Consider this begin-rescue-ensure block:
attempts=0
begin
make_service_call()
rescue Exception
retry unless attempts>2
exit -1
ensure
attemps += 1
end
If you run that code as it is, it raises an exception because there is no function called 'make_service_call()'. So, it retries. But it would be stuck in infinite loop because the control never goes to 'ensure' because of the 'retry'. Shouldn't 'ensure' part of the block ensure that the code in it gets executed no matter what happens in 'begin' or 'rescue'?
Of course I can increment the count in 'begin' - that's not the point. I am just asking the question about 'ensure' to get some clarity.
Ruby Retry Retry is used primarily in the context of exception handling in Ruby. When your program encounters an exception inside a begin block, control moves to the rescue block where the exception is handled. This is conventionally followed by the execution of the code below the rescue block.
The ensure
section is executed when leaving the begin
statement (by any means) but when you retry
, you're just moving around inside the statement so the ensure section will not be executed.
Try this version of your example to get a better idea of what's going on:
attempts = 0
begin
make_service_call()
rescue Exception
attempts += 1
retry unless attempts > 2
exit -1
ensure
puts "ensure! #{attempts}"
end
ensure
code is executed once, just before the code block exits and it will be called at that time.
But because of the unless attempts>2
condition, and the fact that ensure
will only be called "just before the code exits" (e.g. due to exit -1
) the attempts += 1
will not be executed, and so there is an infinite loop.
ensure
is like __finally
in C++ : you could catch
the exception and then use a goto
: but finally
will not be called until the function is actually about to exit.
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With