<!doctype html>
<html lang="en">
<head>
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>Testing ยท paperless-cli docs</title>
<link rel="stylesheet" href="styles.css">
</head>
<body>
<main class="shell">
<section class="grid">
<aside class="nav">
<h2>Navigate</h2>
<ul>
<li><a href="index.html"><span>Docs home</span><small>overview</small></a></li>
<li><a href="readme.html"><span>README</span><small>surface</small></a></li>
<li><a href="architecture.html"><span>Architecture</span><small>layers</small></a></li>
<li><a href="cli-parity.html"><span>CLI parity</span><small>compatibility</small></a></li>
</ul>
</aside>
<article class="doc">
<span class="eyebrow">Testing</span>
<h1>Coverage aimed at behavior, not ceremony</h1>
<p class="lede">The suite is organized around the contracts most likely to regress: transport shaping, config behavior, output guarantees, TUI rendering, and local PDF utilities.</p>
<h2>Test layers</h2>
<ul>
<li>Config and client: round-trips, redaction, query parameters, HTTP mapping, pagination, uploads, downloads</li>
<li>Output and TUI: Markdown shape, JSON envelope shape, TUI rendering, security reviewer behavior</li>
<li>CLI behavior: help surface, missing-config paths, graceful status, config set-url/set-token, document content, and pdf read/info</li>
</ul>
<h2>Compatibility coverage</h2>
<ul>
<li><code>document content</code> returns only extracted text in the happy path.</li>
<li>Document metadata edits add and remove tags without disturbing unrelated fields.</li>
<li><code>tag edit</code> request shaping stays narrow and predictable.</li>
<li><code>PAPERLESS_URL</code> and <code>PAPERLESS_TOKEN</code> override persisted config correctly.</li>
</ul>
</article>
</section>
</main>
</body>
</html>