38 lines
1.3 KiB
Ruby
38 lines
1.3 KiB
Ruby
|
# Test Parfait by parsing it and the tests, and interpreting it.
|
||
|
|
# (Later also run on arm, much like the mains)
|
||
|
|
#
|
||
|
|
# The difference betwen the running parfait and compiled parfait is namespace.
|
||
|
|
#
|
||
|
|
# In the running interpreter (the one that runs the tests/compiler), Parfait is namespaced
|
||
|
|
# to Parfait, because we define Class/Object/Hash/String etc. All that basic stuff that
|
||
|
|
# is not namespaced in ruby and thus would clash in the running interpreter.
|
||
|
|
#
|
||
|
|
# Which leads straight to the point, that in the compiled code parfait is not namespaced.
|
||
|
|
# Parfait files are actually without namespace, and the namespace is wrapped from the
|
||
|
|
# outside, so we are good to just parse parfait files.
|
||
|
|
#
|
||
|
|
# But the tests are defined for the interpreter (also namespaced) and minitest is used,
|
||
|
|
# so we can not just parse the tests. Instead we parse and rewrite the test (removing the
|
||
|
|
# namespace and defining a mini mintest while we are there),
|
||
|
|
# and then write a single main out. And interpret that.
|
||
|
|
# For each test defined in Parfait.
|
||
|
|
require_relative "helper"
|
||
|
|
|
||
|
|
module Parfait
|
||
|
|
|
||
|
|
class InterpretParfait < MiniTest::Test
|
||
|
|
def setup
|
||
|
|
parfait = ["object"]
|
||
|
|
parfait.each do |file|
|
||
|
|
ruby = File.read("./test/parfait/test_#{file}.rb")
|
||
|
|
not_scoped = ParfaitRewriter.new.rewrite(ruby)
|
||
|
|
#puts not_scoped
|
||
|
|
end
|
||
|
|
end
|
||
|
|
|
||
|
|
def test_any
|
||
|
|
|
||
|
|
end
|
||
|
|
end
|
||
|
|
end
|