Integrate AdguardTeam libraries: High-impact roadmap and next steps

[jaypatrick]

Context

This issue summarizes key findings from a review of relevant AdguardTeam repositories and outlines practical next steps to upgrade the adblock-compiler project. The intent is to maximize benefit by leveraging proven tools/libraries/research.

TL;DR - Top 3 Impact Actions

1. Integrate AGLint for Rule Validation

   • Adopt AdguardTeam/AGLint as a linting/validation engine for filter rules. This boosts reliability, supports multiple formats, and offloads future maintenance.
   • Why high value: It's MIT-licensed (easy to adopt), maintained by AdGuard, and will make your project's validation as robust as possible, even for edge-case rules.

2. Leverage tsurlfilter for Smart Deduplication & Simulation

   • Use AdguardTeam/tsurlfilter as a dependency to simulate, parse, and perform semantic deduplication of filter rules beyond string/text checks.
   • Bonus: Enables in-app/live testing ("does this rule work?") for end users.
   • Legal: It's GPL-3.0; mind your own repo's license or wrap usage in a service if that matters.

3. Deep Dive & (Optionally) Port FiltersCompiler

   • Study AdguardTeam/FiltersCompiler. It's the reference implementation for filter transformation and validation, written in JS. Consider porting its core logic (especially filter/metadata combining logic) to C# or .NET if you want native performance or cross-platform CLI.
   • Why: If you want to offer a .NET-native ecosystem for adblock compilation and not just web APIs.

Research Table

Repo	Value to Our Project	My Thoughts
AGLint	Plug directly into validation pipeline; very low effort, very high reward	Absolutely recommended—quickest win, best license, already mature
tsurlfilter	Power semantic deduplication/logical rule analysis	Worth adding for advanced deduplication and functional tests, but careful with GPL dependencies
FiltersCompiler	Model for a true compiler pipeline	Fork or port is only needed if we want deep engine parity or .NET-native CLI; at minimum, use as a reference

My Additional Thoughts

• These three steps would future-proof adblock-compiler, let us focus on UI/automation rather than low-level adblock quirks, and stand on the shoulders of best-in-class research.
• License compatibility and API boundaries matter—AGLint (MIT) is a no-brainer, tsurlfilter (GPL) should be isolated if needed.
• Long term: consider contributing improvements or bug-fixes upstream to AdguardTeam—ongoing synergy is likely given overlap.