Basic instructions for running tests

diff --git a/t/README b/t/README
new file mode 100644 (file)
index 0000000..9f954f3
--- /dev/null
+++ b/t/README
@@ -0,0 +1,50 @@
+CONFIGURING THE TEST SUITE
+
+Before you can run the test suite, you need to configure certain
+information about the SIP server and the ILS data in the file
+SIPtest.pm.
+
+RUNNING THE TESTS
+
+Every file tests a different protocol transaction.
+Unfortunately, a lot of test cases are missing, but the basics
+are tested, as are most of the simple error conditions (invalid
+users, unknown items, checking in item that's not checked out).
+
+To run a single test, just run
+
+   perl -I.. <file>
+
+If the test fails, the output should be pretty clear about what
+went wrong (assuming you can read raw SIP packets).
+
+To run all the tests, just type
+
+   make test
+
+Right now, that will run tests for functionality that isn't
+supported in the Evergreen environment (the two main cases are
+enable patron and hold management).  To run just the Evergreen tests, use
+
+       make test-openils
+
+which will run just the tests
+
+       00sc_status.t
+       01patron_status.t
+       02patron_info.t
+       03checkout.t
+       06patron_enable.t
+       08checkin.t
+       09renew.t
+       11item_info.t
+       05block_patron.t
+
+NOTE: the Block Patron tests are run last because "Patron Enable"
+isn't supported.  Thus, after running the "Block Patron" test,
+manual intervention is required to unblock the test patron.
+
+The Renew All tests will fail when running the stub "ILS"
+implementation unless there's only one ILS server running.  This
+won't be a problem for any real backend implementation that
+properly manages the database of users and items.