CIEL: Duplicate names in 1.11.x for Eparinn & Gratel in 'ht' locale

Tags: #<Tag:0x00007ff4c05f2c80> #<Tag:0x00007ff4c05f26b8>

@akanter, there seems to be a problem with the CIEL for OpenMRS 1.11.x.

Both Eparinn and Gratel have duplicate entries in concept_name table in the ‘ht’ locale. This is causing concept validation errors when trying to export concept packages with the meta data sharing module.

mysql> select concept_id, name, locale, date_created, concept_name_id, uuid, concept_name_type from concept_name where name= “eparinn”; ±-----------±--------±-------±--------------------±----------------±-------------------------------------±------------------+ | concept_id | name | locale | date_created | concept_name_id | uuid | concept_name_type | ±-----------±--------±-------±--------------------±----------------±-------------------------------------±------------------+ | 162305 | Eparinn | ht | 2015-11-05 01:56:15 | 134677 | 134677BBBBBBBBBBBBBBBBBBBBBBBBBBBBBB | FULLY_SPECIFIED | | 77413 | Eparinn | ht | 2015-12-30 19:50:22 | 135018 | 135018BBBBBBBBBBBBBBBBBBBBBBBBBBBBBB | FULLY_SPECIFIED | ±-----------±--------±-------±--------------------±----------------±-------------------------------------±------------------+ 2 rows in set (0.00 sec)

mysql> select concept_id, name, locale, date_created, concept_name_id, uuid, concept_name_type from concept_name where name= “gratel”; ±-----------±--------±-------±--------------------±----------------±-------------------------------------±------------------+ | concept_id | name | locale | date_created | concept_name_id | uuid | concept_name_type | ±-----------±--------±-------±--------------------±----------------±-------------------------------------±------------------+ | 136455 | Gratèl | ht | 2015-11-05 01:40:19 | 134641 | 134641BBBBBBBBBBBBBBBBBBBBBBBBBBBBBB | FULLY_SPECIFIED | | 512 | Gratèl | ht | 2015-11-05 01:46:14 | 134644 | 134644BBBBBBBBBBBBBBBBBBBBBBBBBBBBBB | FULLY_SPECIFIED | ±-----------±--------±-------±--------------------±----------------±-------------------------------------±------------------+ 2 rows in set (0.01 sec)

Full concept_name table

1 Like

@jdegraft, can you please show Andy the specific errors directly in this thread, so that he doesn’t have to click on any links or scroll through any log output?

I got the two examples. It appears that this is an issue where the Haitian term was applied to both the drug and the drug class. I will remove them from the classes.

1 Like

@akanter, would you release a new CIEL version in the near future? Or should @jdegraft manually delete two names/synonyms from the dictionary?

I think I would just retire those concept names from 162305 and 136455

I voided the itching name and changed the heparins name to klas eparinn.

1 Like

@akanter, version 1.11.x of CIEL in the dropbox is still showing the error.

@jdegraft, Andy was saying in the previous messages that he isn’t going to actually do a release with these fixes, but that we should clean them up manually.

He says:

Is it enough to make this change on the mdsbuilder server? (Do it carefully, so as not to change anything else…)

I can remove eparinn and gratel in concept_name in locale ht, but this will effectively remove them in ht. I could also remove the 2nd entry for both, which would leave one entry in ht.

In other words, eparinn refers to itching and rash in ht. I could remove rash which would leave eparinn in ht referring to itching. I would have to try it and see if this works.

This seems to be working.

However, additional duplicates for the record, 120148, 117399 in es

@jdegraft, I understand that

  • on 162305 you should have changed “eparinn” to “klas eparinn”
  • on 136455 you should void the name in the ht locale (yes, this means the concept will have no names in the ht locale)

@darius, sorry I missed that… There are so many duplicates it is difficult to go back. In all cases only one duplicate was removed though. I have left the package traces in the log. I am working on the last package but I should be finished soon.

This is correct. I haven’t done a release since fixing those items as @darius just described. They will be included in the next release which is planned for the beginning of next month.

What other duplicates did you get ?

Here are the CIEL duplicates: CIEL Errors

@raff, these could also be problems with mds validation. What is interesting is after I manually removed the duplicates as best as I could and mds exported the packages successfully with no validation errors, I got an error building referenceapplicationmetadata with the headers of the mds exported packages. Clearly the two modules validate concepts differently.

2 Likes

Thanks for the list

Will fix this for the next release

Let me know if you come up with other errors

Also it would be good to have a list of all the validation rules so we make sure we are running them on our end

@judy, this is a blocker for being able to release Reference Application 2.4. Do you think that you all can do a release soon-ish that has these fixes? (If it were just fixing 2 concept names, I was okay with having James do this directly, but with 10+ changes I don’t feel comfortable releasing a manually-patched version of CIEL.)

@jdegraft, how sure are you that these are all the errors?

I am sure, the exports completed after these errors were resolved.

I think we should get this out this weekend. I am surprised that they haven’t come up before.

Judy, these seem familiar. I think we resolved them before. Not sure how they got back in.

I am investigating the cause of how they got back in

I have a document for you to review (for all the duplicates i found and for the causes )

1 Like