<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Testing on Michael-F-Bryan</title><link>https://adventures.michaelfbryan.com/tags/testing/</link><description>Recent content in Testing on Michael-F-Bryan</description><generator>Hugo</generator><language>en</language><lastBuildDate>Sun, 23 Aug 2026 22:06:27 +0800</lastBuildDate><atom:link href="https://adventures.michaelfbryan.com/tags/testing/index.xml" rel="self" type="application/rss+xml"/><item><title>Run Your Code Generator as a Test</title><link>https://adventures.michaelfbryan.com/posts/run-your-code-generator-as-a-test/</link><pubDate>Sun, 23 Aug 2026 22:05:47 +0800</pubDate><guid>https://adventures.michaelfbryan.com/posts/run-your-code-generator-as-a-test/</guid><description>&lt;p&gt;At some point in a project&amp;rsquo;s life, somebody runs a code generator. Maybe they point a client generator at an OpenAPI document, or feed a &lt;code&gt;.proto&lt;/code&gt; file to &lt;code&gt;protoc&lt;/code&gt;, or write a small script that turns a config file into a lookup table. The output is genuinely useful, so it gets committed, and the project&amp;rsquo;s interfaces improve overnight. Then, being diligent, the author adds a line to the README: &lt;em&gt;&amp;ldquo;After changing the schema, re-run &lt;code&gt;make generate&lt;/code&gt; and commit the result.&amp;rdquo;&lt;/em&gt;&lt;/p&gt;</description></item></channel></rss>