Flag to decide whether to sanitise the output before comparison of test results with expected results (see sanitise method).
Flag to decide whether to sanitise the output before comparison of test results with expected results (see sanitise method). Default is true; override with false if you want actual results compared.
Make tests that process the files in path.
Make tests that process the files in path. name is an identifying
name for this set of tests. All files whose names end in srcext are
processed. Processing is done by the function compile which must
return either Some (s) where s is the output or None if processing
failed. If srcext is .x and resext is .y, then the expected result
for foo.x is found in file foo.y. If optinext is Some (z), then
foo.z is used for standard input, if it exists, otherwise the string
indefault is used. A test fails if either the processing fails or
it succeeds with the wrong result. argslist is used to specify the
sets of command-line arguments that you want to use. Each test is
run with each set of arguments. The default is an empty argument list.
Whether this is a test run or not.
Whether this is a test run or not. Test runs generate all of their code using a single emitter so it can be easily compared to what we expect. A normal run sends the code for each class to a separate file so that they can be compiled by Jasmin.
Process the source tree by analysing it to check for semantic errors.
Process the source tree by analysing it to check for semantic errors. If any messages are produced, print them. If all is ok, translate the program and generate code for the translation. Return whether or not the semantic analysis checks were passed.
(Changed in version 2.9.0) The p0 call-by-name arguments is evaluated at most once per constructed Parser object, instead of on every need that arises during parsing.
Sanitise the output.
Sanitise the output. At the moment this means make any Windows line endings appear in Unix style instead. This allows for either program or test output to use either line ending style, but tests will still pass. This will clearly break any tests where the actual line endings matter.
This expect method has been deprecated. Please replace all invocations of expect with an identical invocation of expectResult instead.
This expect method has been deprecated. Please replace all invocations of expect with an identical invocation of expectResult instead.
(Since version 2.10.0) lastNoSuccess was not thread-safe and will be removed in 2.11.0
(Since version 2.10.0) lastNoSuccess was not thread-safe and will be removed in 2.11.0
Tests that check that the code generator produces the expected byte code.