1036 lines
28 KiB
HTML
1036 lines
28 KiB
HTML
|
||
<!doctype html>
|
||
<html lang="de" class="no-js">
|
||
<head>
|
||
|
||
<meta charset="utf-8">
|
||
<meta name="viewport" content="width=device-width,initial-scale=1">
|
||
|
||
|
||
|
||
<link rel="canonical" href="https://docs.gwoe.toppyr.de/adr/0003-citation-property-tests/">
|
||
|
||
|
||
<link rel="prev" href="../0002-adapter-architecture/">
|
||
|
||
|
||
<link rel="next" href="../0004-deployment-workflow/">
|
||
|
||
|
||
|
||
|
||
|
||
<link rel="icon" href="../../assets/images/favicon.png">
|
||
<meta name="generator" content="mkdocs-1.6.1, mkdocs-material-9.7.6">
|
||
|
||
|
||
|
||
<title>0003 Citation-Property-Tests - GWÖ-Antragsprüfer Docs</title>
|
||
|
||
|
||
|
||
<link rel="stylesheet" href="../../assets/stylesheets/main.484c7ddc.min.css">
|
||
|
||
|
||
<link rel="stylesheet" href="../../assets/stylesheets/palette.ab4e12ef.min.css">
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<link rel="preconnect" href="https://fonts.gstatic.com" crossorigin>
|
||
<link rel="stylesheet" href="https://fonts.googleapis.com/css?family=Roboto:300,300i,400,400i,700,700i%7CRoboto+Mono:400,400i,700,700i&display=fallback">
|
||
<style>:root{--md-text-font:"Roboto";--md-code-font:"Roboto Mono"}</style>
|
||
|
||
|
||
|
||
<script>__md_scope=new URL("../..",location),__md_hash=e=>[...e].reduce(((e,_)=>(e<<5)-e+_.charCodeAt(0)),0),__md_get=(e,_=localStorage,t=__md_scope)=>JSON.parse(_.getItem(t.pathname+"."+e)),__md_set=(e,_,t=localStorage,a=__md_scope)=>{try{t.setItem(a.pathname+"."+e,JSON.stringify(_))}catch(e){}}</script>
|
||
|
||
|
||
|
||
|
||
|
||
</head>
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<body dir="ltr" data-md-color-scheme="default" data-md-color-primary="teal" data-md-color-accent="light-green">
|
||
|
||
|
||
<input class="md-toggle" data-md-toggle="drawer" type="checkbox" id="__drawer" autocomplete="off">
|
||
<input class="md-toggle" data-md-toggle="search" type="checkbox" id="__search" autocomplete="off">
|
||
<label class="md-overlay" for="__drawer"></label>
|
||
<div data-md-component="skip">
|
||
|
||
|
||
<a href="#0003-sub-d-property-verification-zitate-als-substring-der-zitierten-pdf-seite" class="md-skip">
|
||
Zum Inhalt
|
||
</a>
|
||
|
||
</div>
|
||
<div data-md-component="announce">
|
||
|
||
</div>
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<header class="md-header md-header--shadow" data-md-component="header">
|
||
<nav class="md-header__inner md-grid" aria-label="Kopfzeile">
|
||
<a href="../.." title="GWÖ-Antragsprüfer Docs" class="md-header__button md-logo" aria-label="GWÖ-Antragsprüfer Docs" data-md-component="logo">
|
||
|
||
|
||
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"><path d="M12 8a3 3 0 0 0 3-3 3 3 0 0 0-3-3 3 3 0 0 0-3 3 3 3 0 0 0 3 3m0 3.54C9.64 9.35 6.5 8 3 8v11c3.5 0 6.64 1.35 9 3.54 2.36-2.19 5.5-3.54 9-3.54V8c-3.5 0-6.64 1.35-9 3.54"/></svg>
|
||
|
||
</a>
|
||
<label class="md-header__button md-icon" for="__drawer">
|
||
|
||
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"><path d="M3 6h18v2H3zm0 5h18v2H3zm0 5h18v2H3z"/></svg>
|
||
</label>
|
||
<div class="md-header__title" data-md-component="header-title">
|
||
<div class="md-header__ellipsis">
|
||
<div class="md-header__topic">
|
||
<span class="md-ellipsis">
|
||
GWÖ-Antragsprüfer Docs
|
||
</span>
|
||
</div>
|
||
<div class="md-header__topic" data-md-component="header-topic">
|
||
<span class="md-ellipsis">
|
||
|
||
0003 Citation-Property-Tests
|
||
|
||
</span>
|
||
</div>
|
||
</div>
|
||
</div>
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<label class="md-header__button md-icon" for="__search">
|
||
|
||
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"><path d="M9.5 3A6.5 6.5 0 0 1 16 9.5c0 1.61-.59 3.09-1.56 4.23l.27.27h.79l5 5-1.5 1.5-5-5v-.79l-.27-.27A6.52 6.52 0 0 1 9.5 16 6.5 6.5 0 0 1 3 9.5 6.5 6.5 0 0 1 9.5 3m0 2C7 5 5 7 5 9.5S7 14 9.5 14 14 12 14 9.5 12 5 9.5 5"/></svg>
|
||
</label>
|
||
<div class="md-search" data-md-component="search" role="dialog">
|
||
<label class="md-search__overlay" for="__search"></label>
|
||
<div class="md-search__inner" role="search">
|
||
<form class="md-search__form" name="search">
|
||
<input type="text" class="md-search__input" name="query" aria-label="Suche" placeholder="Suche" autocapitalize="off" autocorrect="off" autocomplete="off" spellcheck="false" data-md-component="search-query" required>
|
||
<label class="md-search__icon md-icon" for="__search">
|
||
|
||
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"><path d="M9.5 3A6.5 6.5 0 0 1 16 9.5c0 1.61-.59 3.09-1.56 4.23l.27.27h.79l5 5-1.5 1.5-5-5v-.79l-.27-.27A6.52 6.52 0 0 1 9.5 16 6.5 6.5 0 0 1 3 9.5 6.5 6.5 0 0 1 9.5 3m0 2C7 5 5 7 5 9.5S7 14 9.5 14 14 12 14 9.5 12 5 9.5 5"/></svg>
|
||
|
||
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"><path d="M20 11v2H8l5.5 5.5-1.42 1.42L4.16 12l7.92-7.92L13.5 5.5 8 11z"/></svg>
|
||
</label>
|
||
<nav class="md-search__options" aria-label="Suche">
|
||
|
||
<button type="reset" class="md-search__icon md-icon" title="Zurücksetzen" aria-label="Zurücksetzen" tabindex="-1">
|
||
|
||
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"><path d="M19 6.41 17.59 5 12 10.59 6.41 5 5 6.41 10.59 12 5 17.59 6.41 19 12 13.41 17.59 19 19 17.59 13.41 12z"/></svg>
|
||
</button>
|
||
</nav>
|
||
|
||
</form>
|
||
<div class="md-search__output">
|
||
<div class="md-search__scrollwrap" tabindex="0" data-md-scrollfix>
|
||
<div class="md-search-result" data-md-component="search-result">
|
||
<div class="md-search-result__meta">
|
||
Suche wird initialisiert
|
||
</div>
|
||
<ol class="md-search-result__list" role="presentation"></ol>
|
||
</div>
|
||
</div>
|
||
</div>
|
||
</div>
|
||
</div>
|
||
|
||
|
||
|
||
<div class="md-header__source">
|
||
<a href="https://repo.toppyr.de/tobias/gwoe-antragspruefer" title="Zum Repository" class="md-source" data-md-component="source">
|
||
<div class="md-source__icon md-icon">
|
||
|
||
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 448 512"><!--! Font Awesome Free 7.1.0 by @fontawesome - https://fontawesome.com License - https://fontawesome.com/license/free (Icons: CC BY 4.0, Fonts: SIL OFL 1.1, Code: MIT License) Copyright 2025 Fonticons, Inc.--><path d="M439.6 236.1 244 40.5c-5.4-5.5-12.8-8.5-20.4-8.5s-15 3-20.4 8.4L162.5 81l51.5 51.5c27.1-9.1 52.7 16.8 43.4 43.7l49.7 49.7c34.2-11.8 61.2 31 35.5 56.7-26.5 26.5-70.2-2.9-56-37.3L240.3 199v121.9c25.3 12.5 22.3 41.8 9.1 55-6.4 6.4-15.2 10.1-24.3 10.1s-17.8-3.6-24.3-10.1c-17.6-17.6-11.1-46.9 11.2-56v-123c-20.8-8.5-24.6-30.7-18.6-45L142.6 101 8.5 235.1C3 240.6 0 247.9 0 255.5s3 15 8.5 20.4l195.6 195.7c5.4 5.4 12.7 8.4 20.4 8.4s15-3 20.4-8.4l194.7-194.7c5.4-5.4 8.4-12.8 8.4-20.4s-3-15-8.4-20.4"/></svg>
|
||
</div>
|
||
<div class="md-source__repository">
|
||
tobias/gwoe-antragspruefer
|
||
</div>
|
||
</a>
|
||
</div>
|
||
|
||
</nav>
|
||
|
||
</header>
|
||
|
||
<div class="md-container" data-md-component="container">
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<main class="md-main" data-md-component="main">
|
||
<div class="md-main__inner md-grid">
|
||
|
||
|
||
|
||
<div class="md-sidebar md-sidebar--primary" data-md-component="sidebar" data-md-type="navigation" >
|
||
<div class="md-sidebar__scrollwrap">
|
||
<div class="md-sidebar__inner">
|
||
|
||
|
||
|
||
|
||
<nav class="md-nav md-nav--primary" aria-label="Navigation" data-md-level="0">
|
||
<label class="md-nav__title" for="__drawer">
|
||
<a href="../.." title="GWÖ-Antragsprüfer Docs" class="md-nav__button md-logo" aria-label="GWÖ-Antragsprüfer Docs" data-md-component="logo">
|
||
|
||
|
||
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"><path d="M12 8a3 3 0 0 0 3-3 3 3 0 0 0-3-3 3 3 0 0 0-3 3 3 3 0 0 0 3 3m0 3.54C9.64 9.35 6.5 8 3 8v11c3.5 0 6.64 1.35 9 3.54 2.36-2.19 5.5-3.54 9-3.54V8c-3.5 0-6.64 1.35-9 3.54"/></svg>
|
||
|
||
</a>
|
||
GWÖ-Antragsprüfer Docs
|
||
</label>
|
||
|
||
<div class="md-nav__source">
|
||
<a href="https://repo.toppyr.de/tobias/gwoe-antragspruefer" title="Zum Repository" class="md-source" data-md-component="source">
|
||
<div class="md-source__icon md-icon">
|
||
|
||
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 448 512"><!--! Font Awesome Free 7.1.0 by @fontawesome - https://fontawesome.com License - https://fontawesome.com/license/free (Icons: CC BY 4.0, Fonts: SIL OFL 1.1, Code: MIT License) Copyright 2025 Fonticons, Inc.--><path d="M439.6 236.1 244 40.5c-5.4-5.5-12.8-8.5-20.4-8.5s-15 3-20.4 8.4L162.5 81l51.5 51.5c27.1-9.1 52.7 16.8 43.4 43.7l49.7 49.7c34.2-11.8 61.2 31 35.5 56.7-26.5 26.5-70.2-2.9-56-37.3L240.3 199v121.9c25.3 12.5 22.3 41.8 9.1 55-6.4 6.4-15.2 10.1-24.3 10.1s-17.8-3.6-24.3-10.1c-17.6-17.6-11.1-46.9 11.2-56v-123c-20.8-8.5-24.6-30.7-18.6-45L142.6 101 8.5 235.1C3 240.6 0 247.9 0 255.5s3 15 8.5 20.4l195.6 195.7c5.4 5.4 12.7 8.4 20.4 8.4s15-3 20.4-8.4l194.7-194.7c5.4-5.4 8.4-12.8 8.4-20.4s-3-15-8.4-20.4"/></svg>
|
||
</div>
|
||
<div class="md-source__repository">
|
||
tobias/gwoe-antragspruefer
|
||
</div>
|
||
</a>
|
||
</div>
|
||
|
||
<ul class="md-nav__list" data-md-scrollfix>
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<li class="md-nav__item">
|
||
<a href="../.." class="md-nav__link">
|
||
|
||
|
||
|
||
<span class="md-ellipsis">
|
||
|
||
|
||
Start
|
||
|
||
|
||
|
||
</span>
|
||
|
||
|
||
|
||
</a>
|
||
</li>
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<li class="md-nav__item md-nav__item--active md-nav__item--section md-nav__item--nested">
|
||
|
||
|
||
|
||
<input class="md-nav__toggle md-toggle " type="checkbox" id="__nav_2" checked>
|
||
|
||
|
||
<label class="md-nav__link" for="__nav_2" id="__nav_2_label" tabindex="">
|
||
|
||
|
||
|
||
<span class="md-ellipsis">
|
||
|
||
|
||
Architecture Decision Records
|
||
|
||
|
||
|
||
</span>
|
||
|
||
|
||
|
||
<span class="md-nav__icon md-icon"></span>
|
||
</label>
|
||
|
||
<nav class="md-nav" data-md-level="1" aria-labelledby="__nav_2_label" aria-expanded="true">
|
||
<label class="md-nav__title" for="__nav_2">
|
||
<span class="md-nav__icon md-icon"></span>
|
||
|
||
|
||
Architecture Decision Records
|
||
|
||
|
||
</label>
|
||
<ul class="md-nav__list" data-md-scrollfix>
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<li class="md-nav__item">
|
||
<a href="../" class="md-nav__link">
|
||
|
||
|
||
|
||
<span class="md-ellipsis">
|
||
|
||
|
||
Übersicht
|
||
|
||
|
||
|
||
</span>
|
||
|
||
|
||
|
||
</a>
|
||
</li>
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<li class="md-nav__item">
|
||
<a href="../0001-llm-citation-binding/" class="md-nav__link">
|
||
|
||
|
||
|
||
<span class="md-ellipsis">
|
||
|
||
|
||
0001 LLM-Citation-Binding
|
||
|
||
|
||
|
||
</span>
|
||
|
||
|
||
|
||
</a>
|
||
</li>
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<li class="md-nav__item">
|
||
<a href="../0002-adapter-architecture/" class="md-nav__link">
|
||
|
||
|
||
|
||
<span class="md-ellipsis">
|
||
|
||
|
||
0002 Adapter-Architektur
|
||
|
||
|
||
|
||
</span>
|
||
|
||
|
||
|
||
</a>
|
||
</li>
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<li class="md-nav__item md-nav__item--active">
|
||
|
||
<input class="md-nav__toggle md-toggle" type="checkbox" id="__toc">
|
||
|
||
|
||
|
||
|
||
|
||
<label class="md-nav__link md-nav__link--active" for="__toc">
|
||
|
||
|
||
|
||
<span class="md-ellipsis">
|
||
|
||
|
||
0003 Citation-Property-Tests
|
||
|
||
|
||
|
||
</span>
|
||
|
||
|
||
|
||
<span class="md-nav__icon md-icon"></span>
|
||
</label>
|
||
|
||
<a href="./" class="md-nav__link md-nav__link--active">
|
||
|
||
|
||
|
||
<span class="md-ellipsis">
|
||
|
||
|
||
0003 Citation-Property-Tests
|
||
|
||
|
||
|
||
</span>
|
||
|
||
|
||
|
||
</a>
|
||
|
||
|
||
|
||
<nav class="md-nav md-nav--secondary" aria-label="Inhaltsverzeichnis">
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<label class="md-nav__title" for="__toc">
|
||
<span class="md-nav__icon md-icon"></span>
|
||
Inhaltsverzeichnis
|
||
</label>
|
||
<ul class="md-nav__list" data-md-component="toc" data-md-scrollfix>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#kontext" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Kontext
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#optionen" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Optionen
|
||
|
||
</span>
|
||
</a>
|
||
|
||
<nav class="md-nav" aria-label="Optionen">
|
||
<ul class="md-nav__list">
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#option-a-schema-only-tests-was-wir-vorher-hatten" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Option A — Schema-only Tests (was wir vorher hatten)
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#option-b-property-test-gegen-die-echten-pdfs" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Option B — Property-Test gegen die echten PDFs
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#option-c-online-verifikation-pro-assessment-insert" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Option C — Online-Verifikation pro Assessment-Insert
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
</ul>
|
||
</nav>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#entscheidung" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Entscheidung
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#konsequenzen" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Konsequenzen
|
||
|
||
</span>
|
||
</a>
|
||
|
||
<nav class="md-nav" aria-label="Konsequenzen">
|
||
<ul class="md-nav__list">
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#positiv" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Positiv
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#negativ" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Negativ
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#folgen-fur-andere-adrs" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Folgen für andere ADRs
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
</ul>
|
||
</nav>
|
||
|
||
</li>
|
||
|
||
</ul>
|
||
|
||
</nav>
|
||
|
||
</li>
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<li class="md-nav__item">
|
||
<a href="../0004-deployment-workflow/" class="md-nav__link">
|
||
|
||
|
||
|
||
<span class="md-ellipsis">
|
||
|
||
|
||
0004 Deployment-Workflow
|
||
|
||
|
||
|
||
</span>
|
||
|
||
|
||
|
||
</a>
|
||
</li>
|
||
|
||
|
||
|
||
|
||
</ul>
|
||
</nav>
|
||
|
||
</li>
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<li class="md-nav__item md-nav__item--section md-nav__item--nested">
|
||
|
||
|
||
|
||
|
||
|
||
<input class="md-nav__toggle md-toggle md-toggle--indeterminate" type="checkbox" id="__nav_3" >
|
||
|
||
|
||
<label class="md-nav__link" for="__nav_3" id="__nav_3_label" tabindex="">
|
||
|
||
|
||
|
||
<span class="md-ellipsis">
|
||
|
||
|
||
Archiv
|
||
|
||
|
||
|
||
</span>
|
||
|
||
|
||
|
||
<span class="md-nav__icon md-icon"></span>
|
||
</label>
|
||
|
||
<nav class="md-nav" data-md-level="1" aria-labelledby="__nav_3_label" aria-expanded="false">
|
||
<label class="md-nav__title" for="__nav_3">
|
||
<span class="md-nav__icon md-icon"></span>
|
||
|
||
|
||
Archiv
|
||
|
||
|
||
</label>
|
||
<ul class="md-nav__list" data-md-scrollfix>
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<li class="md-nav__item">
|
||
<a href="../../archive/" class="md-nav__link">
|
||
|
||
|
||
|
||
<span class="md-ellipsis">
|
||
|
||
|
||
Übersicht
|
||
|
||
|
||
|
||
</span>
|
||
|
||
|
||
|
||
</a>
|
||
</li>
|
||
|
||
|
||
|
||
|
||
</ul>
|
||
</nav>
|
||
|
||
</li>
|
||
|
||
|
||
|
||
</ul>
|
||
</nav>
|
||
</div>
|
||
</div>
|
||
</div>
|
||
|
||
|
||
|
||
<div class="md-sidebar md-sidebar--secondary" data-md-component="sidebar" data-md-type="toc" >
|
||
<div class="md-sidebar__scrollwrap">
|
||
<div class="md-sidebar__inner">
|
||
|
||
|
||
<nav class="md-nav md-nav--secondary" aria-label="Inhaltsverzeichnis">
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<label class="md-nav__title" for="__toc">
|
||
<span class="md-nav__icon md-icon"></span>
|
||
Inhaltsverzeichnis
|
||
</label>
|
||
<ul class="md-nav__list" data-md-component="toc" data-md-scrollfix>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#kontext" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Kontext
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#optionen" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Optionen
|
||
|
||
</span>
|
||
</a>
|
||
|
||
<nav class="md-nav" aria-label="Optionen">
|
||
<ul class="md-nav__list">
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#option-a-schema-only-tests-was-wir-vorher-hatten" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Option A — Schema-only Tests (was wir vorher hatten)
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#option-b-property-test-gegen-die-echten-pdfs" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Option B — Property-Test gegen die echten PDFs
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#option-c-online-verifikation-pro-assessment-insert" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Option C — Online-Verifikation pro Assessment-Insert
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
</ul>
|
||
</nav>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#entscheidung" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Entscheidung
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#konsequenzen" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Konsequenzen
|
||
|
||
</span>
|
||
</a>
|
||
|
||
<nav class="md-nav" aria-label="Konsequenzen">
|
||
<ul class="md-nav__list">
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#positiv" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Positiv
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#negativ" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Negativ
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#folgen-fur-andere-adrs" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Folgen für andere ADRs
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
</ul>
|
||
</nav>
|
||
|
||
</li>
|
||
|
||
</ul>
|
||
|
||
</nav>
|
||
</div>
|
||
</div>
|
||
</div>
|
||
|
||
|
||
|
||
<div class="md-content" data-md-component="content">
|
||
|
||
<article class="md-content__inner md-typeset">
|
||
|
||
|
||
|
||
|
||
|
||
<h1 id="0003-sub-d-property-verification-zitate-als-substring-der-zitierten-pdf-seite">0003 — Sub-D Property-Verification: Zitate als Substring der zitierten PDF-Seite<a class="headerlink" href="#0003-sub-d-property-verification-zitate-als-substring-der-zitierten-pdf-seite" title="Permanent link">¶</a></h1>
|
||
<table>
|
||
<thead>
|
||
<tr>
|
||
<th></th>
|
||
<th></th>
|
||
</tr>
|
||
</thead>
|
||
<tbody>
|
||
<tr>
|
||
<td><strong>Status</strong></td>
|
||
<td>accepted</td>
|
||
</tr>
|
||
<tr>
|
||
<td><strong>Datum</strong></td>
|
||
<td>2026-04-10</td>
|
||
</tr>
|
||
<tr>
|
||
<td><strong>Refs</strong></td>
|
||
<td>Issues #50, #54, #60; tests/integration/test_citations_substring.py</td>
|
||
</tr>
|
||
</tbody>
|
||
</table>
|
||
<h2 id="kontext">Kontext<a class="headerlink" href="#kontext" title="Permanent link">¶</a></h2>
|
||
<p>Der LLM-Output enthält pro Assessment N Zitate, jedes mit <code>text</code>, <code>quelle</code>
|
||
(z.B. "GRÜNE NRW Wahlprogramm 2022, S. 58") und <code>url</code>. Wahrscheinlich
|
||
korrekt — aber wie verifizieren wir das, ohne jedes einzeln händisch
|
||
nachzuschlagen?</p>
|
||
<p>Die naheliegenden Test-Optionen sind alle unbefriedigend:</p>
|
||
<ul>
|
||
<li><strong>Mock-LLM-Tests</strong>: prüfen das Schema, sagen aber nichts über die
|
||
inhaltliche Korrektheit.</li>
|
||
<li><strong>Snapshot-Tests</strong> der LLM-Outputs: drift mit jedem Modell-Update.</li>
|
||
<li><strong>Manuelles Stichprobenchecken</strong>: skaliert nicht über mehrere BLs.</li>
|
||
</ul>
|
||
<h2 id="optionen">Optionen<a class="headerlink" href="#optionen" title="Permanent link">¶</a></h2>
|
||
<h3 id="option-a-schema-only-tests-was-wir-vorher-hatten">Option A — Schema-only Tests (was wir vorher hatten)<a class="headerlink" href="#option-a-schema-only-tests-was-wir-vorher-hatten" title="Permanent link">¶</a></h3>
|
||
<p>Pydantic validiert dass jedes Zitat die Felder <code>text</code>, <code>quelle</code>, <code>url</code> hat
|
||
und <code>url</code> mit <code>/static/referenzen/</code> beginnt. Erkennt syntaktische
|
||
Korruption, aber keine Halluzinationen.</p>
|
||
<h3 id="option-b-property-test-gegen-die-echten-pdfs">Option B — Property-Test gegen die echten PDFs<a class="headerlink" href="#option-b-property-test-gegen-die-echten-pdfs" title="Permanent link">¶</a></h3>
|
||
<p>Pro Zitat in der Prod-DB:
|
||
1. <code>quelle</code> per Token-Coverage-Match auf den <code>PROGRAMME</code>-Eintrag mappen.
|
||
2. Seitennummer aus <code>quelle</code> extrahieren.
|
||
3. Per <code>fitz</code> die PDF-Seite lesen, Whitespace + Soft-Hyphen normalisieren.
|
||
4. <code>text</code> muss als Substring (oder 5-Wort-Anker) in der Seite vorkommen.
|
||
5. Bug-Klasse 17 (Cross-Bundesland-Zitat): das aufgelöste Programm muss
|
||
zum Bundesland des Antrags passen, oder ein Grundsatzprogramm sein.</p>
|
||
<p><strong>Vorteile:</strong> prüft die einzige Eigenschaft die wirklich zählt — "war
|
||
das was zitiert wird auch wirklich da". Findet Halluzinationen direkt.</p>
|
||
<p><strong>Nachteile:</strong> braucht eine lokale Kopie der <code>gwoe-antraege.db</code> und der
|
||
Wahlprogramm-PDFs. Test ist Pydantic-Schema-übergreifend (Integration,
|
||
nicht Unit). Skipped sauber wenn DB nicht gemounted ist.</p>
|
||
<h3 id="option-c-online-verifikation-pro-assessment-insert">Option C — Online-Verifikation pro Assessment-Insert<a class="headerlink" href="#option-c-online-verifikation-pro-assessment-insert" title="Permanent link">¶</a></h3>
|
||
<p>Im <code>analyze_antrag</code>-Flow direkt nach LLM-Call jedes Zitat verifizieren
|
||
und bei Failure abbrechen oder retry.</p>
|
||
<p><strong>Vorteile:</strong> kein "stale data in DB"-Risiko.</p>
|
||
<p><strong>Nachteile:</strong> fügt Latenz und Komplexität in den Hot-Path. Die
|
||
Verifikation ist O(N×M), wo N=Zitate und M=Wahlprogramm-Pages.</p>
|
||
<h2 id="entscheidung">Entscheidung<a class="headerlink" href="#entscheidung" title="Permanent link">¶</a></h2>
|
||
<p><strong>Option B als pytest-Integration-Test</strong> — <code>tests/integration/test_citations_substring.py</code>,
|
||
parametrisiert per <code>_load_recent_assessments(limit_per_bl=5)</code> × <code>_flat_zitate()</code>.</p>
|
||
<p><strong>Strict substring</strong> als Default-Match (Whitespace + Soft-Hyphen normalisiert,
|
||
LLM-Truncation-Marker <code>...</code> toleriert), <strong>5-Wort-Anker als Fallback</strong> für
|
||
geringfügige Wort-Drift wie "LLM hat mittendrin gekürzt". Min-Length-Guard
|
||
von 20 Zeichen verhindert false-positive Matches auf trivialen Snippets.</p>
|
||
<p>Marker <code>pytestmark = pytest.mark.integration</code> — der Test läuft nicht in
|
||
der Default-Suite, sondern explizit per <code>pytest -m integration</code>. Skipped
|
||
wenn <code>webapp/data/gwoe-antraege.db</code> nicht existiert (Dev-Setup ohne
|
||
DB-Kopie).</p>
|
||
<p>Match-Helpers (<code>_normalize</code>, <code>_is_substring</code>, <code>_resolve_quelle_to_programm_id</code>,
|
||
<code>_extract_page_number</code>) sind eigene Unit-Tests in <code>TestHelpers</code> — die Match-
|
||
Logik selbst ist nicht-trivial und braucht ihre Eigenkontrolle.</p>
|
||
<h2 id="konsequenzen">Konsequenzen<a class="headerlink" href="#konsequenzen" title="Permanent link">¶</a></h2>
|
||
<h3 id="positiv">Positiv<a class="headerlink" href="#positiv" title="Permanent link">¶</a></h3>
|
||
<ul>
|
||
<li><strong>Findet Halluzinationen direkt</strong>: Issue #60 wurde durch den ersten
|
||
Live-Lauf dieses Tests entdeckt (3 von 36 Citations failed), ohne
|
||
dass ein Mensch Wahlprogramm-PDFs aufmachen musste.</li>
|
||
<li><strong>Re-runnable als Regression-Gate</strong>: nach jedem Deploy einmal <code>pytest -m
|
||
integration</code> gegen die DB → 0 Failures = OK.</li>
|
||
<li><strong>Test-Logik = Production-Logik</strong>: ADR 0001 Option B (<code>reconstruct_zitate</code>)
|
||
nutzt <strong>identische</strong> Match-Heuristiken (<code>find_chunk_for_text</code>,
|
||
<code>_normalize_for_match</code>). Damit kann der Test nichts fangen, was die
|
||
Production nicht auch fangen würde, und umgekehrt — kein Test-/Prod-Drift.</li>
|
||
</ul>
|
||
<h3 id="negativ">Negativ<a class="headerlink" href="#negativ" title="Permanent link">¶</a></h3>
|
||
<ul>
|
||
<li><strong>Lokale DB-Kopie nötig</strong>: vor jedem Sub-D-Run muss <code>data/gwoe-antraege.db</code>
|
||
vom Container gepullt werden. CI-Integration steht aus.</li>
|
||
<li><strong>Test ist langsam-ish</strong>: ~50 Citations × ein PDF-Open pro Programm ist
|
||
bei den ~30 indexierten Programmen ~250ms im Ganzen, nicht trivial aber
|
||
nicht prohibitiv.</li>
|
||
<li><strong>Token-Coverage-Heuristik für Quelle-zu-Programm-Mapping</strong> kann false-
|
||
positive bei sehr ähnlichen Programmen werden (z.B. CDU NRW 2022 vs. CDU
|
||
Niedersachsen 2022 — würde durch Bundesland-Bonus-Check abgefangen).</li>
|
||
</ul>
|
||
<h3 id="folgen-fur-andere-adrs">Folgen für andere ADRs<a class="headerlink" href="#folgen-fur-andere-adrs" title="Permanent link">¶</a></h3>
|
||
<ul>
|
||
<li><strong>ADR 0001</strong> ist von ADR 0003 abhängig — wenn dieser Test entfernt würde,
|
||
hätte der LLM-Citation-Postprocess keinen Backstop und neue Halluzinations-
|
||
Bug-Klassen würden still durchrutschen.</li>
|
||
</ul>
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
</article>
|
||
</div>
|
||
|
||
|
||
<script>var target=document.getElementById(location.hash.slice(1));target&&target.name&&(target.checked=target.name.startsWith("__tabbed_"))</script>
|
||
</div>
|
||
|
||
</main>
|
||
|
||
<footer class="md-footer">
|
||
|
||
<div class="md-footer-meta md-typeset">
|
||
<div class="md-footer-meta__inner md-grid">
|
||
<div class="md-copyright">
|
||
|
||
|
||
Made with
|
||
<a href="https://squidfunk.github.io/mkdocs-material/" target="_blank" rel="noopener">
|
||
Material for MkDocs
|
||
</a>
|
||
|
||
</div>
|
||
|
||
</div>
|
||
</div>
|
||
</footer>
|
||
|
||
</div>
|
||
<div class="md-dialog" data-md-component="dialog">
|
||
<div class="md-dialog__inner md-typeset"></div>
|
||
</div>
|
||
|
||
|
||
|
||
|
||
|
||
<script id="__config" type="application/json">{"annotate": null, "base": "../..", "features": ["navigation.sections", "navigation.expand", "search.highlight"], "search": "../../assets/javascripts/workers/search.2c215733.min.js", "tags": null, "translations": {"clipboard.copied": "In Zwischenablage kopiert", "clipboard.copy": "In Zwischenablage kopieren", "search.result.more.one": "1 weiteres Suchergebnis auf dieser Seite", "search.result.more.other": "# weitere Suchergebnisse auf dieser Seite", "search.result.none": "Keine Suchergebnisse", "search.result.one": "1 Suchergebnis", "search.result.other": "# Suchergebnisse", "search.result.placeholder": "Suchbegriff eingeben", "search.result.term.missing": "Es fehlt", "select.version": "Version ausw\u00e4hlen"}, "version": null}</script>
|
||
|
||
|
||
<script src="../../assets/javascripts/bundle.79ae519e.min.js"></script>
|
||
|
||
|
||
</body>
|
||
</html> |