search

PerseusDL / catalog_pending

Repository to hold new catalog source data pending integration into catalog_data
2 stars 2 forks source link

12 pending MADS records that aren't persons #17

Open cwulfman opened 6 years ago

cwulfman commented 6 years ago
  1. viaf179834949.mads.xml
  2. ActaPauli.mads.xml
  3. ActaPetri.mads.xml
  4. viaf175807631.mads.xml
  5. aethiopis.mads.xml
  6. n84093130.mads.xml
  7. n85115721.mads.xml
  8. no2017086261.mads.xml
  9. viaf183675107.mads.xml
  10. viaf305128391.marcxml.xml
  11. PrecatioOmnium.mads.xml

Books of the Bible, etc. One (viaf305128391.marcxml.xml) is an alternative name for Homer: shouldn't it be added to the record for Homer?

AlisonBabeu commented 6 years ago

Hi @cwulfman. I apologize for my delay in answering this for some reason I forgot to comment last night and only hit preview and then went to bed. Sorry. So the large number of these authority records are for textgroups rather than authors, and I created them deliberately, quite recently in fact, because there are many textgroups in the Perseus Catalog that do have authority records, e.g. Hymni Homerici (http://catalog.perseus.org/catalog/urn:cite:perseus:author.1738). This feature has actually been requested because it allows not only for various work records to get attributed to a particular textgroup, but also the authority record creates a place where you can track the history, make notes, etc. about the textgroup itself. Does that make sense?

In terms of the marcxml.xml record for Homer.Margites, there is also a MADS record in that same directory, with a canonical ID, so I'm curious why this record was pulled as MADS?

cwulfman commented 6 years ago

So the "author" in "urn:cite:perseus:author.NNNN" stands for "authority" and not for "author-the-person-who-wrote-something"? Should all the records in mads pending be assigned a urn:cite:perseus:author:NNNN id, or should they get something else?

AlisonBabeu commented 6 years ago

Um, well, I have absolutely no idea how to answer that in all honesty. Author stands for author in a very broad ethereal sense in that some authorial identity (be it a once actual person or a random textgroup wandering about) created a notional work in some kind of FRBR inspired sense. We really did not stress the semantics so much five years ago I must admit.

cwulfman commented 6 years ago

I've been assuming that the author CITE table contains only authors. Is that right? If so, the presumably I need to sort the "mads pending" records and add them to the appropriate tables.

AlisonBabeu commented 6 years ago

You know I was looking at the both the authors table last night and the textgroups CITE Table, and the truth is I'm less certain about the semantic distictions between them at this point. It appears that all "authors" appear in the "textgroups" list (which makes sense CTS wise I guess) but not all "textgroups" appear in the authors. Would there be an easy way to do a diff to see what values were unique to the textgroups cite_collections table?

cwulfman commented 6 years ago

I'm glad I'm not the only one!...

In developing this eXistdb migration, I used the Catalog's API to pull down the current CITE tables as CSV files and converted them to XML. I've checked them into the "pending review" branch of the catalog_data repo -- try checking out that branch and doing some XPath browsing in Oxygen. I have a meeting at 11 this morning: would you like to chat before or after that?

AlisonBabeu commented 6 years ago

I have meeting most of this morning, and then have to head out by 2:00 as usual. How about first thing tomorrow morning?

AlisonBabeu commented 3 years ago

Issue was "resolved" by new practice of creating MADS records for all authors and for all textgroups (eventually!)

AlisonBabeu commented 3 years ago

Turns out this issue wasn't resolved and will need to look into this.