[Archivesspace_Users_Group] curl help

Thu Jan 22 17:56:01 EST 2015

If you have NodeJS installed and want to try a *very* experimental tool,
you could try this too:

$ npm install as-cli -g
$ as-cli setup
> {ENTER YOUR BACKEND URL, USER, PASSWORD}
$ as-cli repositories
> {SELECT YOUR REPOSITORY}
$ as-cli batch-export --dir {A DIRECTORY}

But I stress that this really an unsupported prototype.

Brian

On Wed, Jan 21, 2015 at 1:04 PM, Steven Majewski <sdm7g at virginia.edu> wrote:

>
> Chris Fitzpatrick has a bunch of examples of Ruby code using RestClient
> and the AS backend API
> at:  https://gist.github.com/cfitz
>
> but if you want to do a bulk export of a repo, there is a script in  the
> current release: archivesspace/scripts/ead_export.sh
>
> Usage: export.rb <username> <password> <repository_id>
>
> which will make a zip file of all of the resources exported from a
> repository.
>
> The ruby code is here ( as another example of using the API ):
>
>
> https://github.com/archivesspace/archivesspace/blob/master/launcher/ead_export/lib/ead_export.rb
>
> Before finding cfitz  more thorough delete_repo
> https://gist.github.com/cfitz/4fefcfa659ef18bfbe6c
> I was using this shell script to delete all resources from a repository (
> so we could retest imports
> after code patches ). I’m including it here as another example of
> scripting the API with curl from
> the shell.
>
>
>
> REPO=http://localhost:8089
> #
> REPO_ID=$1
> #
> user=admin
> pass=admin
>
> session=$(curl -s -F password="$pass" "$REPO/users/$user/login" |
>     sed 's|.*"session":"\([a-z0-9]*\)".*|\1|g')
>
> echo "Repo: $REPO"
> echo "Session $session" 1>&2
>
> curl -H "X-ArchivesSpace-Session: $session" "$REPO/repositories/$REPO_ID"
> | jq .
>
> DELETE='-X DELETE'
> for ID in $( curl  -H "X-ArchivesSpace-Session: $session"
> "$REPO/repositories/$REPO_ID/resources?all_ids=true" | tail -1 | tr '[],' '
> ' )
> do
> curl -H "X-ArchivesSpace-Session: $session" $DELETE
> $REPO/repositories/$REPO_ID/resources/$ID | jq .
> done
>
>
>
> I highly recommend jq http://stedolan.github.io/jq/  for pretty printing
> and exploring json output from
> the API.   ‘jq . ‘  (in the example above just does a json pretty print,
> but you can also select particular fields:
>
> R=http://localhost:8089/repositories
>
> ./curl_as_osx admin admin $R | jq '.[].repo_code’       # list repository
> codes
> "uva-sc"
> "RG_3"
> "Legacy"
> "lsc-test"
> "sdm7g-test"
> "uva-hsl"
> "tsd-test"
> "mhm8m-test"
> "uvalaw-test"
> "img7u-test"
> "elg3e-test"
>
> # get repo-id # from repo-code
> ./curl_as_osx admin admin $R | jq '.[0:-1]| map(select( .repo_code ==
> "img7u-test" ))|.[0].uri'
> "/repositories/12"
>
>
> And, although I haven’t tried it, you could pull json from AS with a GET,
> edit a field on the fly with jq, and post the modified
> json back to ArchivesSpace. However, the jq syntax can be difficult.
> Chris’s bulk_update
>  https://gist.github.com/cfitz/9eb97ad3f6dd772f875f
> may be a better way. It does a merge on the ruby hash of the parsed JSON,
> which may be a bit easier to follow than jq syntax.
>
>
> — Steve Majewski
>
>
> On Jan 21, 2015, at 8:57 AM, Ben Goldman <bmg17 at psu.edu> wrote:
>
> Noah, thanks for the feedback. It's good to know I am on the right track.
> I was able to get that to work. And in fact, we were able to build on that
> and perform a mass export of the finding aids, though for some reason only
> 1914 of 1940 resources were exported.
>
> Now I am wondering how I might adjust the script to export the XML file
> with a file name that uses the value of the resource's identifier field
> rather than the ASpace system identifier for the record.
>
> I agree that it would be great to see the documentation on the API/cURL
> expand in the way you outline. I feel there is so much more I could do with
> the API, that I do not fully understand the capabilities here. I just know
> being able to mass export finding aids with one script is pretty damn cool.
>
> Thanks,
> Ben
>
>
> ------------------------------
> *From: *"Noah Huffman" <noah.huffman at duke.edu>
> *To: *"Archivesspace Users Group" <
> archivesspace_users_group at lyralists.lyrasis.org>
> *Sent: *Tuesday, January 20, 2015 9:32:56 AM
> *Subject: *Re: [Archivesspace_Users_Group] curl help
>
> Hi Ben,
>
>
> I’m a curl novice too, but your last example should work to output a
> single EAD file to the current directory.  The only thing I see is that in
> your example you have a single quote to begin the URL and a double quote to
> end it.  Maybe change the double quote at the very end to a single quote,
> or make both double quotes?
>
>
> Something like:
>
>
> curl --output “EADfilename.xml” -H "X-ArchivesSpace-Session: $TOKEN" ‘
> http://aspace1prod.dlt.psu.edu:9089/repositories/3/resource_descriptions/9235.xml?numbered_cs=true’
>
>
> This works for me in Windows Powershell.
>
>
> More generally, I think it would be helpful for noobs like us if the
> ASpace documentation included a short primer (with examples) on how to use
> curl to interact with the API.  This could supplement the existing API
> documentation at:
> http://archivesspace.github.io/archivesspace/doc/file.API.html
>
>
> The ASpace developer screencasts are a good reference if you haven’t seen
> them, particularly this one
> http://youtu.be/iKd4ZME1uIE?list=PLJFitFaE9AY_DDlhl3Kq_vFeX27F1yt6I
>
>
> -Noah
>
>
> ================
> Noah Huffman
> Archivist for Metadata and Encoding
> David M. Rubenstein Rare Book & Manuscript Library
> Duke University
> noah.huffman at duke.edu
> 919-660-5982
> http://library.duke.edu/rubenstein/
>
>
>
>
>
>
>
>
> *From:* archivesspace_users_group-bounces at lyralists.lyrasis.org [
> mailto:archivesspace_users_group-bounces at lyralists.lyrasis.org
> <archivesspace_users_group-bounces at lyralists.lyrasis.org>] *On Behalf Of *Ben
> Goldman
> *Sent:* Monday, January 19, 2015 3:56 PM
> *To:* Archivesspace Users Group
> *Subject:* [Archivesspace_Users_Group] curl help
>
>
> Hey All,
>
>
> Wondering if anyone can help a cURL novice work through the process of
> exporting a finding aid from the API. I've reviewed some of the past emails
> on this subject from Noah and Kevin but I am still hung up somewhere. A
> quick recap:
>
>
> Login:
>
>
>         curl -F password='***'
> http://aspace1prod.dlt.psu.edu:9089/users/bmg17/login
>
>
> Save token:
>
>
>         token=*****
>
>
> Just to verify I know what I am doing, try retrieving a resource record:
>
>
>         curl -H "X-ArchivesSpace-Session: $token" '
> http://aspace1prod.dlt.psu.edu:9089/repositories/3/resources/1352'
>
>
> Success: {"lock_version":0,"title":"Frederick R. Matson
> papers","publish":true,* etc. etc. etc.*
>
> I'm not going to do a mass publish yet, but went ahead and retrieved all
> the ids anyway:
>
>
>        curl -H "X-ArchivesSpace-Session: $token" '
> http://aspace1prod.dlt.psu.edu:9089/repositories/3/resources?all_ids=true'
>
>
> A long list. Including a very short resource record with the id of 9235.
> Now I am wondering exactly how I put this together to export just this one.
> This is the command mentioned in previous messages:
>
>
>         curl --output “resource_#1.xml” -H "X-ArchivesSpace-Session:
> $TOKEN" '
> http://aspace1prod.dlt.psu.edu:9089/repositories/3/resource_descriptions/$ids.xml?numbered_cs=true
> "
>
>
> I assume if I'm just exporting one, I can replace '$ids" with the actual
> number (9235), and that the value following --output can be whatever I want
> the finding aid file name to be. Is all this correct? And where would this
> actually output the file to?
>
>
> Thanks, in advance, for any help.
>
>
> -Ben
>
>
>
> *Ben Goldman*
> Digital Records Archivist
> Penn State University Libraries
> University Park, PA
> 814-863-8333
> http://www.libraries.psu.edu/psul/speccolls.html
>
>
> _______________________________________________
> Archivesspace_Users_Group mailing list
> Archivesspace_Users_Group at lyralists.lyrasis.org
> http://lyralists.lyrasis.org/mailman/listinfo/archivesspace_users_group
>
> _______________________________________________
> Archivesspace_Users_Group mailing list
> Archivesspace_Users_Group at lyralists.lyrasis.org
> http://lyralists.lyrasis.org/mailman/listinfo/archivesspace_users_group
>
>
>
> _______________________________________________
> Archivesspace_Users_Group mailing list
> Archivesspace_Users_Group at lyralists.lyrasis.org
> http://lyralists.lyrasis.org/mailman/listinfo/archivesspace_users_group
>
>

-- 
Brian Hoffman
brianjhoffman at gmail.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lyralists.lyrasis.org/pipermail/archivesspace_users_group/attachments/20150122/56ef80ed/attachment.html>