Issue Number | 5003 |
---|---|
Summary | Remove Deprecated Links to Cancer.gov |
Created | 2021-06-30 17:10:05 |
Issue Type | Task |
Submitted By | Englisch, Volker (NIH/NCI) [C] |
Assigned To | Englisch, Volker (NIH/NCI) [C] |
Status | Closed |
Resolved | 2023-02-15 17:00:20 |
Resolution | Fixed |
Path | /home/bkline/backups/jira/ocecdr/issue.293248 |
It appears Publish-Preview is still referencing assets from www.cancer.gov/publishedcontent/ That and other paths are destined to go away in the next couple months.
More details to follow.
The following scripts are responsible for the content listed in the log files (number of occurrences in parenthesis):
GlossaryConceptFull.py (40)
PublishPreview (197)
QCforWord.py (127)
QcReport.py (14)
We need to determine whether the locations being retired are controlled by filters (in which case we might be able to fix this outside of a release) or at least partially in code.
I missed the comment from Blair yesterday that we're only interested in links pointing to the "PublishedContent" directory on Cancer.gov. I had thought that all of those links in the log file were problematic.
I found the following files that need to be modified:
./admin/Inetpub/wwwroot/cgi-bin/cdr/PublishPreview.py
./publishing/pdqdocs/PDQ-summary.html
./publishing/pdqdocs/PDQ-summary.xsl
./server/Filters/CDR0000797103.xml
./server/Filters/CDR0000797104.xml
For this ticket we need filter changes and code changes. The ticket will have to go into Ohm (or a hot-fix if it can't wait that long).
The documents in the directory "publishing/pdqdocs" are sample documents for our PDQ partners in order to see the XSLT transformation in action. There is no need to have an identical copy of the HTML output on Cancer.gov available. I extracted the CSS for some of the major elements from the file nvcg.css and included those in the HTML. The following two files won't need the "PublishedContent"-link going forward:
publishing/pdqdocs/PDQ-summary.html
publishing/pdqdocs/PDQ-summary.xsl
The following filter has been updated to remove the unused CSS (pointing to nvcg.css) and JS code.
CDR797103.xml - Cancer Information Summary for Drupal CMS
https://github.com/NCIOCPL/cdr-server/commit/21ae05e
I don't know why CDR797104.xml slipped onto the list of my earlier comment because I can't confirm that the CSS file nvcg.css is used as part of this filter.
The filter CDR797104.xml was listed because it contained a path including ".../PublishedContent/...". However, this path is only listed within a comment, so there is nothing to change for that filter.
I've modified the PublishPreview.py report on DEV to display the report output for the glossary terms without the resources on the 'PublishedContent' directory.
I will work with Blair next week to check the log files and confirm I didn't miss anything.
Blair notified me that he submitted a ticket to shutdown Percussion. He also "moved" the PublishedContent directory into the Cloud. That means these chances were not necessary because "old" links are still available. However, it is good to remove links to resources not in use anymore.
I want to take another look to see if that move of the PublishedContent directory to the Cloud location will have any other impact for us.
I confirmed that the filter changes have already been merged into the main branch. There is nothing more that needs to be done here.
Closing ticket.
File Name | Posted | User |
---|---|---|
DeprecatedLinks.txt | 2021-06-30 17:12:55 | Englisch, Volker (NIH/NCI) [C] |
Elapsed: 0:00:00.001460