Add further tests for sn, sl, nd abbreviations

[hornc]

Closes #

Using test data supplied by @tfmorris in #7652 (comment)

Technical

The original test-publish-sn-sl.mrc test record has the place $a and publisher name $b swapped. I "corrected" it in the test expectations. I should make it case-insensitive, but I don't think we should have the swapped fields in the test example. I was planning to swap them back.

Original MARC:

00764cam a2200241Ia 4500
001 5415173
005 20221110033004.0
008 050902s190u    xx            000 0 eng d
035    $a (OCoLC)ocm61406084
035    $a (NNC)5415173
035    $a 5415173
040    $a ZCU $c ZCU
043    $a a-tu---
050  4 $a BV3170 $b .B65 1900
100 1  $a Bliss, E. E.
245 10 $a Indirect results of missionary labor in northern Turkey / $c by E.E. Bliss, D.D., of Constantinople.
260    $a [S.n.] : $b [s.l.], $c [between 1900 and 1909]
300    $a 7 pages ; $c 24 cm
336    $a text $b txt $2 rdacontent
337    $a unmediated $b n $2 rdamedia
500    $a Caption title.
650  0 $a Missions $z Turkey. $0 http://id.loc.gov/authorities/subjects/sh2010102169
852 80 $b uts,mrlxxp $h 1884

I plan to change the 260 line to:
260 $a [s.l.] : $b [S.n.], $c [between 1900 and 1909]

So it becomes a correct field inupt, but case insensitive, case.

The other test case skips the date entirely, but that seems to be existing behaviour. I think it's checking for a date in 008, but I'll need to figure out exactly how this works. The current behaviour of not adding a date at all (because 008 is empty) doesn't seem entirely wrong. Maybe no date is ok in this case?

Testing

Screenshot

Stakeholders

@tfmorris

[tfmorris]

Thanks!

I plan to change the 260 line to:
260 $a [s.l.] : $b [S.n.], $c [between 1900 and 1909]

Oops, sorry about that! I was just scanning MARC files for my search strings and didn't cast a critical eye over the results. I agree with your correction.

I think this is the offending code for the 008:

openlibrary/openlibrary/catalog/marc/parse.py

Lines 687 to 688 in 524cb8f

	if publish_date.isdigit() and publish_date != '0000':
	edition["publish_date"] = publish_date

It is rejecting the value "190u" and the corresponding 260$c of "[between 1900 and 1909]" isn't being picked either. One or the other should be included.

In particular, .isdigit() should be replaced with the regex r"^[0-9]+u*$"

[hornc]

@tfmorris Thanks for taking a look -- I've implemented your regex to extract the 190u date and the abbreviations are now case insensitive. This is ready for final review.

[tfmorris]

LGTM. Thanks!