Note: The above link is accessible only to members of ASF.
Currently we have an INSAR_ISCE_MULTI_BURST job spec (which takes reference and secondary parameters) and a separate INSAR_ISCE_BURST job spec (which takes a granules parameter) because it was the easiest way to merge the multi-burst work to main without breaking backwards compatibility for INSAR_ISCE_BURST.
For this issue, we'll want to merge the two job specs as follows:
[ ] Add the granules parameter to INSAR_ISCE_MULTI_BURST.yml and enforce that either granules or reference and secondary must be used.
[ ] Rename INSAR_ISCE_MULTI_BURST.yml to INSAR_ISCE_BURST.yml (replacing the existing INSAR_ISCE_BURST.yml)
[ ] Rename the job type to INSAR_ISCE_BURST (first line of the job spec)
[ ] Limit reference and secondary each to length 1 (this can be changed back to 30 in the hyp3-multi-burst-sandbox deployment for testing purposes)
[ ] Remove INSAR_ISCE_MULTI_BURST.yml from hyp3-multi-burst-sandbox deployment
[ ] Currently the validate_jobs function calls convert_single_burst_jobs to convert the parameters for all INSAR_ISCE_BURST jobs from granules to reference and secondary in validation.py. This will need to be updated to only perform the conversion if the job has a granules parameter. Or we could just handle the granules parameter in the actual validator functions that use the burst granule names (currently that should just be the check_same_burst_ids validator), which is probably the better option.
[ ] Rename INSAR_ISCE_MULTI_BURST to INSAR_ISCE_BURST everywhere it's used in the code (see git grep INSAR_ISCE_MULTI_BURST); should just be set_batch_overrides.py, test_set_batch_overrides.py, and test_validation.py
[ ] Add a note to the Changelog that granules is now deprecated for INSAR_ISCE_BURST
Jira: https://asfdaac.atlassian.net/browse/TOOL-3145
Note: The above link is accessible only to members of ASF.
Currently we have an
INSAR_ISCE_MULTI_BURST
job spec (which takesreference
andsecondary
parameters) and a separateINSAR_ISCE_BURST
job spec (which takes agranules
parameter) because it was the easiest way to merge the multi-burst work tomain
without breaking backwards compatibility forINSAR_ISCE_BURST
.For this issue, we'll want to merge the two job specs as follows:
granules
parameter toINSAR_ISCE_MULTI_BURST.yml
and enforce that eithergranules
orreference
andsecondary
must be used.oneOf
keyword (see https://swagger.io/specification/#discriminator-object and https://stackoverflow.com/a/25637397), which would require modifying our job spec parsing hereINSAR_ISCE_MULTI_BURST.yml
toINSAR_ISCE_BURST.yml
(replacing the existingINSAR_ISCE_BURST.yml
)INSAR_ISCE_BURST
(first line of the job spec)reference
andsecondary
each to length1
(this can be changed back to30
in thehyp3-multi-burst-sandbox
deployment for testing purposes)INSAR_ISCE_MULTI_BURST.yml
fromhyp3-multi-burst-sandbox
deploymentvalidate_jobs
function callsconvert_single_burst_jobs
to convert the parameters for allINSAR_ISCE_BURST
jobs fromgranules
toreference
andsecondary
invalidation.py
. This will need to be updated to only perform the conversion if the job has agranules
parameter. Or we could just handle thegranules
parameter in the actual validator functions that use the burst granule names (currently that should just be thecheck_same_burst_ids
validator), which is probably the better option.INSAR_ISCE_MULTI_BURST
toINSAR_ISCE_BURST
everywhere it's used in the code (seegit grep INSAR_ISCE_MULTI_BURST
); should just beset_batch_overrides.py
,test_set_batch_overrides.py
, andtest_validation.py
granules
is now deprecated forINSAR_ISCE_BURST