A high-performance Python utility that fetches and merges multiple adlists into domain-only output for DNS blockers like Pi-hole, AdGuard, and similar DNS filtering solutions.

• Features
• Quick Start
• Installation (optional)
• Configuration
• Output Format
• Performance
• How It Works
• Architecture
• VS Code Tasks
• Redundancy Analysis
• Troubleshooting
• Example Output
• Requirements
• Use Cases
• Contributing
• License
• Acknowledgments

• Fast Concurrent Processing: Processes 1.6M+ entries from 50+ sources in ~50-60 seconds
• Zero Dependencies: Uses only Python standard library (3.8+)
• Dual Output: Generates both adlists and whitelists simultaneously
• Smart Content Processing: Handles domains, wildcards, regex patterns, and Pi-hole format conversions
• ABP Filter Support: Converts Pi-hole regex patterns to AdBlock Plus (ABP) format with automatic wildcard normalization
• Intelligent Separation: Automatically separates exception rules (@@||) from blocklist to whitelist
• Domain Validation: Validates and filters invalid domain entries during post-processing
• Real-time Progress: Animated progress spinners with detailed status updates
• Error Resilient: Failed fetches don't crash the pipeline; they're logged and filtered out

1. Clone the repository:

   git clone https://github.com/Toomas633/Adlist-Parser.git
   cd Adlist-Parser

2. Run the parser:

   • From a repo checkout:

     python -m adparser

   • If installed as a package:

     adlist-parser

   On Windows PowerShell:

   python -m adparser
   # or, if installed
   adlist-parser

3. Find your results:

   • output/adlist.txt - Merged blocklist (~1.6M entries)
   • output/whitelist.txt - Merged whitelist (~2K entries)

You can run from a checkout (above) or install locally to get the adlist-parser command:

# Editable install for development
python -m pip install -e .

# Then run
adlist-parser

Configure your sources in JSON files:

data/adlists.json - Blocklist sources:

{
  "lists": ["blacklist.txt", "old_adlist.txt"],
  "urls": [
    "https://raw.githubusercontent.com/StevenBlack/hosts/master/hosts",
    "https://adaway.org/hosts.txt",
    "https://v.firebog.net/hosts/AdguardDNS.txt"
  ]
}

data/whitelists.json - Whitelist sources:

{
  "lists": ["whitelist.txt"],
  "urls": [
    "https://raw.githubusercontent.com/hagezi/dns-blocklists/main/domains/whitelist-referral.txt"
  ]
}

• URLs: HTTP/HTTPS links to remote lists
• Local files: Relative paths to files in the data/ directory
• Mixed format: Each source can contain domains, wildcards, regex patterns, or Pi-hole entries

Notes:

• Path resolution: relative paths inside the JSON files are resolved relative to the JSON file location (not the CWD).
• Accepted keys: both files accept any of lists, urls, adlists, or sources for compatibility; they are merged.

The parser intelligently handles multiple input formats:

• Plain domains: example.com
• Wildcards: *.example.com
• Pi-hole regex: (\.|^)example\.com$
• AdBlock patterns: /pattern/flags
• Host file entries: 0.0.0.0 example.com
• Comments: Lines starting with #, !, //, or ;

1. Domain Extraction: Extracts clean domains from various host file formats
2. Wildcard Handling: *.domain.com is preserved as a domain token (wildcard not expanded). In the final domain output the leading *. is stripped to domain.com.
3. ABP Normalization: Fixes broken ABP patterns automatically:
   • ||*cdn.domain.com^ → ||*.cdn.domain.com^ (missing dot after wildcard)
   • ||app.*.adjust.com^ → ||*.adjust.com^ (wildcard-only label removed)
   • ||domain.google.*^ → ||domain.google^ (wildcard TLD removed - not supported)
   • -domain.com^ → ||-domain.com^ (adds missing || prefix)
   • @@|domain.com^| → @@||domain.com^ (fixes single pipe + trailing pipe)
4. ABP Conversion: Pi-hole regex patterns convert to ||domain^ format when possible
5. Blocklist/Whitelist Separation: Automatically moves @@|| exception entries from blocklist to whitelist
6. Domain Validation: Validates and removes invalid domain entries during post-processing
7. Regex Handling: Complex regexes that can't convert to ABP are discarded (pipeline doesn't crash)
8. Deduplication: Preserves first-seen order during normalization; final outputs are sorted case-insensitively during post-processing
9. Comment Filtering: Strips whole-line and inline comments (#, !, //, ;)
10. HTML Filtering: Removes HTML tags and attributes from lists
11. Error Resilience: Failed fetches logged and filtered during normalization
12. Adlist merge: Adlist pipeline merges with prior output/adlist.txt before writing to preserve entries across transient source failures (whitelist writes directly)

• Outputs use LF-only line endings.
• Sorting is deterministic and case-insensitive; deduplication is case-insensitive and whitespace-trimmed.
• Headers are regenerated during post-processing (don’t hand-edit outputs).

Each output starts with a generated header like this:

# Adlist - Generated by Adlist-Parser
# https://github.com/Toomas633/Adlist-Parser
#
# Created/modified: 2025-01-01 00:00:00 UTC
# Total entries: 1,684,272
# Domains: 400,527
# ABP-style rules: 1,283,745
# Sources processed: 50
#
# This file is automatically generated. Do not edit manually.
# To update, run: adlist-parser or python -m adparser
#

• Concurrency: Fetches multiple sources simultaneously (max 16 workers)
• Async Processing: Adlists and whitelists processed in parallel
• Memory Efficient: Line-by-line processing for large datasets
• Real-world Scale: Tested with 1.6M+ entries from 50+ sources

• Concurrency: network fetching uses up to 16 workers (see adparser/fetcher.py). You can adjust the cap there if needed for your environment.
• I/O: most heavy I/O runs off the event loop using asyncio.to_thread(); disk speed can impact total time.
• Output size: output/adlist.txt can reach ~1.6–1.7M lines depending on sources.

Two concurrent pipelines run via asyncio.gather() in adparser/cli.py:

Pipeline Flow (for both adlist and whitelist):

1. Load Sources → Parse JSON configs and resolve paths
2. Fetch Content → Concurrent downloads (16 workers max)
3. Generate List → Normalize, categorize, and convert entries
4. Adlist-only merge → Merge new entries with prior output/adlist.txt before write (preserves previous content across transient source failures); whitelist writes directly
5. Write Output → Save with auto-generated headers (LF-only line endings)
6. Post-Processing → Separate blocklist/whitelist entries, validate domains, regenerate headers, and re-write both files
7. Redundancy Report → Analyze duplicates and overlaps

Key Processing Steps:

• All heavy I/O wrapped with asyncio.to_thread() to keep event loop responsive
• Progress displayed via animated spinners (adparser/status.py)
• Domain validation using DOMAIN_RE regex pattern
• Pi-hole regex → ABP conversion when patterns are simple enough
• ABP wildcard normalization fixes malformed patterns automatically
• Post-processing separates @@|| exception entries to whitelist
• Cross-list deduplication ensures no conflicts between blocklist and whitelist
• Failed sources tracked separately and reported at end
• Files are written with LF-only line endings

The codebase follows a modular async architecture with strict separation of concerns:

adparser/cli.py           # Main orchestrator with async/await
├── adparser/io.py        # JSON parsing, path resolution, file I/O
├── adparser/fetcher.py   # Concurrent HTTP fetching (ThreadPoolExecutor)
├── adparser/content.py   # Domain extraction, normalization, regex conversion
├── adparser/models.py    # Source descriptor dataclass (URL vs local files)
├── adparser/status.py    # Progress spinners and terminal UI updates
├── adparser/reporting.py # Results summary with emoji formatting
├── adparser/redundancy.py# Duplicate detection and overlap analysis
└── adparser/constants.py # File path constants

Design Principles:

• Single Responsibility: Each module handles one concern (fetch/parse/write separated)
• Error Isolation: Failed sources don't crash the pipeline
• Async I/O: Heavy operations run in thread pool via to_thread()
• Progress Feedback: Global status_display coordinates concurrent spinners
• Order Preservation: Deduplication via _dedupe_preserve_order() before final sort

This repository includes ready-to-run tasks for Windows PowerShell:

• Adlist-Parser: runs the tool end-to-end (equivalent to python -m adparser).
• Tests: Pytest (coverage): runs pytest with coverage as configured in pyproject.toml.
• Lint: Pylint (report): runs pylint and writes pylint-report.txt; non-zero exit is allowed while still generating the report.

Quality gates expectations:

• Build: N/A (pure Python). Runtime entry point is adparser.cli:main.
• Tests: PASS on local data set; full runs may take ~50–60s.
• Lint: PASS is ideal; if not, review pylint-report.txt.

The parser includes built-in redundancy detection to help optimize your source lists:

Features:

• Duplicate Detection: Identifies sources with identical content
• Local File Analysis: Shows which entries in local files are already covered by remote sources
• Removal Suggestions: Lists first 20 redundant entries with count of remaining

Example Output:

🔁 Duplicate sources (identical content): 2 groups
├─ 🌐 https://example.com/list1.txt
└─ 🌐 https://example.com/list2.txt
💡 Tip: Keep one source from this group, remove the others

📄 Local file redundancy analysis:
• blacklist.txt: 150/200 entries (75.0%) already in remote sources

The parser intelligently processes various input formats and converts them appropriately:

Common Issues:

• Network Errors: Failed sources are listed as "UNAVAILABLE SOURCES" in the final report with 🌐 (remote) or 📄 (local) indicators
• Proxy Issues: Configure system proxy settings or mirror remote sources locally in data/ and update JSON configs
• Large Files: output/adlist.txt can be 30MB+; use command-line tools (grep, wc -l) for inspection
• Slow Performance: Check network speed; adjust worker count in adparser/fetcher.py (default: 16)
• Memory Usage: The parser uses line-by-line processing, so memory footprint stays low even with 1.6M+ entries

• Why are element-hiding rules (e.g., ##, #@?#) missing from outputs?
  • This tool targets DNS blocklists. Element-hiding is cosmetic (browser-side), so such rules are dropped during normalization.
• Why do some regex rules disappear?
  • Only simple, anchored Pi-hole patterns are converted to ABP (||domain^). Complex/JS-like regex is discarded for safety and DNS relevance.
• My local file entries are already covered by remotes—how do I find them?
  • Check the redundancy section at the end of the run; it lists duplicates and local entries already provided by remote sources.

🚀 Starting Adlist-Parser...
⚡ Processing adlists and whitelists concurrently...

⚡ Adlist: Fetching content... |/-\ [48/50 (96%)]
⚡ Whitelist: Processing domains...

Adlist: ✅ Complete - 1684272 entries (400527 domains, 1283745 ABP rules)
Whitelist: ✅ Complete - 2337 entries (1346 domains, 991 ABP rules)

=== Adlists redundancy analysis ===
Analyzed 50 sources.
✅ No redundancy issues detected

============================================================
🎉 ALL PROCESSING COMPLETED IN 53.16 SECONDS! 🎉
============================================================
📊 RESULTS SUMMARY:
┌──────────────────────────────────────────────────────────┐
│ 🛡️  ADLIST:    50 sources → 1684272 entries              │
│   📝 Domains:  400527 | ABP rules: 1283745              │
├──────────────────────────────────────────────────────────┤
│ ✅ WHITELIST:  6 sources →    2337 entries               │
│   📝 Domains:    1346 | ABP rules:     991              │
├──────────────────────────────────────────────────────────┤
│ 📁 Output files:                                         │
│   • output/adlist.txt                                    │
│   • output/whitelist.txt                                 │
└──────────────────────────────────────────────────────────┘

• Python 3.8 or higher
• No external dependencies (uses only standard library)

• Pi-hole: Use output/adlist.txt as a blocklist and output/whitelist.txt as an allowlist
• AdGuard Home: Import both files as custom filtering rules
• DNS Filtering: Any DNS-based ad blocker that supports domain lists
• Network Security: Corporate firewall domain blocking lists

This project is licensed under the MIT License - see the LICENSE file for details.

Contributions are welcome! Here's how to get started:

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Make your changes following the existing patterns:
   • Keep modules single-responsibility
   • Use asyncio.to_thread() for I/O operations
   • Wrap long operations with spinner.show_progress()
   • Add inline documentation for complex logic

• Keep runtime stdlib-only; do not add dependencies or modify pyproject.toml beyond [project.optional-dependencies].dev
• Preserve public contracts: fetcher.fetch, content.generate_list, content.separate_blocklist_whitelist, io.write_output
• Do not widen domain regex or IDN heuristics; keep _maybe_extract_domain and DOMAIN_RE conservative

1. Test thoroughly with python -m adparser
2. Verify output files are generated correctly
3. Submit a pull request with clear description

Development Tips:

• Read .github/copilot-instructions.md for architecture overview
• Check adparser/content.py for parsing rules and regex patterns
• Use existing regex patterns rather than adding new ones
• Maintain backward compatibility with existing JSON configs

For quick iterations, limit sources to local files to reduce runtime:

• Edit data/adlists.json and data/whitelists.json to only include the local files:

  { "lists": ["blacklist.txt"], "urls": [] }

  { "lists": ["whitelist.txt"], "urls": [] }

• Add a few test lines to data/blacklist.txt and data/whitelist.txt and run:

  python -m adparser

This exercises the full pipeline (status UI, normalization, separation, reporting) in seconds.

• Built for the DNS filtering community
• Inspired by the need for fast, reliable adlist aggregation
• Uses high-quality sources from the community (StevenBlack, Hagezi, FadeMind, and others)