Closed Gyukat closed 4 years ago
(Related issue: https://github.com/EQAR/eqar_backend/issues/217)
I think in general if we start to include related institutions in the search how and where they display, we (I) should seriously redesign the whole institution webAPI which takes time and planning.
Plus we should also think about how filters are effecting the search results. (If you search for a specific year, how parents/children should show up in the hitlist?) This is now basically rewriting the whole background logic of the institution search. Just to prepare, this will take some time.
Somehow it would be nice to have some examples how hits should look like / appear. So let's say someone searches for a string which exists in a parent unit name and one of the child unit's name. Currently you are getting two hits...
How this should be changed?:
What about the rest of the filters?
I have dealt with the above in the related card: #217
Regarding this particular request, I am also attaching the former spec.
Institutional Relationships_2018-04-04.docx
The relevant section is: III. Display of related institution information in institution records
When an institution with no active/valid reports is hit then you will simply retrieve a record with no reports displayed and you will have to activate the historical function to see old reports.
THIS WILL CHANGE. (WE ARE DEFAULTING TO SEEING HISTORICAL REPORTS NOW...NOT IMPORTANT FOR YOU I THINK.)
Relationships of historical type (type 2, 3, 4) should be displayed simply as a link from the record with note beneath. Label can read “Historical note” followed by bulleted list:
CHANGED TO "HISTORICAL RELATIONSHIPS"
SLIGHTLY CHANGED FORMAT, BUT MORE OR LESS TRUE.
NOT SURE IF THIS WAS DONE BY SLIK.
NEW SPECS. FOR SHOWING HISTORICALLY RELATED REPORTS ON INSTITUTION PAGES. Should include reports from:
- On SUCCEEDING TARGET record: all PRECEDING SOURCE record institutional and programme reports whose report validity extends beyond the succession, merger or split date. (Note, this does not work in inverse as it would violate the relationship logic.)
- On ABSORBing SOURCE record: all ABSORBED TARGET record institutional and programme reports whose report validity extends beyond the absorption date. (Note, this does not work in inverse as it would violate the relationship logic.)
- On SPUN OFF TARGET record: all SOURCE of SPIN OFF record institutional and programme reports whose report validity extends beyond the spin off date. (Note, this does not work in inverse as it would violate the relationship logic.)
- In ALL cases, these are effectively earlier records whose reports show up on later records.
- Display of Reports should follow naming convention used in hierarchical records.
Here is an example of how reports might show up on a historically related record:
Sibelius (my favourite absorbed institution) has a 2013 report that is valid until 2019
Sibelius was absorbed by University of the Arts Helsinki in 2013.
The report was still valid at the moment that Sibelius was absorbed by University of the Arts (actually the date of absorption PRECEDES the report, which is a bit strange)...therefore the Sibelius report should show up on the University of the Arts page (under a heading that includes the name Sibelius: "Institutional audit, Sibelius Academy (by FINEEC)")
Let's check through this!
The reports show up, but the data is lacking the institution_relationship_context
information (name, etc.).
Still a problem...for instance, the report from Sybillius shows up as a "normal" report for University of the Arts Helsinki.
We don't see this...it didn't work on the record that we use as an example...https://www.eqar.eu/qa-results/institution/?id=4269
@JoshBone Sorry, I realise the info is there but just not "consumed" by the template. But I now remember the reason, I think we discussed also in Vienna: hierarchical info so far came out of the institution_relationship_context
array, while all info (hierarchical relations, historical relations, and institutions with whom a joint programme is offered) is in the new institutions
array.
So that is definitely better and more flexible, it just misses one info: the type of relationship (e.g. institution_relationship_context
included parent or child). So it would be great if we could simply add parent, child, historic, joint_programme, ... or some sort of clarifier to see what is the relationship.
I could then change the template to use all info from institutions
and we can ultimately retire institution_relationship_context
.
REMINDER!!!! PLEASE LOOK OVER THIS!
WE have found a record that breaks this rule. See the report: https://www.eqar.eu/qa-results/search/by-institution/institution/?country=Austria&id=15
The 2013 report FINEEC is related to the University of Graz, which this institution spun off from in 2004. The report should only appear IF it was valid before the spin off OR after a MERGER.
If a report's validity date extends beyond an historical "event" (i.e. merger, split, absorption, spin off), then the report should show up on both the original institution record and also on the later related institution records. (The search should also reflect this logic.)