Skip to content

Documentation
Documentation
Commit 26b8e722 by Tom Laudeman
Options

Fix tables, fields, update prose

parent 3fa26c09
Showing with 64 additions and 43 deletions
Specifications/Resource Relation.md
......@@ -2,15 +2,22 @@
### Overview
Archival resources as described in the CPF resourceRelation should have more data than captured by
EAC-CPF. SNAC had begun this work by adding an objectXMLWrap element to the resourceRelation. This XML snippet
contained some MODS or a fragment of EAD. The data includes details about the resource and it's holding
EAC-CPF. SNAC has begun this work by adding an objectXMLWrap element to the resourceRelation. This XML snippet
contains some MODS or a fragment of EAD. The data includes details about the resource and it's holding
institution.
See these files:
https://www.dropbox.com/home/SNACCooperativeProgram/ResourceRelationsExamples/ResourceRelationsExamples.xml
https://www.dropbox.com/home/SNACCooperativeProgram/ResourceRelationsExamples/ResourceRelationstoPostrgres.txt
### Description of new fields and tables
We will add fields to table related_resource, and create new table related_resource_creator (was
related_resource_name). We will have to go back to the originally extracted SNAC CPF files because we need 2
additional fields from MODS (datafield 040$a, 300), and 1 field from EAD (prefercite).
We will add fields to table related_resource, and create new table related_resource_origination_name (was
related_resource_name). We will have to go back to the original source files because we need 3
additional fields from WorldCat MARC (datafield 040$a, 041, 300), and 1 field from EAD (prefercite).
From WorldCat MARC we need:
......@@ -19,7 +26,9 @@ Identifier. We already have this data, and a software pipeline is processing thi
in table related_resource, as well as new constellations for each repository. Ideally, the new constellations
will have an address and other geographic information as part of their place element.
2) The 300 extent data which is human readable information about the size/extent of the archival materials
2) The 041 if it exists may have a subfield with a language code.
3) The 300 extent data which is human readable information about the size/extent of the archival materials
In the case of EAD we need the \<prefercite> element where there is no \<repository> element. We will review
......@@ -35,15 +44,11 @@ a foreign key relation. A repository's role is always "repository" or
http://id.loc.gov/vocabulary/relators/rps
```
todo: are there other role types? Answer: There sort of can't be, but we need to do something with them.
Need to scan through the data for non-repo, non-orig.
The only current role types are creator and repository. Every WorldCat record was checked.
Note: repository is superceded by looking up oclc/marc org code or ead repo.
related_resource_creator is creator aka origination aka originationName as from ResourceRelationstoPostgres.txt
related_resource_origination_name is creator aka origination aka originationName as from ResourceRelationstoPostgres.txt
Example: A constellation has resourceRelation to a field book (archival object) written by Clausen, Jens (Jens
Christian), 1891-1969. The related resource creator is Clausen, Jens (Jens Christian), 1891-1969.
......@@ -52,18 +57,17 @@ Resource language is a many-to-one relation related resource. We use a reverse f
record in the language table, back to the related_resoure.id. This means we have to parse the EAD
langmaterial/language elements.
We do not have language for MARC derived records. While it is possible to get some language info from the MARC
546, the values are discursive as opposed to language codes, or even a textual language name.
Daniel wants to use language codes from MARC 041, probably $a, and implicitly, we must parse the 041 subfields
for a language code. See ResourceRelationstoPostrgres.txt
```
language.fk_id=related_resource.id and language.fk_table='related_resource'
```
Resource creator name is a many-to-one. Use a reverse foreign key from related_resource_creator.fk_id to related_resource.id.
Resource creator name is a many-to-one. Use a reverse foreign key from related_resource_origination_name.fk_id to related_resource.id.
```
related_resource_creator.fk_id=related_resource.id and related_resource_creator.fk_table='related_resource'
related_resource_origination_name.fk_id=related_resource.id and related_resource_origination_name.fk_table='related_resource'
```
The place and address associated with a repository is handled via a place in the constellation. Address data
......@@ -87,7 +91,7 @@ Several new fields are added to the related_resource table:
As noted above, we did not capture the MARC 300 extent, so I will have to parse the original WorldCat records for that data.
The related_resource_creator table is new.
The related_resource_origination_name table is new.
```
......@@ -95,44 +99,61 @@ The related_resource_creator table is new.
--
-- The role aka roleTerm of repo_ic_id is always http://id.loc.gov/vocabulary/relators/rps
--
create table related_resource (
id int default nextval('id_seq'),
is_deleted boolean default false,
-- this constellation information
version int not null,
ic_id int not null,
is_deleted boolean default false, --
role int, -- @xlink:role, fk to vocabulary.id, type document_type, e.g. ArchivalResource
arcrole int, -- @xlnk:arcrole, fk to vocabulary.id type document_role, creatorOf, referencedIn, etc
type int, -- @xlink:type, fk to vocabulary.id type source_type, always "simple"?
href text, -- @xlink:href, link to the resource
relation_type text, -- @resourceRelationType, only from AnF, maybe put in a second table
relation_entry text, -- relationEntry (name) of the related eac-cpf record (should be unnecessary in db)
relation_entry_type text, -- relationEntry@localType, AnF, always "archival"?
-- connection between this constellation and resource
arcrole int, -- @xlnk:arcrole, fk to vocabulary.id type document_role, creatorOf, referencedIn, etc
-- information about the resource
relation_entry text, -- relationEntry — “shortcut” description of this resource
title text, -- resource title, from EAD/MODS
abstract text, -- resource abstract, from EAD/MODS
extent text, -- resource extent, from EAD and from original MARC
href text, -- @xlink:href, URL link to the resource
repo_ic_id int, -- holding repository ic_id, fk to ic_id
-- has name, address, and location information of the holding institution
role int, -- @xlink:role, fk to vocabulary.id, type document_type, e.g. ArchivalResource
-- to be deprecated
type int, -- @xlink:type, fk to vocabulary.id type source_type, always "simple"?
-- to be deprecated
relation_type text, -- @resourceRelationType, only from AnF, maybe put in a second table
relation_entry_type text, -- relationEntry@localType, AnF, always "archival"?
-- to be deprecated
object_xml_wrap text, -- contains more information about the resource (title, extent, etc)
-- other useful information
descriptive_note text,
object_xml_wrap text, -- from objectXMLWrap, xml
repo_ic_id int, -- holding repository ic_id, fk to ic_id
res_title text, -- resource title, from EAD?
res_abstract text, -- resource abstract, from EAD/
res_extent text, -- resource extent, from EAD and from original MARC
primary key(id, version)
);
);
--
-- This is origination name aka creator.
-- No need for field role aka roleTerm since all these names are creators.
-- This is origination name aka originationName aka creator.
-- No need for field role aka roleTerm since all these names are creators. The term "origination" seems to be from EAD.
--
-- http://id.loc.gov/vocabulary/relators/cre
--
-- This table can be updated via the web UI, so it needs the id,version,ic_id. Name is simply a string and we
-- will not require them to be SNAC identities at this time. Later we will link these names to constellations
-- via ic_id.
-- will not require names to be SNAC identities at this time. Later we will link these names to constellations
-- via field name_ic_id.
--
create table related_resource_creator (
id int default nextval('id_seq'),
version int not null,
ic_id int, -- ic_id of this creator, eventually filled in by humans
name text, -- name of creator of the related resource
fk_id int, -- fk to related_resource.id
fk_table text default 'related_resource' -- related table name
create table related_resource_origination_name (
id int default nextval('id_seq'),
version int not null,
ic_id int not null,
name_ic_id int, -- ic_id of this creator, eventually filled in by humans
name text, -- name of creator of the related resource
fk_id int, -- fk to related_resource.id
fk_table text default 'related_resource' -- related table name
);
```
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment