Increase number of lock "tries" by 1.

So that "retry_count" actually means the number of retries.

Users that had set retry_count: 1 to disable retries at all will
need to set the parameter to 0 now.

Fixes #59.

lib/redlock/client.rb

@@ -149,7 +149,7 @@ def recover_from_script_flush
     end
 
     def try_lock_instances(resource, ttl, options)
-      tries = options[:extend] ? 1 : @retry_count
+      tries = options[:extend] ? 1 : (@retry_count + 1)
 
       tries.times do |attempt_number|
         # Wait a random delay before retrying.

spec/client_spec.rb

@@ -113,14 +113,14 @@
         expect(lock_info).to eql(false)
       end
 
-      it 'retries up to \'retry_count\' times' do
+      it 'tries up to \'retry_count\' + 1 times' do
         expect(lock_manager).to receive(:lock_instances).exactly(
-          lock_manager_opts[:retry_count]).times.and_return(false)
+          lock_manager_opts[:retry_count] + 1).times.and_return(false)
         lock_manager.lock(resource_key, ttl)
       end
 
       it 'sleeps in between retries' do
-        expect(lock_manager).to receive(:sleep).exactly(lock_manager_opts[:retry_count] - 1).times
+        expect(lock_manager).to receive(:sleep).exactly(lock_manager_opts[:retry_count]).times
         lock_manager.lock(resource_key, ttl)
       end