Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

Jira Legacy
serverSystem

...

JIRA
serverId01505d01-b853-3c2e-90f1-ee9b165564fc
keyUXPROD-3345

Problem(s):

  • Vendors often supply cataloging and shelfready information to libraries in MARC bibliographic records at point of shipment or invoicing. That data may be used to update brief Instances, Holdings, and Items created at point of order. The vendor does not have HRID or UUID information for any of the Inventory records. Without that, the best matchpoint is the FOLIO Purchase Order Line number (POL) or the Vendor Reference number (VRN). 

...

Firm order (FOLIO POL and Vendor Reference Number)

Note: POL and VRN can be in various MARC fields, usually 9xx, but sometimes 024; below are examples, not all possibilities

LDR  02524cam a2200493 i 4500
001  on1141040024
003  OCoLC
005  20210610144226.0
008  200206t20202020paua\\\\\b\\\s001\0\eng\\
010  \\$a  2020001402
040  \\$aPSt/DLC$beng$erda$cDLC$dYDX$dOCLCO$dOCLCF$dBDX$dPIT$dUKMGB$dYDX$dUPM$dYUS$dFNE$dGYG
019  \\$a1229937418
020  \\$a9780271086040$q(hardcover)
020  \\$a0271086041$q(hardcover)
024  8\$a90030228025
035  \\$a(OCoLC)1141040024$z(OCoLC)1229937418
042  \\$apcc
050  00$aND237.J76$bV36 2020
082  00$a759.13$223
100  1\$aVanDiver, Rebecca,$d1982-$eauthor.
245  10$aDesigning a new tradition :$bLoïs Mailou Jones and the aesthetics of Blackness /$cRebecca VanDiver.
264  \1$aUniversity Park, Pennsylvania :$bThe Pennsylvania State University Press,$c[2020]
264  \4$c©2020
300  \\$axii, 241 pages :$billustrations (chiefly color) ;$c27 cm
336  \\$atext$btxt$2rdacontent
336  \\$astill image$bsti$2rdacontent
337  \\$aunmediated$bn$2rdamedia
338  \\$avolume$bnc$2rdacarrier
386  \\$mGender group:$ngdr$aWomen$2lcdgt
386  \\$mNationality/regional group:$nnat$aAmericans$2lcdgt
386  \\$mOccupational/field of activity group:$nocc$aUniversity and college faculty members$2lcdgt
504  \\$aIncludes bibliographical references (pages 213-223) and index.
505  00$tIntroduction : claiming middle ground --$tSeeking success : school, society, and career aspirations --$tRoutes to roots : from Black Washington to Black Paris --$tDiasporic directions : Haiti, collage, and composite aesthetics --$tIn and out : Africa and the academy --$tConclusion : composite naming practices and art histories.
520  \\$a"A critical analysis of the art and career of African American painter Loïs Mailou Jones (1905-1998). Examines Jones's engagement with African and Afrodiasporic themes as well as the challenges she faced as a black woman artist"--$cProvided by publisher.
600  10$aJones, Lois Mailou$xCriticism and interpretation.
600  17$aJones, Lois Mailou.$2fast$0(OCoLC)fst00051276
650  \0$aAfrican diaspora in art.
650  \7$aAfrican diaspora in art.$2fast$0(OCoLC)fst01902318
655  \7$aCriticism, interpretation, etc.$2fast$0(OCoLC)fst01411635
924  \\$a90030228025
935  \\$a12345-1
949  \1$aND237.J76$bV36 2020$fART$g1$i00053505045$laada$s-$t0$u-$z05000
980  \\$a210622$b6613$d000$e5423$f928288
981  \\$bMHUMAAA$caa

...

Question

Status

Conclusion

Comments

Can we limit to 0xx and 9xx fields?

Status
subtletrue
colourGreen
titleresolved

Probably OKDevs need to confirm if MARC-MARC matching capabilities can be expanded, e.g. 924$a to 924$a
Any different considerations if the match-from and/or match-to field is repeatable?

Status
subtletrue
colourGreen
titleresolved

so

So long as matches to single instance, holdings, or item, should be able to update

From: If copies ordered at the same time, but on separate POLs (for .

Recommend that libraries use 9xxs for POLs and VRNs, instead of 024s and 035s, since less likely that there will be other kinds of numbers in specified 9xx fields, but likely to be other kinds of numbers in 024s/035s

From: If copies ordered at the same time, but on separate POLs (for different acq units or locations)

To: Instance linked to multiple orders, Holdings linked to multiple orders

Is an item ever linked to multiple orders? A-M asking on Acq channel

Needs to try each of the froms (if multiples)

Needs to try each of the tos (if multiples)

Confirm appropriate test cases for E-to-E automated tests

Status
subtletrue
colourGreen
titleresolved


See use cases above

Test 1: POL, Instance source = FOLIO, update all record types

Test 2: POL, Instance source = MARC, update all record types

Test 3: VRN, Instance source = FOLIO, update all record types

Test 4: VRN, Instance source = MARC, update all record types

Add negative test cases - no matches or multiple matches

Do we need to break out vendor ref number types, like we do for Identifiers?

Status
subtletrue
colourGreen
titleresolved


For now, do not break out the different vendor ref number types; if use case arises in the future, consider breaking out, similar to how the Instance identifiers are broken out
Can we remove some unused match data elements?

Status
subtletrue
colourGreen
titleresolved


Leave for now; as users test more of the match elements in the future, correct or amend on a case-by-case basis, Besides, that enlarges the scope of this feature
Match on POLs with which statuses?

Status
subtletrue
colourGreen
titleRESOLVED

Per Christie, would want to be able to choose the POL status for successful matches

Per Raegan, this is not a scenario that comes up for them

Open = OK to match

Pending = DO NOT MATCH

Closed = DO NOT MATCH (no use case identified yet)

If multiple are matched (like the same VRN in Law and Main order, leading to the same Instance, but different holdings and items), stop if multiple matches. Or maybe use location as a submatch to get to the right holdings/item

Christie: would be rare to want to match based on a closed PO, especially if it has cataloging implications and could result in accidental overlay of previous cataloging.

Same for Jennifer E, Leeda, Dung-Lan

A-M to talk with Devs on Friday - would it be more complexity to take status into account when matching? a little more, but infrastructure already there

Maybe have 4 matching options

POL with status = Open

POL with status = Open or Closed

VRN with status = Open

VRN with status = Open or Closed

What about multiple copies?

Status
subtletrue
colourGreen
titleRESOLVED

For now, we can only do 1 instance/holdings/item update until the multiples feature is developed

Jira Legacy
serverSystem JiraJIRA
serverId01505d01-b853-3c2e-90f1-ee9b165564fc
keyUXPROD-2741

Right now, can only create single instance, holdings, item from an incoming MARC Bib. What happens if you're trying to updating multiple holdings or items from the same MARC Bib (if you have multiple holdings/items HRIDs) (find the feature for updating multiples from the same MARC Bib)

Write a couple tests to see what happens when trying to update more than 1 item or holdings from the same MARC Bib


...

  • Thunderjet: 
    • Create API endpoint that would allow lookup by POL for related Instances, Holdings, Items
    • Already have a query for using VRNs to find POLs
    • No additional indexes needed, since there are already searches and indexes for POLs and VRNs, which probably will be fast enough
  • Folijet:
    • Spike to review the feature and create BE stories: MODDATAIMP-399
    • Create match queries and matching logic for POLs and VRNs
    • Will need cross-index query if POL/VRN matches need to be limited to specific statuses (e.g. Open, but not Pending or Closed)
    • Add stories for happy path and exception Karate tests
  • Prokopovych:
    • No dev work required
  • Other
    • Have PTF help with performance check once dev work is completed
    • Analyze MARC-MARC matching for 0xx and 9xx fields (based on the MARC query work previously done)
      • May not be needed if can use MARC-Instance matching to create/update SRS MARC Bib

...

Automated Testing Scenarios (for API Karate and UI E-to-E)

...

Pull POL or VRN from the MARC field/subfield designated in the match profile

Question

Status

Conclusion

Comments

What changes are needed to the UI matching screen?

Status
subtletrue
colourGreen
titleRESOLVED

Add POL and VRN options for each Inventory record type
Any different considerations if the match-from and/or match-to field is repeatable?

Status
subtletrue
colourBlueGreen
titleOpen

Where do we pull the POL and VRN from?

Status
subtletrue
colourGreen
titleRESOLVED

resolved

  • For MARC-to-Inventory matching, not an issue as long as incoming data is in a unique field/Ind1/2/Subfield. 
  • Recommend that libraries use 9xxs for POLs and VRNs, instead of 024s and 035s, since less likely that there will be other kinds of numbers in specified 9xx fields, but likely to be other kinds of numbers in 024s/035s

  • For MARC-to-MARC, still TBD, after spike. Probably not until Morning Glory, but prep for it in Lotus
Where do we pull the POL and VRN from?

Status
subtletrue
colourYellow
titlein process

Pull POL or VRN from the MARC field/subfield designated in the match profile

  • Need to review the work that was done on MARC Search, to see if the match data can be pulled from 024, 035, and any 9xx field. Kate: add a spike to review the MARC Search. Cannot commit to this for Lotus.
  • For MARC-to-Inventory record match, best if the value in the incoming record is in a unique MARC field (or unique indicators at least), and that field is not used for any other data in the incoming file. (e.g. if POL is in 980$a, have no other 980$a's in the incoming record)
If we only wanted to match on POLs or VRNs for orders that are Open (and maybe Closed), but NOT Pending or Closed, would that be an issues? (we already use similar logic for EDIFACT invoice matching between Invoice line and POL/VRN)

Status
subtletrue
colourYellow
titlein process

No, not an issue, but confirm with Thunderjet exactly what story/work is needed from them

POs with Ongoing status (in receipt or payment) are included with Open

Order status is at the PO level, but POL/VRN are at the POL level. May need to hit a composite endpoint to get the number and the status.

Per Siarhei H, can use cross-index query to also find the status of a POLquery to also find the status of a POL

Kate to confirm with Thunderjet:

  • Need to clarify and finalize the story for the consolidated API
  • How to get from a VRN to its associated POL
  • How to ensure that we only retrieve Inventory info for Open POLs
E-to-E automated tests; how many, and happy path only, or negative too?

Status
subtletrue
colourYellow
titlein process

See draft UI test Jira FAT-1107

See draft Karate test Jira

UIDATIMP-1050

FAT-1105

A-M to clean up the Jiras and add sample data and profiles

  • For the 4 happy path scenarios
  • Possible negative scenarios
    • No match (instance, holdings, item)
    • Multiple matches (holdings, item; any need for instance?)

...