QA is mistakenly complaining about unexpected calibrationInformation datasets

[gshiroma]

Not sure if #9 addresses this issue, but the current GCOV QA code seems to be erroneously listing some calibrationInformation LUTs as "UNEXPECTED XML DATASET PRESENT IN HDF5:" in the log file. However these datasets are present in the GCOV specs.

/science/LSAR/GCOV/metadata/calibrationInformation/frequency{freq}/{pol}/rfiLikelihood
/science/LSAR/GCOV/metadata/calibrationInformation/frequency{freq}/elevationAntennaPattern/{pol}
/science/LSAR/GCOV/metadata/calibrationInformation/frequency{freq}/noiseEquivalentBackscatter/{pol}

Such as:

/science/LSAR/GCOV/metadata/calibrationInformation/frequencyA/HH/rfiLikelihood
/science/LSAR/GCOV/metadata/calibrationInformation/frequencyA/elevationAntennaPattern/HH
/science/LSAR/GCOV/metadata/calibrationInformation/frequencyA/noiseEquivalentBackscatter/HH

GSLC and RSLC QA may be presenting the same problem.

[gmgunter]

Thanks for the report, @gshiroma.

I don't think this will have been resolved by #9 -- this seems to be related to the XML spec validation, which is a separate step in the QA workflow.

Would you mind pointing us to an example GCOV product that we could run QA on to reproduce the spurious warnings? Maybe @Tyler-g-hudson could take a look since he implemented that part of the QA code.

[nemo794]

Actually, this is a known issue that occurs on all GCOV datasets. The issue has been documented on the internal QA repo, but I don't think the proposed solutions are robust enough to avoid false negatives from occurring.

No need for further example products nor for @Tyler-g-hudson to look into the issue further at this time. Thanks!

[gshiroma]

Thank you for taking a look a this, @gmgunter ! I can provide a GCOV product, but any GCOV product generated with ISCE3 R4.0.6 (or even R4.0.4) should include these datasets.

[gshiroma]

I wrote my comment at the same time as @nemo794 lol. Thank you, @nemo794 , for also looking at this issue.