Opened 11 years ago
Closed 11 years ago
#516 closed defect (fixed)
Oracles to be independent of configured SECORE resolver
Reported by: | Piero Campalani | Owned by: | Piero Campalani |
---|---|---|---|
Priority: | major | Milestone: | Future |
Component: | systemtest | Version: | development |
Keywords: | secore resolver | Cc: | |
Complexity: | Medium |
Description
The systemtests for WCS service should not fail if a non-default SECORE resolver is configured. With the new systemtest architecture, this can be possible by using a companion bash script, like is done with the first WCS test on GetCapabilities.
This does not hold for every test, but only for those where the configured resolver is used, that is when compound CRS URIs are built (this happend for the eobstest
dataset, e.g. test n.06).
Change History (5)
comment:1 by , 11 years ago
Owner: | changed from | to
---|---|
Status: | new → assigned |
comment:2 by , 11 years ago
Mh, I'm not sure why I talked about test n.06, anyway looking at the oracle e.g. n.12 (12-compound_crs.kvp.oracle
) one can understand the problem: SECORE URI is part of the oracle and is hence non system-independent.
I can take it, ok to reassign.
comment:3 by , 11 years ago
Actually the oracle comparison code strips off the SECORE_URI from the output and the oracle and then makes the comparison. Does it not work for you?
sed -i 's|'$SECORE_URL'||g' "$output_tmp" "$oracle_tmp"
comment:4 by , 11 years ago
That is for SECORE systemtest, but not for WC*S ones: you have the SECORE URL there as well. I'll just repeat that operation in the other else
branch.
comment:5 by , 11 years ago
Resolution: | → fixed |
---|---|
Status: | assigned → closed |
Not sure what exactly do you mean. Is test n.06 = 06-single_crs_misspelled_key.kvp ?
I'll reassign to you, I think it will be easier for both.