<html>
<head>
<base href="https://llvm.org/bugs/" />
</head>
<body><table border="1" cellspacing="0" cellpadding="8">
<tr>
<th>Bug ID</th>
<td><a class="bz_bug_link
bz_status_NEW "
title="NEW --- - lit substitutions don't work well when a substitution is a prefix of another"
href="https://llvm.org/bugs/show_bug.cgi?id=31237">31237</a>
</td>
</tr>
<tr>
<th>Summary</th>
<td>lit substitutions don't work well when a substitution is a prefix of another
</td>
</tr>
<tr>
<th>Product</th>
<td>Test Suite
</td>
</tr>
<tr>
<th>Version</th>
<td>trunk
</td>
</tr>
<tr>
<th>Hardware</th>
<td>PC
</td>
</tr>
<tr>
<th>OS</th>
<td>Linux
</td>
</tr>
<tr>
<th>Status</th>
<td>NEW
</td>
</tr>
<tr>
<th>Severity</th>
<td>normal
</td>
</tr>
<tr>
<th>Priority</th>
<td>P
</td>
</tr>
<tr>
<th>Component</th>
<td>lit
</td>
</tr>
<tr>
<th>Assignee</th>
<td>unassignedbugs@nondot.org
</td>
</tr>
<tr>
<th>Reporter</th>
<td>dan@su-root.co.uk
</td>
</tr>
<tr>
<th>CC</th>
<td>daniel@zuster.org, llvm-bugs@lists.llvm.org, modocache@gmail.com
</td>
</tr>
<tr>
<th>Classification</th>
<td>Unclassified
</td>
</tr></table>
<p>
<div>
<pre>I was surprised to find today that lit was doing a substitution for me that I
wasn't expecting. It just so happened to be the "right thing" but this
behaviour could really bite us one day.
Concrete example:
Let's say we have
```
config.substitutions.append(('%klee', '/bin/klee'))
```
in our lit configuration file and we have a test that just contains
```
RUN: not echo %klee-stats
```
The test will fail with something like
```
******************** TEST 'KLEE :: regression/x.c' FAILED ********************
Script:
--
not /bin/klee-stats
--
Exit Code: 1
```
We can see that lit expanded the '%klee' in '%klee-stats'. Okay so I guess this
isn't too surprising but what if the lit configuration was this.
```
config.substitutions.append(('%klee', '/bin/klee'))
config.substitutions.append(('%klee-stats', '/bin/klee'))
```
then lit actually does the same thing it expands the RUN line to
```
not echo /bin/klee-stats
```
lit preferred the `%klee` substitution rather than the `%klee-stats` because it
is early in its list of substitutions than `%klee-stats`.
This might not be the behaviour that the user of lit expected.
I'm not sure the best way to handle this but I would suggest emitting a warning
if one of the user supplied substitutions is a prefix of another and the prefix
appears before the longer string in the list of substitutions to apply.</pre>
</div>
</p>
<hr>
<span>You are receiving this mail because:</span>
<ul>
<li>You are on the CC list for the bug.</li>
</ul>
</body>
</html>