after reading https://golang.org/pkg/database/sql/, i'm still not quite sure how to do error handling with respect to tx (transaction) operations. coming from a java/c# background, where once an exception is raised, i know what to do in the "catch" and "finally" blocks, i'm not as sure with go errors, so i thought i'd askfor help being a go newbie.
supposing i'm doing something like this in postgresql within a go func: tx, err := db.Begin() //Question 1: what happens if err != nil in the above statement? will tx be nil or will i get a non-nil tx that is just unuseable? if the latter, would it ever make sense for me to just wait until after the tx.Commit() statement and just do error checking and handling once? i know that if i get an error with db.Begin(), i really should just forget about trying to execute the remaining code but i was just curious what other options -- if any -- might make sense. rows, err := tx.Query(...) result, err := tx.Exec(...) err = tx.Commit() additional questions: (2) kinda related to question 1 above: do i have to check for errors after each statement above, or is it the case that once an error has occurred, the tx object is unuseable because it "remembers" the error(s) that have occurred? (3) when an error occurs, is the transaction automatically rollback'ed or do i have to do this explicitly? (4) if tx.Commit() fails, does the "go" runtime take care of necessary resource cleanups / givebacks, etc? any potential gotchas i need to be aware of? thanks for any help! -- You received this message because you are subscribed to the Google Groups "golang-nuts" group. To unsubscribe from this group and stop receiving emails from it, send an email to golang-nuts+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
