Re: [PATCH] Implement `equal' for hash tables

On Sun, Aug 16, 2026 at 6:13 AM Sean Whitton <[email protected]>
wrote:

> Björn A. Lindqvist [14/Aug  9:27pm +02] wrote:
> > On 8/14/26 13:13, Sean Whitton wrote:
> >> I think we'd really like to see examples.  I don't think I have ever
> >> done it so I have a hard time believing "extremely"
> >
> > Suppose you have same JSON configuration. The most natural
> > representation of it is a list of hash tables that contain nested hash
> > tables and lists. You perform some expensive operation only if the
> > configuration changes:
> >
> > (when (not (equal old-config new-config))
> >    (expensive-op))
> >
> > The configuration may be a parse tree, key-value store, or any other
> > kind of database for which nested hash tables is a convenient
> > representation.
>
> Oh, interesting example, thanks.
>

If hash table insertion order matters, then the equal operation would need
an argument to compare (nested) hash tables in an order-insensitive way.  I
believe the Java Jackson package's equal operation ignores element order.
It's been a while in Python but IIRC, one has to perform key sorting
shenanigans to compare json documents parsed into a dictionary.  And then
there are the data type comparator issues where two json documents might be
"human equal" but not technically equal; the most common thing I've seen is
that json has no standards for dates and times so it's a mess +/- your
level of control over the json documents.