News Content Filter

A lightweight userscript that filters unwanted news articles based on customizable keywords. Built for Tampermonkey/Greasemonkey browser extensions.

Table of Contents

• Overview
• Features
• Project Structure
• Installation
• Configuration
• How It Works
• Troubleshooting
• Privacy & Security
• Contributing
• License
• Credits

Overview

ZeroNoise automatically hides news articles containing specific keywords from supported news websites. It works on both static and dynamically loaded content (infinite scroll, AJAX updates), ensuring a cleaner, quieter browsing experience.

Features

• Single-file simplicity: All configuration and logic in one JavaScript file
• Easy customization: Keywords and site templates defined at the top for quick editing
• Dynamic content support: Monitors page changes and filters new articles automatically
• Multi-site support: Pre-configured for 8 major Spanish news sites
• Whole-word matching: Intelligent pattern matching avoids false positives
• Case-insensitive: Matches keywords regardless of capitalization
• Zero dependencies: Pure vanilla JavaScript, no external libraries

Project Structure

zeronoise/
│
├── readme.md              # Documentation (this file)
│
└── zeronoise.js           # Main userscript
    ├── Configuration      # Keywords list (edit here)
    ├── Site Templates     # CSS selectors for each news site (advanced users only)
    └── Core Logic         # Filtering and detection algorithms

Single File Architecture

ZeroNoise consolidates everything into one file for easier management and portability.

Installation

1. Install a Userscript Manager

Choose one for your browser:

• Tampermonkey (Recommended - Chrome, Firefox, Safari, Edge, Opera)
• Greasemonkey (Firefox only)
• Violentmonkey (Chrome, Firefox, Edge)

2. Install ZeroNoise

1. Click on the Tampermonkey icon in your browser toolbar
2. Select "Create a new script"
3. Delete the default template content
4. Copy and paste the entire contents of zeronoise.js
5. Save the script (Ctrl+S or Cmd+S)

3. Verify Installation

1. Visit a supported news site (e.g., elpais.com, elmundo.es)
2. Open browser console (F12 → Console tab)
3. Look for ZeroNoise log messages:

   [ZeroNoise] Initializing for elpais.com with 37 keywords
   [ZeroNoise] Filtered 12 articles on initial load
   [ZeroNoise] Watching for new content
   

Configuration

Adding/Removing Keywords

Open zeronoise.js in the Tampermonkey editor and modify the KEYWORDS array:

const KEYWORDS = [
    'Trump',
    'Mazón',
    // Add your keywords here
    'new keyword',
    'last keyword'
    // Remove any keywords you don't want
];

Tips for Keywords:

• Multi-word phrases are supported: 'Donald Trump'
• Keywords are case-insensitive: 'trump' matches "Trump", "TRUMP", "trump"
• Whole-word matching: 'Trump' won't match "trumpet"
• Special characters work: 'Mazón'

Performance Note

ZeroNoise uses explicit @match directives in the script header, which means:

• The script only loads on news sites (not on every website you visit)
• Zero overhead when browsing non-news sites
• Faster browser performance overall
• When adding a new site, you must update both the @match directive and the SITE_TEMPLATES section

Supported News Sites

ZeroNoise comes pre-configured for these Spanish news sites:

Site	Domain	Type
ABC	abc.es	General news
El Mundo	elmundo.es	General news
El País	elpais.com	General news
La Vanguardia	lavanguardia.com	General news
La Razón	larazon.es	General news
Menéame	meneame.net	News aggregator
Marca	marca.com	Sports
AS	as.com	Sports

Adding a New Site

To add support for a new news website, you need to make two changes:

Step 1: Add @match directive (header)

1. Open zeronoise.js in Tampermonkey editor
2. Locate the @match directives at the top (lines 7-22)
3. Add two new lines for your site:

// @match        *://yournewssite.com/*
// @match        *://*.yournewssite.com/*

The *:// matches both http and https. The second line with *. catches subdomains like www.yournewssite.com.

Step 2: Add site template (configuration)

1. Scroll down to the SITE_TEMPLATES section
2. Add a new entry following this structure:

const SITE_TEMPLATES = {
    // ... existing sites ...
    
    'yournewssite.com': {
        // Selector for article containers
        articleSelector: 'article, div.news-item',
        
        // Selector for headlines within articles
        headlineSelector: 'h2, h3, .title',
        
        // Selector for descriptions (optional)
        descriptionSelector: 'p.summary, .description'
    }
};

Finding the Right Selectors

1. Visit the news site
2. Open Developer Tools (F12)
3. Use the Element Inspector (Ctrl+Shift+C)
4. Click on an article, then find:
   • Article container: The element wrapping the entire article
   • Headline: The link or heading with the article title
   • Description: The summary or excerpt text (if present)

Example walkthrough:

<!-- If the site HTML looks like this: -->
<article class="story">
    <h2 class="story-title">
        <a href="/news/123">Breaking News Title</a>
    </h2>
    <p class="story-summary">This is the article summary...</p>
</article>

<!-- Your template should be: -->
'example.com': {
    articleSelector: 'article.story',
    headlineSelector: 'h2.story-title a',
    descriptionSelector: 'p.story-summary'
}

How It Works

Filtering Process

1. Page Load: When you visit a news site, ZeroNoise checks if it's supported
2. Initial Scan: Finds all articles using the site template selectors
3. Text Analysis: Extracts headline and description text from each article
4. Keyword Matching: Checks if text contains any blacklisted keywords
5. Article Hiding: Sets display: none on matching articles
6. Continuous Monitoring: Watches for new articles loaded via scroll/AJAX

Matching Algorithm

Article text: "Trump announces new policy on trade"
Normalized:   "Trump announces new policy on trade"
Keyword:      "Trump"
Regex:        (^|\s)Trump(\s|$)
Result:       ✓ MATCH (whole word at start)

Article text: "The trumpet player wins award"
Normalized:   "The trumpet player wins award"
Keyword:      "Trump"
Regex:        (^|\s)Trump(\s|$)
Result:       ✗ NO MATCH (not a whole word)

Performance

• Lightweight: Pure JavaScript, no heavy frameworks
• Efficient: Only processes new content, marks filtered articles
• Non-blocking: Uses MutationObserver for asynchronous updates
• Memory-friendly: Minimal DOM manipulation

Troubleshooting

Articles aren't being filtered

1. Check the console (F12) for ZeroNoise messages

   • No messages? Script might not be enabled in Tampermonkey
   • "No template found"? Site isn't supported yet

2. Verify selectors for the site

   • Site may have updated their HTML structure
   • Use browser DevTools to find new selectors

3. Test keywords

   • Try a very common word to see if filtering works
   • Check if keywords match exactly (accounting for punctuation)

Too many articles are hidden

• Review your keywords list for overly broad terms
• Consider removing generic keywords that match unrelated content

Script slows down the page

• Reduce the number of keywords (fewer = faster)
• The 1-second delay in init() can be adjusted if needed

Site is supported but not working

The site may have changed their HTML structure. Steps to fix:

1. Right-click an article → Inspect Element
2. Find the CSS selectors for:
   • Article container
   • Headline
   • Description
3. Update the template in SITE_TEMPLATES

Privacy & Security

• No external requests: All processing happens locally in your browser
• No data collection: ZeroNoise doesn't track or store any data
• No remote code: No code is loaded from external sources
• Open source: All code is visible and auditable in the userscript

Contributing

To contribute new site templates:

1. Test the selectors thoroughly on the target site
2. Verify they work with dynamic content (scroll, load more)
3. Use specific selectors to avoid false positives
4. Document any site-specific quirks

License

This project is open source and available for personal use.

Credits

• Concept initially inspired by Antiayuseame.

---

Version: 1.0
Last Updated: January 2026
Compatibility: All modern browsers with Tampermonkey/Greasemonkey support