installing and invoking the MMQA test harness
% $HopeName: MMQA_harness!script:help:problems(trunk.2) $

The MMQA test system consists of the test harness (in hope as MMQA_harness!),
suites of individual tests (MMQA_test_function!, MMQA_test_argerr!,
MMQA_test_conerr!, MMQA_test_misc!), data files (MMQA_data!) and a number of
test sets (MMQA_testsets!). To run tests, you need at least the test harness
and the individual tests you want to run. To checkout everything, use the
hope command:
   checkout -recursive -missing-dir create MMQA!(.)

(There are no useful checkpoints or branches on the MMQA compound.)
This will give you a directory structure as follows:

  mmqa         -- MMQA top-level directory, i.e. MMQA!
    test       -- test harness, i.e. MMQA_harness!
      qa       -- perl script for compiling and running test &c
      script   -- perl programs and other files used by 'qa'
      testlibs -- test libraries (.c and .h files)
      obj      -- compiled tests and libraries go in here
      report   -- test reports (unused)
    function   -- mps functional tests, i.e. MMQA_test_function!
    conerr     -- context error tests, i.e. MMQA_test_conerr!
    argerr     -- argument error tests, i.e. MMQA_test_argerr!
    misc       -- miscellaneous tests, i.e. MMQA_test_misc!
    data       -- data files, i.e. MMQA_data!
    testsets   -- test sets, i.e. MMQA_testsets!

Actually, it doesn't matter where you put the directories of individual
tests (function, conerr, argerr, misc), but it will be simplest if you
use the structure above.

All qa commands are invoked by running the perl script in the mmqa/test
directory, as follows:

 perl mmqa/test/qa <command> <parameters>

Perl is required: if the above fails, try

 perl5 mmqa/test/qa <command> <parameters>

You will probably find it useful to set up an alias, batch file or similar
so that you can just type "qa <command> <parameters>" to invoke qa commands.