Re: [maildropl] Maildrop handling subject containing unicode characters

[Milan O. <cou...@di...>]

On Wed, 07 Sep 2022 18:21:09 -0400
Sam Varshavchik <mr...@co...> wrote:

[ snip ]

> You can use reformime to reencode unicode back into ascii-only:
> 
> $ reformime -o 'Hóla!'
> =?UTF-8?B?SMOzbGEh?=
> 
> So, you probably want something like:
> 
> Subject=`reformime -o "$MATCH1"`
> 
> and, with everything else, it should work.
> 
> maildrop does not do variable substitution inside backticks, so this
> command gets passed directly for execution to the shell, and it
> safely substitutes $MATCH1, from the inherited environment. 

Unfortunately, it does not work for me.

My snippet coming into action:

if (/^Subject: *(Test .*)/)
 {Subject=`reformime -o "$MATCH1"`
  echo "subject $MATCH1 -> $Subject"
  xfilter 'reformail -i"Subject: $Subject"'
 }

Mail sent with subject 'Test ľščťž', encoded form is

Subject: Test =?UTF-8?B?xL7FocSNxaXFvg==?=

(end of) line in log from echo statement above produced is

subject Test ľš�M-^Mťž -> Test

(letter č is not properly displayed here, which could be different
issue) and mail being delivered has subject Test. Pure ASCII, no UTF-8,
neither raw nor encoded.

> Note that, either way, changing the subject line will likely break
> any signatures that include the original subject line.

That's accepted here, I have no problem with this - if necessary, any
signature like DKIM could be removed or modified as necessary. I
understand this issue and can handle it, if it causes any issue.

Regards,
Milan