Content

The latest version of FP uses Trino as the query engine, which imply slight changes in the queries' statements.

Known functions(frequent used) change:

Previous	Current	Example
`current_date()`	`current_date`	-
`current_timestamp()`	`current_timestamp`	-
`timestamp('xxxx-xx-xx')`	`timestamp'xxxx-xx-xx'`	-
`date_add (date,INTERVAL expr type)`	`date_add(unit, value, timestamp)`	Change `date_add(current_date(),INTERVAL 1 day)` to `date_add('day',1,current_date)`
`datetime_add(datetime,INTERVAL expr type)`	`date_add(unit, value, timestamp)`	-
`timestamp_add(datetime,INTERVAL expr type)`	`date_add(unit, value, timestamp)`	-
`date_sub(date,INTERVAL expr type)`	`date_add(unit, value, timestamp)`	Change `date_sub(current_date(),INTERVAL 1 day)` to `date_add('day',-1,current_date)`
`timestamp_sub(datetime,INTERVAL expr type)`	`date_add(unit, value, timestamp)`	-
`datediff(date1,date2)`	`date_diff(unit, date1, date2)`	datediff(date1,date2) Change to date_diff('day',date2,date1)
`field=""`	`field=''`	Change `select \* from xx where field1="xy"` to `select \* from xx where field1='xy'`
/	Time fields cannot be compared to strings	Change `on_date >= '2021-12-01'` to `on_date >= date'2021-12-01'` or `on_date >= timestamp'2021-12-01'`
`group by` alias	`group by` origin	Change `select on_date as "day",count(1)as cnt from xx group by day` to `select on_date as "day",count(1)as cnt from xx group by on_date`
`having` alias	`having` origin	Change `select on_date as "day",count(1)as cnt from xx group by on_date having cnt>100` to `select on_date as "day",count(1)as cnt from xx group by on_date having count(1)>100`
`ifnull(expr1, expr2)`	`coalesce(expr1, expr2)`	-
`str_to_date()`	`date_parse()`	-
`float`	`double` or `real`	-
`string`	`varchar`	-
`lag(expr, offset, default)`	`lag(expr, offset)`	-
`lead(expr, offset, default)`	`lead(expr, offset)`	-