Reputation: 627
I have the following lines of code:
keys = [:a, :b, :c, :d, :e, :f, :g]
if keys.any? do |key|
node[:app][:alphabets][key].empty?
error << " "
end
unless error.empty?
raise error
end
How can I error out what was the empty key that was causing the error?
Upvotes: 2
Views: 701
Reputation: 10546
You can set a proc to the hash that will be executed when a key is not found (don't forget to set the origin value back after using this hash, or to duplicate the original hash):
irb(main):017:0> a = {:a => 1, :b => 2}
irb(main):018:0> a.default_proc = lambda{|h,v| raise Exception.exception(v)}
irb(main):019:0> a[:a]
=> 1
irb(main):020:0> a[:b]
=> 2
irb(main):021:0> a[:c]
Exception: :c
irb(main):022:0> a.default_proc = nil
For your example:
keys = [:a, :b, :c, :d, :e, :f, :g]
alphabets = node[:app][:alphabets].dup
alphabets.default_proc = lambda{|h,v| raise Exception.exception(v)}
if keys.any? {|key| alphabets[key].empty? } then
# do stuff
end
but probably you don't need an Exception anyway:
keys = [:a, :b, :c, :d, :e, :f, :g]
alphabets = node[:app][:alphabets].dup
missing = nil
alphabets.default_proc = lambda{|h,v| missing = v }
if keys.any? {|key| alphabets[key].empty? } then
# do stuff
else
# value inside missing was missing
end
Upvotes: 1
Reputation: 182000
Unless you want to pass a state-changing block to the any?
method, which is bad practice, you'll need to loop over the keys with each
:
keys.each do |key|
error << key if node[:app][:alphabets][key].empty?
end
Upvotes: 2