-sender.  Might be
>>    controversial.
>>
>> 4. Stop using unnecessary let-bindings, though how "unnecessary"
>>    these bindings are is somewhat subjective.
>
> I like these two changes.
>
>>
>>      Cheers,
>>      Jonas
>
> Tomi
>
> PS: I used the following code to drop cr's (\r) from base64-encoded
>     text blobs; just enough to work in this particular case.
>
> #!/usr/bin/perl
> # -*- mode: cperl; cperl-indent-level: 4 -*-
>
> use 5.8.1;
> use strict;
> use warnings;
> use MIME::Base64;
>
> while (<STDIN>) {
>     print $_;
>     if (/^Content-Transfer-Encoding: base64/) {
> 	while (<STDIN>) { print $_; last if /^\s*$/ }
> 	my @lines;
> 	while (<STDIN>) { last if /^\s*$/; push @lines, $_ }
> 	my $decoded = decode_base64 join('', @lines);
> 	$decoded =~ tr/\r//d;
> 	print(encode_base64 $decoded);
> 	print "\n"
>     }
> }