rdb <rddebl...@gmail.com> added the comment:
> > From a cursory glance at the RFC3339 spec it looks like the only other
> > change needed to fully support RFC3339 would be to support an arbitrary
> > number of sub-second digits, whereas fromisoformat() currently requires
> > either exactly 3 or 6.
>
> There are other differences, for example a comma can be used in place of a
> dot as the delimiter for fractional seconds. Looking at the grammar in the
> RFC, it seems that it might also support datetimes like 2018-W03-D4, but I
> don't see any mention of that in the text.
I think you're looking at the appendix, which collects the ABNF from
ISO 8601, but this is not part of RFC3339. The grammar for RFC3339 is
purposefully very restrictive to make parsing it simple. The comma
for delimiter is in though, good catch; also a trivial change.
> > So, I can bundle this together with a change making it more lenient about
> > the number of decimal places for seconds, and we can change the docs for
> > `fromisoformat()` to be "it accepts any RFC3339 timestamp, including those
> > generated by isoformat()".
>
> No, because the isoformat outputs are not a subset of RFC 3339. For example,
> 2015-01-01T00:00:00 is not a valid RFC 3339 datetime string, nor is
> 2015-01-01Q00:00:00, but they are valid outputs of datetime.isoformat().
> datetime.fromisoformat() also supports fractional seconds on time zone
> offsets, which is not part of ISO 8601.
Fair enough (though I'd say "isoformat()" is a misnomer then). I was
just going by your option #2. We would change the wording to imply
"supports RFC 3339 or anything produced by isoformat()"
>
> > Because what I'm trying to use it for technically falls outside the
> > intended use, I say it would make the most sense to expand the intended use
> > a bit.
>
> Is there a reason you can't use `dateutil.parser.isoparse`? The contract of
> that function is to parse any valid ISO8601 datetime, and fromisoformat is
> adapted from it.
It seems a little odd to need to pull in a third-party library for
this; it seems far more tempting for me to just do
"datetime.fromisoformat(str.replace('Z', '+00:00'))" instead since I
know my dates are produced by a JSON API.
I don't intend to get argumentative about whether supporting RFC3339
belongs in the standard library; that is clearly a decision for the
Python maintainers, and I'm not sure what criteria they follow on
this. I just find it odd to point people to a third-party library for
parsing a simple but ubiquitous date standard when there are many
modules in the standard library for far more specific use cases.
FWIW, I do think that fromisoformat() is the right function to provide
RFC3339 support. I don't think users would benefit from having to
choose between several different functions that parse similar but
subtly different date formats; this seems likely to cause confusion.
Thanks for your consideration!
----------
_______________________________________
Python tracker <rep...@bugs.python.org>
<https://bugs.python.org/issue35829>
_______________________________________
_______________________________________________
Python-bugs-list mailing list
Unsubscribe:
https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
