| Issue Number | 4697 |
|---|---|
| Summary | [Summary] Markup in modules not showing in linked summaries in RLSO report |
| Created | 2019-11-04 12:43:09 |
| Issue Type | Bug |
| Submitted By | Osei-Poku, William (NIH/NCI) [C] |
| Assigned To | Englisch, Volker (NIH/NCI) [C] |
| Status | Closed |
| Resolved | 2020-01-30 14:00:19 |
| Resolution | Fixed |
| Path | /home/bkline/backups/jira/ocecdr/issue.252020 |
Crystal reported that in the RLSO report for the HP summaries proposed/approved text in the modules is not showing up in the linked summaries. The proposed and approved text does appear in the RLSO for the modules but not in the main summary.
I confirmed this by looking at the Genomics of Wilms Tumor module (777841) in the Wilms Tumor summary (62789). There is proposed mark up in the module that shows on the RLSO report for the module, but does not show when you run the RLSO report for the whole summary.
I put William as the assignee on this, but realized I really meant to have him be the owner since Robin is out of town. I added Robin and Crystal as watchers. Who should this be assigned to? (I wasn't sure what the component was.)
I re-assigned to ~volker and changed the reporter from ~mbeckwit to me.
We just modified the denormalization filter on Friday to only pull in the publishable version of a module in order to prevent text not yet versioned to make it to Cancer.gov.
We may need to explore different denormalization paths depending on the end result.
The isQC flag parameter will be applied to all QC reports in Kepler. Then Volker will make filter changes at a later time.
The following filter has been copied to QA:
CDR0000712005.xml: Denormalization Filter: Summary Module
I ran diff reports for all of the summary QC report and vendor publishing output and the documents were identical.
~oseipokuw, please go ahead preparing documents on QA for additional testing.
We've tested the changes and they look good on QA. Thanks!
~oseipokuw, I just wanted to confirm that you did test the different paths this filter change provides. I can see some changes to the document CDR62792 (Childhood Rhabdomyosarcoma Treatment) where some marked-up insertion text had been added to both, the parent document as well as the module document (CDR777839 - Genomics of Rhabdomyosarcoma). This is a useful test but doesn't address the problem this ticket has been requested to solve. In order to test that everything is working correctly you should also have text changes in the current working document for both, the parent as well as the module document.
This brings up another item that should be clearly understood because I am certain it will trigger questions in the future. As we had decided we've made changes to denormalize the modules and are picking up the publishable version for publishing jobs and the CWD for QC reports. This is the default and what one would expect in the general case. However, there can be situations when the output shown might not be what is expected. Here are some cases to consider:
The user opens a parent document and runs a QC report
--> The QC report displays the CWD of the parent and the CWD of the
module
The user opens a version (or publishable version) of a parent
document and runs a QC report
--> The QC report displays the (publishable) version of the parent
and the CWD of the module
The user opens a parent document and runs a PP report
--> The QC report displays the current (opened) version of the parent
and the publishable version of the module
Please let me know if you've already covered the scenarios mentioned with your testing and I'll push the changes.
Thanks ~volker Yes, we have reviewed all these use cases and that is exactly what we expect to happen. Everything looks good.
The following filter has been copied to PROD and STAGE:
CDR0000712005.xml: Denormalization Filter: Summary Module
https://github.com/NCIOCPL/cdr-server/commit/ce069cc
Verified on PROD. Thanks ~volker.
Elapsed: 0:00:00.001725