AstroResearch/Cargo.toml
Asfmq 3f1935678b feat: LLM/Embedding 客户端模块化、侧边栏折叠交互、arXiv→ADS 下载回退与前端体验重构
**后端架构**
  - 抽取翻译服务中内嵌的 LLM HTTP 调用为独立的 LlmClient /
    EmbeddingClient(src/clients/llm.rs),翻译模块改为委托调用,消除
    对 reqwest/serde 的直接耦合
  - Config 新增 EMBEDDING_API_KEY/EMBEDDING_API_BASE/EMBEDDING_MODEL
    三项配置,默认 fallback 至 LLM 对应值,补齐向量嵌入基础设施

  **下载策略优化**
  - arXiv 直连下载失败后自动回退至 ADS 网关 PUB_PDF→EPRINT_PDF→CrossRef
    多级通道,替换此前单路径策略;批量同步同步应用此逻辑
  - PDF/HTML 任一方成功时,失败方的 path 字段不再存储 "error:" 报错字符串,
    改为置 NULL,防止日志污染数据

  **前端交互增强**
  - 侧边栏支持折叠/展开:收起为仅图标模式(w-16),展开恢复完整模式(w-64);
    收起后点击 Logo 展开,含流畅 cubic-bezier 过渡动画
  - 阅读面板新增 PDF 内嵌预览:已下载 PDF 时可通过 iframe 切换查看
    /api/files 下的本地文献
  - reader/citation 面板未选文献时展示带图标的空状态引导页,替代空白页
  - 文献详情面板改为固定高度弹性布局(h-[460px]),各区块按比例分配避免
    内容挤压;期刊名过长截断+悬停tooltip;关键词无数据显式占位
  - 全局移除 emoji Unicode,统一替换为 lucide-react 图标组件,
    消除跨平台字体渲染差异

  **反爬检测精细化**
  - 按响应长度分层:>150KB 跳过检测(完整文献),<5KB 才扫描通用 HTTP
    错误关键字,杜绝长文献误触 Cloudflare/503 模式匹配
  - 新增 Radware Bot Manager、ShieldSquare WAF 特征识别

  **健壮性**
  - Obscura 下载校验失败后自动清理硬盘残留坏文件
  - 健康检查工具:文献已有有效 HTML 但 PDF 字段为旧报错时自动判定可修复
  - 上传接口 body limit 提升至 100MB,新增 /api/files 静态文件服务路由
  - StandardPaper 新增 has_pdf/has_html 字段区分格式级下载状态
2026-06-13 11:11:33 +08:00

62 lines
1.6 KiB
TOML

[package]
name = "astroresearch"
version = "0.1.0"
edition = "2021"
default-run = "astroresearch"
[lib]
path = "src/lib.rs"
[[bin]]
name = "astroresearch"
path = "src/main.rs"
[[bin]]
name = "health_check"
path = "src/bin/health_check.rs"
[dependencies]
tokio = { version = "1", features = ["full"] }
axum = { version = "0.7", features = ["macros", "multipart"] }
tower-http = { version = "0.5", features = ["cors", "fs", "trace"] }
sqlx = { version = "0.7", features = ["runtime-tokio-rustls", "sqlite", "chrono", "json"] }
serde = { version = "1.0", features = ["derive"] }
serde_json = "1.0"
reqwest = { version = "0.12", default-features = false, features = ["json", "stream", "multipart", "cookies", "rustls-tls"] }
dotenvy = "0.15"
quick-xml = { version = "0.31", features = ["serialize"] }
anyhow = "1.0"
thiserror = "1.0"
tracing = "0.1"
tracing-subscriber = { version = "0.3", features = ["env-filter", "json"] }
futures-util = { version = "0.3", features = ["io"] }
rand = "0.8"
regex = "1.10"
chrono = { version = "0.4", features = ["serde"] }
sha1 = "0.10"
hmac = "0.12"
base64 = "0.22"
urlencoding = "2.1"
url = "2.5"
html2md = "0.2"
flate2 = "1.1.9"
zip = "8.6.0"
uuid = { version = "1.23.2", features = ["v4"] }
tracing-appender = "0.2.5"
obscura-browser = { path = "libs/obscura/crates/obscura-browser", optional = true }
obscura-net = { path = "libs/obscura/crates/obscura-net", optional = true }
[features]
default = []
obscura-inprocess = ["dep:obscura-browser", "dep:obscura-net", "obscura-browser/stealth", "obscura-net/stealth"]
[profile.release-min]
inherits = "release"
opt-level = "s"
lto = true
codegen-units = 1
strip = true