From 730c99a7ab64b917f46338f0542f8ba5b70371be Mon Sep 17 00:00:00 2001 From: Mats Wichmann Date: Thu, 25 Jun 2020 06:19:46 -0600 Subject: A few more edits to testing doc Signed-off-by: Mats Wichmann --- testing/framework/test-framework.rst | 17 ++++++++--------- 1 file changed, 8 insertions(+), 9 deletions(-) diff --git a/testing/framework/test-framework.rst b/testing/framework/test-framework.rst index c73afb0..265b82e 100644 --- a/testing/framework/test-framework.rst +++ b/testing/framework/test-framework.rst @@ -528,8 +528,8 @@ Test infrastructure The main test API is defined in the ``TestSCons`` class. ``TestSCons`` is a subclass of ``TestCommon``, which is a subclass of ``TestCmd``. All those classes are defined in Python files of the same name -in ``testing/framework``. Start in ``testing/framework/TestCmd.py`` -for the base API definitions, like how +in ``testing/framework``. +Start in ``testing/framework/TestCmd.py`` for the base API definitions, like how to create files (``test.write()``) and run commands (``test.run()``). Use ``TestSCons`` for the end-to-end tests in ``test``, but use @@ -618,11 +618,10 @@ This leads to a suggestion for E2E test organization because the framework doesn't have a way to indicate a partial skip - if you executed 200 lines of test, then found a condition which caused you to skip the last 20 lines, the whole test is marked as a skip; -it also doesn't have a way to indicate a partial pass. To improve of -this, keep tool tests which don't need the underlying program in -separate files from ones which do - -it is clearer what is going on if we can see in the test results that the -plumbing tests worked but the ones using the underlying program were skipped -rather than seeing all the tests for a tool reported passing (even though -some weren't actually run) or being entirely skipped. +it also doesn't have a way to indicate a partial pass. +To improve on this, keep tool tests which don't need the +underlying program in separate files from ones which do - +that way one can see in the test results that the "plumbing" +tests worked even if the the ones using the underlying program +maybe were skipped. -- cgit v0.12