Update harvard ordering command

[quevon24]

I added a small change to update the way we get the harvard queryset, which has over ~350k objects.

The original approach could cause memory issues because we get the ~350k clusters and use prefetch_related to get their opinions, if we still wanted to use prefetch_related, we will have to use an iterator with chunk_size param to avoid getting everything.

The new approach is similar to the one bill implemented for columbia ordering, we get the cluster_ids of the clusters that we need to update their opinions, and then use the id to get the opinions and order them.

[sentry-io]

🔍 Existing Issues For Review

Your pull request is modifying functions with the following pre-existing issues:

📄 File: cl/corpus_importer/management/commands/update_opinions_order.py

Function	Unhandled Issue
sort_columbia_opinions	NameError: name 'x' is not defined cl.corpus_impo... Event Count: 1
sort_columbia_opinions	FileNotFoundError: [Errno 2] No such file or directory: '/opt/courtlistener/_columbia/washington/supreme_court_opini... ... Event Count: 1
sort_harvard_opinions	ValueError: chunk_size must be provided when using QuerySet.iterator() after prefetch_related(). cl.corpus_importer.management.command... Event Count: 1

---

_{Did you find this useful? React with a 👍 or 👎}