CDR Tickets

Issue Number 3882
Summary Modify Summary Filter to adjust URLs for NVCG
Created 2015-03-16 18:28:15
Issue Type Task
Submitted By Englisch, Volker (NIH/NCI) [C]
Assigned To Englisch, Volker (NIH/NCI) [C]
Status Closed
Resolved 2015-06-05 17:58:08
Resolution Fixed
Path /home/bkline/backups/jira/ocecdr/issue.156960
Description

The CDR summary filter creates links to the PDQ boards. These links contain the path 'cancertopics' which will go away on NVCG.
We will need to modify the filters in order to avoid broken links for the summaries.

Comment entered 2015-06-02 17:28:13 by Englisch, Volker (NIH/NCI) [C]
Comment entered 2015-06-02 18:01:49 by Englisch, Volker (NIH/NCI) [C]
Comment entered 2015-06-02 18:22:44 by Englisch, Volker (NIH/NCI) [C]

Note to myself:
These changes will pretty much require every summary to be reprocessed when the filters go in production.

Comment entered 2015-06-05 15:49:30 by Englisch, Volker (NIH/NCI) [C]

Another note:
It is not clear yet if the Spanish genetics link will need to be modified or if it should continue to point to the English document.

Comment entered 2015-06-05 17:23:52 by Englisch, Volker (NIH/NCI) [C]

I'm not sure anymore if the change for the DrugInfo summaries for the clinical trials search needs to be modified:
http://www.cancer.gov/Search/ClinicalTrialsLink.aspx?id=43130&idtype=1

I'll need to find out the format for the link within NVCG.

Comment entered 2015-06-05 17:51:10 by Englisch, Volker (NIH/NCI) [C]

I learned that there exists a ClinicalTrialsLink handler meaning that any URL containing the string ClinicalTrialsLink.aspx will correctly resolve to the clinical trials search results. This means there are no changes necessary for this link.

This is ready for review on DEV.

Comment entered 2015-06-09 17:27:34 by Juthe, Robin (NIH/NCI) [E]

I tried running Pub Preview for a couple of summaries (62908 and 256714) on DEV to check some of the URLs but I'm getting an error:

CDRPreview web service error: Xml data validation error,The required attribute 'nct_id' is missing.Validation error occurred when validating the instance document.,928,35

CDRPreview web service error: Xml data validation error,The required attribute 'nct_id' is missing.Validation error occurred when validating the instance document.,49,313

Comment entered 2015-06-10 11:00:51 by Englisch, Volker (NIH/NCI) [C]

Bob, are the new filters to add the nct_id attribute not installed on DEV?

Robin, it appears that you're getting this error for summaries that contain a ProtocolRef element. We have 164 of those, so you have a 1 in 4 chance to pick the "wrong" one. :-)

The problem is that the CTRP changes require a new mandatory attribute for the ProtocolRef element and it seems not everything has been set up properly on the DEV tier.

Comment entered 2015-06-10 11:02:19 by Juthe, Robin (NIH/NCI) [E]

Well that's just my luck! 🙂 I'll try a few more.

Comment entered 2015-06-10 11:19:47 by Kline, Bob (NIH/NCI) [C]

That's right. My marching orders were to move directly from Blair's VM to QA. Perhaps we should bring DEV into the fold. :-) Don't know what that would mess up of what you and Alan are working on.

Comment entered 2015-06-10 11:26:05 by Englisch, Volker (NIH/NCI) [C]

If your changes are in subversion I can update my sandbox and include your changes or we'll have to restore the DTD on DEV.
Marching order or not, it makes testing difficult when a bunch of the documents fail due to validation errors.

Comment entered 2015-06-10 11:32:43 by Englisch, Volker (NIH/NCI) [C]

Well that's just my luck!

I could give you the list of summaries not to try, Robin.

Comment entered 2015-06-10 11:56:06 by Kline, Bob (NIH/NCI) [C]

The changes are in Ampere (they're the three filters for which I gave you the diffs to review on Monday). The fourth modified filter is the one for the import of the trials. Doesn't make sense to install the changes for that filter without the changes to the download/import software (which I don't think will work until the DEV tier has been migrated out of its funky subnet, unless we talk CBIIT into allowing the server to talk to CTRP). Go ahead and install the other three. Or I can do it if you prefer.

Comment entered 2015-06-18 17:56:57 by Englisch, Volker (NIH/NCI) [C]

The following filters have already been updated for other stories:

  • R13191: CDR000134.xml (Vendor Filter: Term)

  • R13221: CDR000157.xml (Vendor Filter: Summary)

These filters have been updated in subversion:

  • R13224: CDR505580.xml (Vendor Filter: DrugInfoSummary)

  • R13224: CDR502087.xml (CTGov Export Filter)

  • R13224: CDR685090.xml (CTRP Export Filter)

Comment entered 2015-06-18 17:59:17 by Englisch, Volker (NIH/NCI) [C]

This filters have been updated in subversion:

  • R:13225: CDR315588.xml (Module: Vendor Cleanup Templates)

Comment entered 2015-06-25 09:12:48 by Juthe, Robin (NIH/NCI) [E]

The URL for the Spanish Adult Treatment Board page has an extra "s" after "tratamiento", so I'm getting a page not found error.

http://www.cancer.gov/espanol/publicaciones/pdq/consejos-editoriales/tratamientos-adultos

Comment entered 2015-06-25 09:42:02 by Juthe, Robin (NIH/NCI) [E]

The URL for the Spanish Pediatric Treatment Board has an extra "s" after both "tratamiento" and "pediatrico".

http://www-qa.cancer.gov/espanol/publicaciones/pdq/consejos-editoriales/tratamientos-pediatricos

Comment entered 2015-06-25 11:59:17 by Englisch, Volker (NIH/NCI) [C]

Bummer!!!
We have a problem with our versions for the summary filter. The version of the filter CDR157 that's in the CDR is not the same as what's in subversion. I'll have to do a little compare and adjust.

Comment entered 2015-06-25 17:33:23 by Englisch, Volker (NIH/NCI) [C]

The URLs for the Spanish board pages have been updated and tested.

Comment entered 2015-06-30 12:13:28 by Englisch, Volker (NIH/NCI) [C]

I've updated the filter to fix the Spanish URLs and ran another publishing job. The URLs were updated correctly and there were no other, unexpected changes.

  • R13252: CDR000157 - Vendor Filter: Summary

Comment entered 2015-06-30 13:06:44 by Beckwith, Margaret (NIH/NCI) [E]

I have checked several documents in Pub Preview (Adult and Ped. treatment and Supportive Care, HP and Patient), and the links in the About This PDQ Summary section to the Board pages all seem to be working correctly. The only issue I noticed (and Robin J confirmed) is that for the Adult and Ped. HP summaries, the link to the LOE document creates a Python error. It works fine for other document types. I think we can say this is verified in QA.

Comment entered 2015-06-30 16:24:35 by Kline, Bob (NIH/NCI) [C]

... the link to the LOE document creates a Python error.

So this comment from OCECDR-3793 (Robin, June 18) no longer holds true?

I have confirmed that we are no longer receiving the LOE pop-up box with an error message.

Comment entered 2015-06-30 17:15:50 by Beckwith, Margaret (NIH/NCI) [E]

Apparently not. This is the error I am seeing:

A problem occurred in a Python script.
D:\cdr\Log\tmp1iupi8.html contains the description of this error.

Not sure if it is the same one, but it only occurs with the links to the LOE summary for the treatment summaries.

Comment entered 2015-06-30 17:31:49 by Kline, Bob (NIH/NCI) [C]

I'm not able to reproduce the problem. Can you tell me the exact steps you're using when it hits?

Comment entered 2015-06-30 17:36:09 by Beckwith, Margaret (NIH/NCI) [E]

Run Publish Preview for an Adult or Pediatric HP summary. Go to About this PDQ Summary section. Click on link that says "formal evidence ranking system". It is supposed to go to the appropriate LOE summary for that Board.

Comment entered 2015-06-30 17:43:20 by Beckwith, Margaret (NIH/NCI) [E]

Victoria reminded me that the pop-up that you might be referring to is when you click on an LOE ref in a summary document and you get the pop-up definition (which comes from the glossary). A completely different thing than what I was referring to.

Comment entered 2015-07-01 11:06:52 by Englisch, Volker (NIH/NCI) [C]

The problem related to the Python error is a Gatekeeper issue.
Currently, the summary content on Gatekeeper is still created by pages/sections. Each page is created as an object in Percussion and it just happened that somebody decided to name one of the summary sections ‘Summary’ which is also the name of the document type used in Percussion. Now, there exists an object named ‘Summary’ indicating the document itself and when the object/page ‘Summary’ is getting created for the section, Percussion doesn’t like it because it would contain two objects in the same directory with the same name.
I renamed the section to ‘VE Summary’ and now all is well.

I've talked to Aarti to confirm that this error will be resolved with the Feline release because future releases will not handle summaries by pages anymore.

Comment entered 2015-09-04 12:54:27 by Englisch, Volker (NIH/NCI) [C]

Many summaries have already been published to Cancer.gov with the updated URLs. These are all English treatment summaries and most of the Spanish, adult treatment summaries. The other summaries (suppotive care, genetics, etc) will be updated over the weekend,

Please confirm, for instance, by hovering over the link PDQ ... Editorial Board:

Comment entered 2015-09-18 11:34:17 by Englisch, Volker (NIH/NCI) [C]

Can this Curie ticket be closed or is there anything left to be done?

Comment entered 2015-09-18 11:40:14 by Juthe, Robin (NIH/NCI) [E]

Yes, this can be closed. I've verified the new links on PROD.

Elapsed: 0:00:00.001315