[Archivesspace_Users_Group] Question Regarding the REST API batch_imports Operation

[Andrew Morrison]

If you want to load EAD via the backend API you can use the 
/jobs_with_files/ endpoint. You send that both JSON (to specify what job 
to run and give it the necessary parameters) and the EAD for the job to 
process.


Here is an example a cURL command to start a job to import an EAD file 
in the current local directory called *myeadfile.xml* to an 
ArchivesSpace system running on a machine called *myhost*:


curl -H "X-ArchivesSpace-Session: *MYTOKEN*" -F 
"job={\"jsonmodel_type\": \"job\", \"status\": \"queued\", \"job_type\": 
\"import_job\", \"job\": {\"jsonmodel_type\": \"import_job\", 
\"filenames\": [\"*myeadfile.xml*\"], \"import_type\": \"ead_xml\"}}" -F 
"files[]=@*myeadfile.xml*" 
http://*myhost*:*8089*/repositories/*2*/jobs_with_files


That assumes the backend is set up to listen (and isn't blocked by a 
firewall between you and it) on post 8089. And that you want any records 
created to belong to repository number 2. Change those as appropriate.


*MYTOKEN* is a session token which you can get from the response to this:


read -p "Username: " USER; read -s -p "Password: " PASS; curl 
-Fpassword=$PASS http://*myhost*:*8089*/users/$USER/login


You could probably specify multiple EAD files, but it is safer to always 
do one file per import job.


The HTTP response is only the status of the job, usually "Created". It 
doesn't tell you whether the import succeeds, nor the ID(s) of any 
records created. I've only ever used this to populate testing systems, 
when a few failures didn't matter. The jobs and their logs can be viewed 
in the staff interface.


Andrew.




On 07/05/2020 15:44, James R Griffin III wrote:
> I am terribly sorry, I have now found that the payload is a JSON 
> serialization of an array of  resource objects: 
> https://github.com/archivesspace/archivesspace/blob/master/backend/spec/controller_batch_import_spec.rb#L41
>
> Would one then model the payload as one of these arrays of JSON 
> objects, each with a URI internal to the EAD file?
> ------------------------------------------------------------------------
> *From:* James R Griffin III
> *Sent:* Thursday, May 7, 2020 10:30 AM
> *To:* Archivesspace Users Group 
> <archivesspace_users_group at lyralists.lyrasis.org>
> *Subject:* Question Regarding the REST API batch_imports Operation
> Hello Everyone,
>
> I have recently been reviewing the documentation for the REST API, and 
> was looking to explore the possible usage of 
> https://archivesspace.github.io/archivesspace/api/#import-a-batch-of-records
>
> Please forgive for my ignorance, but does the body of the POST request 
> contain a payload of EAD XML? Would this be a string of concatenated 
> EAD documents?
>
> Additionally, my understanding is that the response from this request 
> contains an identifier for the created object. As the job for 
> importing the records would be asynchronous, should one poll for the 
> status of this new object by repeatedly transmitting a GET request 
> against 
> https://archivesspace.github.io/archivesspace/api/#find-resources-by-their-identifiers until 
> it has been fully imported?
>
> Thank you for your patience and assistance.
>
> Sincerely,
> James
>
> -- 
>
> my.pronoun.is/he <http://my.pronoun.is/he>
>
> James R. Griffin III
>
> Digital Infrastructure Developer
>
> Princeton University Library <https://library.princeton.edu/>
>
> Princeton, NJ 08544
>
>
>
> _______________________________________________
> Archivesspace_Users_Group mailing list
> Archivesspace_Users_Group at lyralists.lyrasis.org
> http://lyralists.lyrasis.org/mailman/listinfo/archivesspace_users_group
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lyralists.lyrasis.org/pipermail/archivesspace_users_group/attachments/20200507/c99aac41/attachment.html>