Collection short names that differ from standard filenames

[houchin]

Hi,

Is there a comprehensive list of collection short names that differ from what is written in the filename?

For example, the PolarWander files all start with "off_USNO-PolarWander-UT1-ANC", but the actual CSN is "USNO-PolarWander-UT1-ANC-Int".

It's easy enough to just strip off the "off_", but I need to either know algorithmically when to add the "-Int" or a list to match what I pull out of the filename.

[houchin]

In absence of an exhaustive list, is it an "official" standard that when files are converted from some external format to an IDPS internal format, that "_Internal" is appended to the filename? If so, I can check for that, and then append "-Int" to the CSN pulled from the beginning of the filename.

[kbisanz]

I talked to a couple people and we're not aware of a list that contains the info you're looking for. It'll basically be all the ANC and AUX products which are named funny.

Yes, if you see _Internal on the file name, you should be able to assume that -Int is on the collection short name. It won't always start with off_ (for official). When it's substitute data, it'll start with sub_. The short name should be the second field (delimited by underscores).

The CDFCB-X Volume 1 contains info on file naming conventions in section 3.4. I believe it is document number D34862.

What's the problem you're trying to solve?

[houchin]

Thanks. That helps. I'll add sub_ and the other prefixes to my code.

The problem we're dealing with is packaging up all of the specific inputs actually used for a specific ADL run to submit to DPE for LUT update verification. The issue that came up today is that the Solar Diffuser requires many outputs, but in the metadata for the Agg History file, none of those inputs are listed. So what I was trying to do is was to go back to the OBC-IP and look at its metadata, then find those files in my segmented DMS.

However, in the IDPS generated OBC-IPs, there is zero directory info in the N_Aux_Filename values, so to avoid doing a "find", I need to take one of those filenames and get the actual collection short name from the file. Thus the issue with the internal formats, as in many cases the filename itself uses the non-internal CSN and appends _Internal to the name instead of using the internal CSN.

[kbisanz]

I understand your problem. Unfortunately I don't have a current solution.

A future solution could be that ADL add ADL-only metadata on outputs which would list the inputs used by the algorithm and their full path on disk. I would recommend feeding that feature request through Paul Meade.