fill does not work with SQL server #651

Activity

1. mgirlich commented on May 7, 2021

   

   mgirlich

   on May 7, 2021

   Collaborator

   More actions

   Ah, it seems the IGNORE NULLS part should be outside of LAST_VALUE(). Can you check if your code works after you executed the following code

   `last_value_sql.Microsoft SQL Server` <- function(con, x) {
     build_sql("LAST_VALUE(", ident(as.character(x)), ") IGNORE NULLS", con = con)
   }
2. nassuphis commented on May 7, 2021

   

   nassuphis

   on May 7, 2021

   Author

   More actions

   will do + revert.
   the whole dbplyr package is awesomely useful.
   thanks for the all the good work!
3. mgirlich commented on Jul 7, 2021

   

   mgirlich

   on Jul 7, 2021

   Collaborator

   More actions

   Argh, seems like IGNORE NULLS is only supported by Azure SQL Edge
4. mgirlich

   closed this as completedin #677on Dec 9, 2021
5. fabkury commented on Dec 22, 2022

   

   fabkury

   on Dec 22, 2022

   Last edited by fabkury

   More actions

   Same thing happens with Databricks ODBC connections:

   Error in new_result(connection@ptr, statement, immediate) : 
     nanodbc/nanodbc.cpp:1412: 42000: [Simba][Hardy] (80) Syntax or semantic analysis error thrown in server while executing query. Error message from server: org.apache.hive.service.cli.HiveSQLException: Error running query: org.apache.spark.sql.catalyst.parser.ParseException: 
   mismatched input 'IGNORE' expecting {')', ',', '-'}(line 21, pos 38)
   
   SELECT
         `row_id`,
         `pid`,
         `ts`,
         LAST_VALUE(`guia_total_cobrado` IGNORE NULLS) OVER (PARTITION BY `pid` ORDER BY `pid`, `ts`) AS `guia_total_cobrado`,
   --------------------------------------^^^
   

   I was able to make it work by running this prior to fill():

   `last_value_sql.Spark SQL` <- function(con, x) {
     dbplyr:::build_sql("LAST_VALUE(", ident(as.character(x)), ", true)", con = con)
   }