Hey Pier,

I’m not very knowledgeable about styles, but I think you may have the
context pattern wrong. You can use C-x h to see the context at any point
when writing a command and you can use that output as a guide when building
a pattern for a style. For example, I have a command that I wrote that is
called pb and this is the output I get when I do C-x h right after typing
pb<space>:

~% pb
tags in context :completion::complete:pb::
    argument-1 options  (_arguments _pb (eval))

The full context is :completion::complete:pb::, I think in your call to
zstyle you are missing the last two colons, so what if you try:

zstyle ':completion:*:approximate:my_command:*' max-errors 0

You don’t need the -e option because you don’t need 0 to be evaluated.
​

On Mon, Jun 18, 2018 at 2:41 PM Pier Paolo Grassi <pierpaolog@gmail.com>
wrote:

> Hello everybody, I must confess I don't have a clue when it comes to
> configure completion styles. I made various attempts but for now, I don't
> seem to grasp the syntax of ztyle.
> I have in my .zshrc (blindly copied years ago from who knows where):
>
> zstyle -e ':completion:*:approximate:*' max-errors \
> 'reply=( $(( ($#PREFIX+$#SUFFIX)/3 )) numeric )'
>
> that, as I understand, enables approximate match for all the commands when
> trying completion for a number of errors proportional to the number of
> characters of the word to be completed etc.
>
> I would like to disable this completion for a command that has a completer
> function I wrote that pass to compadd various values, since these are so
> many that the completion is very slow to show matches. So I thought that
> maybe disabling the approximate match for this command would speed things
> up.
>
> My best bet doing this was:
> zstyle -e ':completion:*:approximate:my_command' max-errors 0
>
> but this does not work. I tried also various variants to no success.
> Can someone shed some light on this please?
>
> --
> Pier Paolo Grassi
> email: pierpaolog@gmail.com
> linkedin: https://www.linkedin.com/in/pier-paolo-grassi-19300217
> founder: https://www.meetup.com/it-IT/Machine-Learning-TO
>