Skip to content

CLAUDE.md

Latest commit

 

History

History
102 lines (75 loc) · 3.93 KB

CLAUDE.md

File metadata and controls

102 lines (75 loc) · 3.93 KB
Raw

CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

Build and Run Commands

Rust Backend

cargo run                    # Run in development mode
cargo build --release        # Build for release
cargo test                   # Run tests
cargo test --lib <test_name> # Run a specific test

Web Frontend (Vue 2)

cd web
npm install                   # Install dependencies
npm run serve                # Development server
npm run build                # Production build
npm run lint                 # Lint code

Configuration

Configuration is loaded from environment variables. Default values are defined in src/app/config.rs. Key settings:

  • SERVER_HOST / SERVER_PORT - Server binding (default: 0.0.0.0:8080)
  • DATABASE_URL - SQLite path (default: sqlite:storage/reader.db?mode=rwc)
  • WEB_ROOT - Static web files directory (default: ../reader/web)
  • STORAGE_DIR / ASSETS_DIR - Storage paths
  • LOG_LEVEL - Logging verbosity (default: info)
  • REQUEST_TIMEOUT_SECS - HTTP request timeout (default: 15)

Environment variable separator is __ (double underscore).

Architecture Overview

This is a Rust implementation of "阅读3.0" (Reader 3.0) - a book reading API server. It provides book source management, search, chapter retrieval, and content parsing.

Module Structure

  • src/api/ - HTTP handlers and routing (axum)

    • router.rs - Route definitions for all /reader3/* endpoints
    • handlers/ - Request handlers by domain (book, bookmark, rss, user, etc.)
    • AppState - Shared application state (config, services)

  • src/service/ - Business logic layer

    • book_service.rs - Book search, info, chapters, content
    • book_source_service.rs - CRUD for book sources (stored in SQLite)
    • user_service.rs - User management and authentication

  • src/parser/ - Content extraction engine

    • rule_engine.rs - Main entry point, auto-detects content type
    • html.rs - CSS selector parsing (using scraper crate)
    • jsonpath.rs - JSONPath queries (using jsonpath_lib)
    • js.rs - JavaScript execution (rquickjs) for js: prefixed rules

  • src/crawler/ - HTTP fetching

    • http_client.rs - Configurable reqwest client with compression support
    • fetcher.rs - Page fetching with URL resolution

  • src/model/ - Data structures

    • book_source.rs - Book source JSON schema with rule definitions
    • rule.rs - SearchRule, BookInfoRule, TocRule, ContentRule types

  • src/storage/ - Persistence

    • db/ - SQLite via sqlx, with BookSourceRepo
    • cache/ - File-based chapter content cache (MD5 key)
    • fs/ - Filesystem operations for storage/assets

Request Flow

  1. api/handlers/ receives HTTP request
  2. Handler calls service/ layer
  3. Service fetches remote content via crawler/http_client.rs
  4. Content is parsed using parser/rule_engine.rs with rules from BookSource
  5. Results returned as JSON

Rule Parsing

Book sources define parsing rules in JSON. The RuleEngine auto-detects parsing mode:

  • CSS selectors - Default for HTML (.class, #id, tag)
  • JSONPath - Auto-detected for JSON content ($.data.list)
  • XPath - Lines starting with / or ./
  • JavaScript - Rules prefixed with js: or @js:
  • Regex - Rules starting with :

Rule prefixes can be explicit: @css:, @json:, @xpath:, @regex:.

Book Source Format

Book sources are JSON objects containing:

  • bookSourceUrl / bookSourceName - Source identity
  • searchUrl / exploreUrl - Search/discovery URLs with ${key} placeholders
  • ruleSearch / ruleBookInfo / ruleToc / ruleContent - Parsing rules for each stage
  • Each rule has fields like bookList, name, author, bookUrl, etc.

Web Frontend

The web/ directory contains a Vue 2 frontend (阅读3.0 web client). It connects to the Rust backend API. Build output goes to web/dist/, which the backend serves as static files.