Error fetching checkpoint for any document causes rest of the batch to fail in deleteSummarizedOps

[arafat-java]

Describe the bug

Presently when using the deleteSummarizedOps.ts when the aggregated data regarding all the unique documentIds is passed, the process iterates through each of those and then adds soft delete markers to deltas (where applicable) and does hard delete of deltas (where applicable). And this is done in a sequential manner for each of the documents which are passed in.
Now the issue is whenever there is an error occurring for any of the documents, all the documents there after are not handled and the loop breaks in between

To Reproduce

Steps to reproduce the behavior:

1. Intentionally delete the content block from one of the DOCUMENTS#tenant#documentId collection item
2. Enable the softDeletionEnabled flag
3. Restart scriptorium
4. See error as follows

Expected behavior

Ideally the error should be logged for the document for which the failure happens and all the subsequent documents deletion should take place

Logs

  "line": {
    "eventName": "LogMessage",
    "exception": {
      "message": "Cannot read properties of null (reading 'scribe')",
      "stack":"TypeError: Cannot read properties of null (reading 'scribe')\n	at deleteSummarizedOps (/usr/src/app/node_modules/@fluidframework/server-services-utils/dist/deleteSummarizedOps.js:27:66)\n	at process.processTicksAndRejections (node:internal/process/task_queues:95:5)",
      "name": "TypeError"
    },
    "message": "Error running deleteSummarizedOps",
    "id": "d5b9ced8-5c0e-4ecf-9518-22c147ea7419",
    "type": "Log",
    "label": "winston",
    "level": "error",
    "properties": "{}",
    "timestamp": "2024-04-05T14:41:41.505Z"
  },

[arafat-java]

This bug is fixed by PR #20576

[arafat-java]

Fixed by PR #20576
Commit c7d68e6