bug#65137: 29.1; completion-substring-try-completion doesn't return the longest common substring

[Stefan Monnier]

> Patch with test:

Thanks, pushed.

>> Similarly, I see that
>>
>>     (completion-pcm--merge-completions '("abr" "absabr") '(prefix "br"))
>>
>> returns
>>
>>     ("br" prefix)
>>
>> whereas it should arguably return
>>
>>     ("br" "a" prefix)
>>
>> [ Tho this may have the side effect that after this completion, `absabr`
>>   won't be considered any more, if the `basic` completion comes before
>>   `substring` :-(  ]

> I did notice this too.  I could try fixing/changing this too, but it
> does seem annoying when basic comes before substring - as it does by
> default in a number of completion categories.

Yup :-(

I have often wished for there to be a way to remember the style that was
used so as to try and avoid such "style capture".

> I wonder if we should move basic to after substring in those
> categories in completion-category-defaults?

It's a tradeoff: the default is designed to reduce the factor of
surprise for people used to the default (i.e. to mostly prefix completion).

> Or just remove basic from them.

Same difference :-)

> It doesn't seem like having both basic and substring in those lists
> has much point.

If you're used to relying on `substring` indeed it's not helpful.
But if you're used to using mostly prefix-based completion (which may
have the effect that you choose your names such that prefix completion
works well), it can be helpful to have `substring` as a fallback when
you can't remember what prefix to use for the thing you're looking for.


        Stefan