Updating function to avoid collisions

[icaoberg]

Consider the following payload

{"entity_type": "FILE", "parent_ids": ["21cfe0f6c3368b06237716a6b48f5825"], "file_info": [{"path": "21cfe0f6c3368b06237716a6b48f5825-metadata.tsv", "size": 1092, "checksum": "bf3bae1aeaff3a6010a713380ef5271c", "base_dir": "DATA_UPLOAD"}, {"path": "extras/contributors.tsv", "size": 480, "checksum": "e3bb5373a2875380e0dd2e5aa60c5b2b", "base_dir": "DATA_UPLOAD"}, {"path": "hg19.exonic+intronic/fastq/Puck_200501_25.read1.fastq.gz", "size": 3826680202, "checksum": "eca2160974c2d68850b96a3640154b9d", "base_dir": "DATA_UPLOAD"}]}

If I submit the same payload twice, the entries will appear more than once in db with different UUIDs. I would like to update the code so that each entry must have a unique combination of

• parent_ids
• file_info.path
• file_info.checksum

to help avoid collisions. That is, if there already exists a UUID for an entry with a combination of those 3 fields, then the API should return the existing UUID rather than generating a new one.

cc @pdblood

[icaoberg]

@shirey any updates on this ticket? Has it been implemented? When will it be implemented?