<li><a href="#restrictfeature">Restricting tests to feature configurations</a></li>
<li><a href="#mpi">Running tests with MPI</a></li>
<li><a href="#binary">Tests with binary output</a></li>
+ <li><a href="#variants">Tests with multiple comparison files</a></li>
<li><a href="#expect">Changing condition for success</a></li>
<li><a href="#layoutaddtests">Adding new tests</a></li>
<li><a href="#layoutaddcategory">Adding new categories</a></li>
files for different numbers of MPI processes.
</p>
+
<a name="binary"></a>
<h3>Tests with binary output</h3>
<p>
files.
</p>
+
+ <a name="variants"></a>
+ <h3>Tests with multiple comparison files</h3>
+ <p>
+ Sometimes it is necessary to provide multiple comparison files for a
+ single test. Additional comparison files have the same path as the
+ main comparison file (in this case <code>test.output</code>) followed
+ by a dot and a variant description:
+<pre>
+category/test.output
+category/test.output.2
+category/test.output.3
+category/test.output.4
+</pre>
+ The testsuite will try to match the output against all variants in
+ alphabetical order starting with the main output file.
+ </p>
+
+ <p>
+ <b>Warning:</b> This mechanism is only meant as a last resort for
+ tests where <i>all</i> no alternative approach is viable. Especially,
+ consider first to
+ <ol>
+ <li> make the test more robust such that differences can be
+ expressed in round-off errors detectable by numdiff.
+ <li> restrict comparison files to specific feature versions
+ <a href="#restrictfeature">versions of an external feature</a>.
+ </ol>
+ </p>
+
+ <p>
+ <b>Note:</b> The main comparison file (i.e., the one ending in
+ <code>output</code> is mandatory. Otherwise, no test will be
+ configured.
+ </p>
+
+
<a name="expect"></a>
<h3>Changing condition for success</h3>
<p>
</p>
-
<a name="layoutaddtests"></a>
<h3>Adding new tests</h3>