'Seed URL' when calling list_seed_uris() – possible API format change

[bllin001]

While attempting to use the list_seed_uris() method, I encountered an error due to a missing "Seed URL" column in the Archive-It seed report CSV. As a result, the seed URIs could not be retrieved using the library. This issue may be due to recent changes or inconsistencies in the Archive-It API response format.

Error Details

• Method Used: aic.list_seed_uris() / aic.return_all_metadata_dict()
• Error Message:

Traceback (most recent call last):
  File "/Users/brianllinas/Library/CloudStorage/OneDrive-OldDominionUniversity/VMASC/Projects/archive-collections/code/test.py", line 30, in <module>
    print(aic.return_all_metadata_dict())
          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/brianllinas/miniforge3/lib/python3.11/site-packages/aiu/archiveit_collection.py", line 575, in return_all_metadata_dict
    self.load_seed_metadata()
  File "/Users/brianllinas/miniforge3/lib/python3.11/site-packages/aiu/archiveit_collection.py", line 376, in load_seed_metadata
    seed_report_metadata = get_seed_metadata_from_seed_report(self.collection_id, self.session)
                           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/brianllinas/miniforge3/lib/python3.11/site-packages/aiu/archiveit_collection.py", line 274, in get_seed_metadata_from_seed_report
    seed_metadata[ row["Seed URL"] ] = {}
                   ~~~^^^^^^^^^^^^
KeyError: 'Seed URL'

Affected Collection (for context)

• Collection ID: 3649
• Collection Name: 2013 Boston Marathon Bombing
• Collection URL: https://archive-it.org/collections/3649

Expected Behavior

The list_seed_uris() method should return a list of seed URLs. If the "Seed URL" field is renamed or omitted in the CSV, the code should either handle it gracefully or raise a clearer error.

[shawnmjones]

Thank you for finding this! Because of API changes, I'm impressed this lasted as long as it did.

I am no longer working on this library and will soon be under some IP constraints, but I can give suggestions.

Try this for now:

1. pip uninstall aiu (or pip3 uninstall aiu depending on your installation)
2. clone the repository
3. comment out lines 376 - 383 in aiu/archiveit_collection.py:

aiu/aiu/archiveit_collection.py

Lines 376 to 383 in a71c586

	seed_report_metadata = get_seed_metadata_from_seed_report(self.collection_id, self.session)
	seed_report_timestamp = datetime.now()
	for uri in seed_report_metadata:
	self.seed_metadata.setdefault("seeds", {})
	self.seed_metadata["seeds"].setdefault(uri, {})
	self.seed_metadata["seeds"][uri].setdefault("seed_report", {})
	self.seed_metadata["seeds"][uri]["seed_report"]= seed_report_metadata[uri]

4. inside the repository root (i.e., the directory on your machine you cloned to): pip install .

I'm not in a position to edit the code at this moment, but let me know if this works for you.

[bllin001]

Thank you, Dr. Jones, for your quick response!

I’m currently working with Dr. Nelson, and I plan to build on your dissertation topic for my own research, focusing on storytelling in Archive-It collections and integrating LLMs for analysis.

I’ll follow your suggestions and let you know if the changes resolve the issue.

Again, thank you so much for your time and support!