Closed ColeDCrawford closed 5 months ago
@aweakley do you want to remove the old parentheses grouping examples from BAD_EXAMPLES
here? Those are from the old spec.
Do we have any examples of the difference between null
and ..
for extended intervals? It doesn't look like the current version of python-edtf
has any open
syntax?
Do the parentheses ones make things more complex? I quite like that it's clear from the tests that inputs like that will raise an error, but if it'll make our lives easier to get rid of them then I'm happy with that.
Do the parentheses ones make things more complex? I quite like that it's clear from the tests that inputs like that will raise an error, but if it'll make our lives easier to get rid of them then I'm happy with that.
Nope, they fail just fine as they currently are so I'm also happy to leave them.
Not sure why the tests are hanging but they look like they are passing if you click in.
I've updated the tests, specifically EDTF level 1 "Qualification of a date (complete)" dates. These parse as UncertainOrApproximate
classes.
This looks great, thank you.
This PR updates the natural language parser to work with the 2018 spec. As noted in the EDTF docs,
Checklist for the PR:
u
withX
?~
into%
for uncertain + approximateunknown
withnull
syntax in intervalsopen
with..
syntax syntax in intervalsy
,e
, andp
withY
,E
, andS
syntax for exponential years and significant digits