Build a support framework for testing if your data can be read by flamingo

[agstephens]

It would be nice if there was a simple way for users to test out if their data could be read by flamingo and made available under the subset WPS in the CEDA WPS UI. Here is a suggestion for how it might work:

1. Create an installation under /gws/smf/j04/cedaproc/ceda-wps-tools that includes:
   • roocs-utils, ceda-wps-assets, clisops, daops, flamingo
2. Write a new tool that would allow a CEDA Staff member to define a simple YAML file and test it to see if clisops has done the right thing with a list of files.
3. Proposed tool would have:
   • Input: YAML file
   • Output: Report to file/terminal of:
     1. Input dataset (before subsetting) - in terms of main dimensions and metadata, min, max
     2. Output dataset (after subsetting) - in terms of main dimensions and metadata, min, max
     3. Outcome: SUCCESS or FAIL
4. Then CEDA staff could say:
   • I want to put dataset X under the CEDA WPS subsetter
   • I write a YAML file
   • I run the test script
   • I get a result
   • If all good, I propose that this is added to the flamingo WPS by use of a nicely crafted github issue template that asks them to fill in all required info for adding the dataset (including test files)

YAML file content The YAML file will need:

• files: list or glob pattern
• subset parameters: dictionary of key-value pairs stating the spatial, temporal and other inputs
• subset tests: one, or more, tests that should be run on the dataset in order to test subsetting works