Issue #19857 has been updated by ioquatix (Samuel Williams). We found it was discussed here: https://bugs.ruby-lang.org/issues/19008#note-5 I'm a little embarrassed I didn't consider or fix it at the time. Anyway... Here is the proposed fix: https://github.com/ruby/ruby/pull/8330 Specifically, when an iseq is compiled code, it resets all the coverage to zero for lines that are executable. Instead of resetting to zero, we should leave it as is if it already contains coverage information. ---------------------------------------- Bug #19857: Eval coverage is reset after each `eval`. https://bugs.ruby-lang.org/issues/19857#change-104410 * Author: ioquatix (Samuel Williams) * Status: Open * Priority: Normal * Assignee: ioquatix (Samuel Williams) * Backport: 3.0: UNKNOWN, 3.1: UNKNOWN, 3.2: UNKNOWN ---------------------------------------- It seems like `eval` based coverage is reset every time eval is invoked. ```ruby #!/usr/bin/env ruby require 'coverage' def measure(flag) c = Class.new c.class_eval(<<~RUBY, "foo.rb", 1) def foo(flag) if flag puts "foo" else puts "bar" end end RUBY return c.new.foo(flag) end Coverage.start(lines: true, eval: true) # Depending on the order of these two operations, different computation is calculated, because the evaluation of the code is considered different, even if the content/path is the same. measure(false) measure(true) p Coverage.result ``` Further investigation is required. -- https://bugs.ruby-lang.org/