Re: [hibernate-dev] 6.0 - formalized standard functions

For whatever reason JPA and a few of the databases do not categorize coalesce and nullif as functions - hence the "not technically a function" part. However in Hibernate I have defined them as functions for the sake of users being able to plug in custom function templates.

On Tue, May 30, 2017 at 5:14 AM Sanne Grinovero <sanne(a)hibernate.org&gt; wrote: > > Hi Steve, > just a curiosity; for `coalesce` and `nullif` you stated "not > technically a function, but supported as such to allow overriding". > > What does it mean? > > Thanks! > Sanne > > On 29 May 2017 at 22:29, Steve Ebersole <steve(a)hibernate.org&gt; wrote: > > Hibernate has always defined a set of functions that should always be > > available on every Dialect/database. JPA formalized some of these, but > > Hibernate still had a superset. So I wanted to formalize the complete > > set > > of supported functions. Here is the formalized list of what Hibernate > > has > > always considered the "supported list": > > > > - those defined by JPA > > - abs > > - avg > > - concat > > - count > > - current_date > > - current_time > > - current_timestamp > > - length > > - locate > > - lower > > - max > > - min > > - mod > > - sqrt > > - substring > > - sum > > - trim > > - upper > > - Hibernate extensions > > - bit_length > > - coalesce - not technically a function, but supported as such to > > allow overriding > > - cast > > - day - generally defined as `extract(day from ?1)` > > - extract > > - hour - generally defined as `extract(hour from ?1)` > > - minute - generally defined as `extract(minute from ?1)` > > - month - generally defined as `extract(month from ?1)` > > - nullif - not technically a function, but supported as such to > > allow > > overriding > > - second - generally defined as `extract(second from ?1)` > > - str - generally defined as `cast(?1 as CHAR )` > > - year - generally defined as `extract(year from ?1)` > > > > > > Additions/subtractions anyone would like to discuss? > > > > One question I has was in regards to str being defined as `cast(?1 as > > CHAR > > )` by default - specifically using CHAR instead of VARCHAR. Some > > databases > > apply padding for the size of the datatype when casting a numeric to a > > CHAR > > which is not an unreasonable thing to do. > > _______________________________________________ > > hibernate-dev mailing list > > hibernate-dev(a)lists.jboss.org > > https://lists.jboss.org/mailman/listinfo/hibernate-dev