Allow multi data file granules

[lisakaser]

Question: What needs to be in the UMM-G size? (Amy)

Example data set?

We could check how this is currently handled in existing UMM-G

Type in CNM message might have to change

[lisakaser]

Specific questions for @afitzgerrell blocking this ticket: What is the file size that needs to go into the UMM-G for multi science file granules?

[afitzgerrell]

The file size reported in UMM-G (SizeInBytes) would be the sum of all the data files comprising the multi-file granule (i.e., everything packaged in a granule that is not the external, file-level umm-g metadata file, browse image, or ancillary files like QA and/or PH files). I don't recall ever receiving QA and PH files with a data producer-delivered data set). There's an example of this here in the UMM-G documentation, but I've attached a visual here in case helpful.

To state this another way as: the umm-g SizeInBytes should be the sum of all the files in a product where the files are defined as "type": "data" in the CNM.

[lisakaser]

SIPSmetgen has a common file name and what file types are allowed as addition to a multifile granule.

[lisakaser]

In this story only deal with additional science data; A separate ticket should take care of ancillary files.

[juliacollins]

The requirements of this issue should be satisfied by issue-61. Currently metgenc can handle multiple data files and multiple browse file, but it does this by assuming anything not a browse file is a data file. The logic will need some refinement when we tackle ancillary file support. Once issue-61 is ready for Ops testing I'll move this issue into testing as well.

[juliacollins]

Update! The definition of "common file name" was clarified during our Slack conversation on March 5, and it is not what's implemented by issue-61! The remaining work for this story is to refine the algorithm/regex used to collect up all files related to a single granule.

[lisakaser]

Work started. 2SP left from original 3SP issue.