#SQL

Hi! I have a complex SQL select query which I execute via %ResultSet. It also includes a UNION. The select statement within the union should only be done when an external parameter is set, and I'm not sure of way of doing that within SQL, other than something like

SELECT FOO FROM BAR

WHERE (some conditions)

UNION

SELECT FOO2 FROM BAR2

WHERE :doUnion = 1 AND (some conditions)

...this works, but I don't think it's particularly optimal as I believe it will still execute the second SELECT regardless of :doUnion. It's also sort of hard to read...

When using "IDENTITY" as my primary key, I can select the last inserted ID with

SELECT LAST_IDENTITY() FROM %TSQL_sys.snf;

Actually this is how Hibernate + Iris' Driver acquires the inserted ID when mapping with

@GeneratedValue(strategy = GenerationType.IDENTITY)

Now, considering that I am using the type "SERIAL" as my primary key instead, how can I get the last inserted ID?

Note that with "SERIAL" I can forcefully insert any value for this ID, from which Iris will continue generating values...

Hello,

I have established an ODBC connection with our client based on credentials they provided. The test connection was successful and after using the Link Table wizard I can now see the client's tables and properties. However when we try to execute a query in IRIS we are getting error code -226. I'm trying to determine if this could be just a simple setting problem in IRIS or something with the client.

We have tried with PostgreSQL as well as with TSQL. Same result for each.

In the Inbound SQL Adapter settings, is it possible to specify more than 1 field as the Key Field Name?

Because of the way the Query is being index in Ensemble by the Key Field Name, sometimes transactions get missed and I would like to see if we can add an additional key to the mix to ensure all the transactions are picked up. In this case the InterfaceTrigger is an ID that is auto generated by the table, and I would like to use that as well to ensure we don't miss transactions, and it does not throw any warning messages when it executes the Delete Query.

Does anybody know if there is an easy way to know if the execution of the command $System.SQL.PurgeForTable went well/wrong?

The documentation describes that a string is returned, but in fact there is a nice Quit "" in the code.

Maybe inside PurgeForTable^%apiSQL there is any process variable set when it goes well/wrong?

Thank you very much!

Given a complex method flagged with [ SqlProc ] so it is available as an SQL stored procedure, what's the best way to report a non-system error detected in that method - say, for example, an error %Status - so that the SQL query calling it fails descriptively? Is it best to create and throw an exception, or are there special % variables involved (like in a trigger)? I haven't been able to find an answer in the documentation.

Thanks in advance!

I have an SQL text index defined like this

Index TextIndex On (Text) As %iFind.Index.Basic(IGNOREPUNCTUATION = 1, INDEXOPTION = 0, LANGUAGE = "en", LOWER = 1);

If I run a query like this:

SELECT
Text
FROM Post
WHERE %ID %FIND search_index("TextIndex",'ABC')

I get 20 results with posts containing the term ABC.

But each post can contain the term ABC several times.

Hi folks!

Do we have an embedded REST API in IRIS which accepts arbitrary SQL and returns data in JSON?

Or if anyone did something like this?

The idea is to have something like MDX2JSON but SQL2JSON.

Thanks!

Hello community,

I need to know what permissions or resources, apart from read for the tables, a user would need to view the catalog via ODBC? The ODBC user can run queries sucessfully but cannot get the catalog from the DB via ODBC. Whereas a a user with %All can fetch the catalog.

Any help would be much appreciated.

Regards,

Juergen

Hi All, I am a Data Architect and am trying a method of deploying alter statements onto Cache DB.

My Data Modeling tool generates alters as this.. So question is if the Cache has a way to rename table? If so what is it?

ALTER TABLE <SchemaName>.GROUP_TYPE_NODES RENAME TO GROUP_TYPE_10082019140110000
;

This is giving error.

Hi,

I've a Service utilising the Adapter EnsLib.SQL.InboundAdapter, which uses a Credentials item set with the details of a local SQL account. This currently works, however, we're looking to use the credentials of an AD domain account.

The domain account is a member of an AD security group, which has the required permissions on the source SQL database. I've checked that access is possible with this account via SQL studio.

Hey Intersystems Community-Member,

Following Issue: When I update a SQL table like (update OnlineV3Admin.ParameterApp set popuptext = '' where ID = '1') and then read it for my Json WebService there is a /"x00" in my Json.

What is the best option to disable that? I need here a empty String and not /x00.

Any ideas?

Best regards,

Florian Hansmann

In MSSQL I think you can do something like this:

select *
from HS_IHE_ATNA_Repository.COLUMNS
where TABLE_NAME='Aggregation'

How can I do this in Cache SQL?

[%msg: < Table 'HS_IHE_ATNA_REPOSITORY.COLUMNS' not found>]

Good Morning,

I have this question for a long time. When I make an insert by SQL from an external application in the fields of type %String, if they are empty, it writes the character $c(0) in the global.

Is there a way so that if an Insert is received in a field of type% String with null value instead of the $ c (0) it leaves it empty?: