archives_tools

Python libraries for working with archives data

aspace

A Python library for working with the ArchivesSpace API

• This is a partially complete library I've added to as I write scripts. I set this up to avoid writing requests multiple times and hopefully serve as an example for a community-directed library. There are more functions that are not yet documented. Any suggestions are welcome.

Requirements

• Python 2.7 or Python 3.4+
• ArchivesSpace version 1.5+

Setup

• you can clone the repo, and install with setup.py:

`

git clone https://github.com/UAlbanyArchives/archives_tools
python setup.py install

• aspace can optionally be set up to us local config file called local_settings.cfg which places itself in you Python lib directory. Three values stored there can be edited in the .cfg file manually or by using the library:

`

from archives_tools import aspace as AS
	
AS.setURL("http://localhost:8089")
AS.setUser("admin")
AS.setPassword("admin")

• This prevents you from having to re-enter credentials each time, but you can also pass a tuple with each function

• Keep in mind that these values are stored in plain text in your python library directory

Principles

• Lower the complexity of working with the ArchivesSpace API
• Write commonly used code once, making it faster and easier to write scripts in the future
• Be consistent and transparent about what the library is doing
• Maintain the full functionality of the API in as much as possible

Usage

Authentication

• Once these values are stored, you can connect with ArchivesSpace by using AS.getSession():

`

from archives_tools import aspace as AS

session = AS.getSession()

• As with all functions you can also use a tuple each time:

`

from archives_tools import aspace as AS

loginData = ("http://localhost:8089", "admin", "admin")	
session = AS.getSession(loginData)

Get Data from ArchivesSpace

Using Data Objects

• Data objects returned from ArchivesSpace can use dot syntax (collection.title) but are also a dictionary of native ASpace API data, so you can also use that syntax (collection["title"]) and get a list of fields with collection.keys().

• There are a few functions for viewing the API data:

  • AS.pp(collection) pretty prints the native JSON to the console
  • AS.fields(collection) prints all of the data fields on different lines
  • AS.serializeData(collection, "/output/path/data.json") writes the data to a pretty printed JSON file to the output file path entered as a string

Repositories

• Returns a repository object

`

>>> from archives_tools import aspace as AS
>>> session = AS.getSession()
>>> repositories = AS.getRepositories(session)
>>> print(repositories[0].name)
My Test Repository

Resources

• getResource() returns a single resource object
  • Can iterate through child objects using same keys as JSON API
  • Requires:
    • session
    • login data tuple is optional second argument
    • Repository number as a string

`

from archives_tools import aspace as AS

session = AS.getSession()
repo = "2"

collection = AS.getResource(session, repo, "512")
print (collection.uri)

> /repositories/2/resources/512

for note in collection.notes:
	if note.type == "accessrestrict":
		note.publish = True
		for subnote in note:
			subnote.publish = True

• getResources() returns a list of resource objects, limited by parameter
  • Can iterate through child objects using same keys as JSON API
  • Requires:
    • session
    • Repository number as a string
    • Limiting parameter such as:
      • "all" for all resources (may be problematic for large repositories)
      • A string of comma separated resoruce numbers: "334, 352, 445, 707"
      • A range of the total resources for pagination: "50-100"
        • Note that these are not resource numbers, but the second set of 50 resoruces

`

from archives_tools import aspace as AS

session = AS.getSession()
repo = "2"

for collection in AS.getResources(session, repo, "all"):
	if collection.title == "David Baldus Papers":
		print (collection.ead_id)
		for note in collection.notes:
			if note.type == "abstract":
				print (note.content)

>apap329
>Consists of records from legal scholar and death penalty researcher David C. Baldus whose work informed the 1987 McCleskey v. Kemp decision.

• getTree() returns a object that lists the arrangement of archival objects assigned to a resource object
  • Can be iterated over to get basic archival object information and to retrieve archival object uris to call with getArchObj()
    • Requires: *