The erldns repository is an Erlang/OTP-based authoritative DNS server that processes DNS queries through a modular pipeline architecture. It loads zone data from JSON or standard zone files, provides DNSSEC validation and signing, and includes features like query throttling, EDNS0 processing, and an HTTP-based Admin API for runtime zone management and monitoring. The system is built using OTP supervision trees, with separate supervisors for listeners, pipeline workers, and zone loading. Key architectural strengths include a reusable pipeline pattern for extensible query processing, a worker pool for asynchronous operations, a segmented cache abstraction for high-concurrency scenarios, and admission control mechanisms using CoDel and Cubic algorithms. The Admin API follows a middleware pattern allowing custom extensions, and the zone loading subsystem supports pluggable codecs for custom record types. The codebase demonstrates strong engineering practices with comprehensive testing, dependency management, and observability via telemetry.
This repository provides a library and tools for training and deploying CNN models (RotNet, RotNetR) to predict the rotation angle of images, primarily designed for cracking rotate captchas. The system includes a training pipeline with customizable dataset preprocessing (random rotations, cropping, normalization), model definitions using ResNet50 and RegNet_Y backbones, a loss function optimized for angular classification (Circular Smooth Label), and an HTTP server for inference. The codebase is well-structured as a Python package with clean separation of concerns, showcasing advanced patterns like generic pipe operators for data transformation composition and context-managed training loops. It is actively maintained and demonstrates solid software engineering practices.
PILKit is a Python library providing image processing utilities for PIL/Pillow. It offers a collection of processor classes (e.g., ResizeToFit, SmartCrop, GaussianBlur) that can be chained via a Pipeline pattern to manipulate images. The library also includes utility functions for image format handling, saving, and transparency management. Architecture-wise, it's a modular library with clear separation of concerns: processors handle specific transformations, utilities manage format conversions, and the Pipeline enables composition. The codebase demonstrates solid engineering practices with version compatibility layers and thoughtful error handling for PIL's quirks.