aspace_pyscripts

Python scripts to support ongoing ArchivesSpace work at Harvard.

The first script is designed to create PUI-side PDF files to be stored in an AWS S3 bucket, so that they can be served from the Harvard's ArchivesSpace Public User Interface, as opposed to generating them on-the-fly.

This work depends heavily on ArchivesSnake.

Functionality that is potentially useful for additional scripts has been factored out into sub-folders.

Also included is a node script useful for periodically running the scripts,

Requirements and Dependencies

These scripts were written with Python 3.6.5

The following python libraries must be installed; they are all available in PyPi:

• ArchivesSnake
• boto3
• SolrClient
• python-magic

To install dependencies, do pip install -r requirements.txt

Important Note:

The scripts in this repository that use ArchivesSnake are relying on there being an .archivessnake.yml configuration file in the home directory of the user. An example configuration file:

baseurl: http://localhost:4567
username: admin
password: admin

Installation

• Make sure you have Python v 3.6.5 or higher installed, with the appropriate PIP tool
• Clone or download this repository as a zip
• Install the required python libraries
• Create .archivessnake with the baseurl, username, and password for your ArchivesSpace instance
• Create pdf_store.yml and s3.yml, using the included templates.

Script: Batch-create and store PDFs

pdfStorer.py runs through one or all repositories, creating a PDF and storing it in an AWS S3 bucket when the Resource is marked as published, and removing the analogous PDF from the S3 bucket if the Resource has been marked as unpublished.

The script uses pickle the json library in the utils/savestate.py script to create a file, pdfs/savedstate.json to keep track of the last time that the PDF was created for the resource. However, if that file is missing, or the targeted resource isn't in it, the code will query the bucket for the last update; the combination of these two mechanisms allows using a cron job to create a new PDF if the resource is subsequently updated.

Use:

python3 pdfStorer.py [-a] [-r {repository_code}] [-t {email_address} -f {email_address}]

Flag	Description
-a	clears the saved state completely. Use this if you want to completely refresh your S3 PDF holdings (for example, if you've changed the PDF output)
-r {repository_code}	For those institutions (like Harvard 😄) that have more than one repository, you may choose to run this script serially for each repository (or just some of them).
-t {email_address}	Address to receive a completion message
-f {email_address}	the "From" address for that completion message

Note that, at the moment, only one instance of the script may run at a time.

Configuration:

This script requires two yaml files: pdf_store.yml (see template), which is expected to be in the same directory as the script, and s3.yml (see template), which can be anywhere, as its path is specified in pdf_store.yml

At the moment, the logging configuration is hard-coded such that the logs will be written to the /logs folder, and have the format pdf_storer_YYYYMMDD.log. This may change in subsequent releases.

In addition, the already_running function in the utilities subpackage, which is used to determine if there already is a pdfStorer process running, assumes that the operating system is linux. Feel free to fork and contribute back!

Script: Identify Users without emails

missing_emails.py runs through ArchivesSpace's entire User table, identifies those user records that have an empty or null email field, and reports out, sorting by Repository.

Use:

python3 missing_emails.py

The results are reported in logs/missing_emails.log.

Configuration:

This script uses ArchivesSnake, so the .archivessnake file (as defined above in the Installation section) must be present.

Script: Report permissions for all or selected users

report_permissions.py reports on the permissions granted to users, by repository. If no usernames are specified, the entire user list will be reported on.

Use:

python3 report_permissions.py   # to report on all users

OR

python3 report_permissions.py "username1,username2,username3"  # a comma-delimited list on those you want reports on

The results are reported in logs/report_permssions.log.*.

Configuration:

This script uses ArchivesSnake, so the .archivessnake file (as defined above in the Installation section) must be present.

Reusable functionality

Handling S3 buckets

The S3 class is used to manage an S3 bucket. An S3 object can be instantiated with keyword arguments as follows:

Key	Value
configpath	The filepath to an s3.yml file
accesskey	The AWS S3 accesskey
secret	The AWS S3 secret
bucket	The name of the bucket to be addressed

If you define configpath, and the s3.yml file is properly filled in, you need not use the other keyword arguments.

Use:

  s3 = s3(configpath = '/my/s3.yml')
  s3.upload('/my/object/path', 'mykey')
  s3.get_object_head('mykey')
  s3.download('mykey', '/my/new/object/path')
  s3.remove('mykey')

Method	Function
initialize	See above
download	downloads the file from the bucket
get_object_head	returns the head_object as defined by AWS
remove	removes the file from the S3 bucket
upload	loads the file into the S3 bucket

Script: Empty out an S3 bucket

bucket_clear.py is a convenience script that empties a given S3 bucket without deleting the actual bucket.

Use:

  python3 bucket_clear.py

If the s3.yml file cannot be found in the directory from which you are running this script, you will be asked for the filepath.

Once the script has found the S3 bucket as defined in the yml file, you will be asked to confirm.

Pickler

The pickler class was abstracted out so that more than one script could use it, rather than endlessly cutting and pasting. The object to be "pickled" is assumed to be a dict datatype.

Use:

  pkl = pickler(filepath)
  if 'foo' not in pkl.obj:
     pkl.obj['foo'] = 'bar'
  pkl.save()

where filepath is the path to the file that will hold/holds the pickled object.

Method	Function
initialize	Opens the file and invokes the load method. the pickled object.
load	loads the pickled object from the file. If the file doesn't exist, the object is created as an empty dict
clear	empties the object
save	pickles the object and writes it to the file (pickle.dump)

One-off scripts

Scripts that have been written for one-off or one-use purposes are included here as exemplars:

• publish_unpublished_dos.py

Utilities

[TBD: a description of the utilities in utils/utils.py ]

Contributors

• Bobbi Fox: @bobbi-SMR (maintainer)
• h/t @helrond of the Rockefeller Archives Center, some of whose ideas in https://github.com/RockefellerArchiveCenter/asExportIncremental I "borrowed".