实现 spec(2026-05-29-model-download-design)的模型分发功能: - ModelManifest: 硬编码功能文件清单 + base URL https://file.myv0.com/ - FileDownloader: URLSessionDataDelegate 分块写盘,HTTP Range 断点续传 + 大小校验 (根因修复:URL.resourceValues 会缓存文件大小,续传时先读 offset 再读 finalSize 会拿到下载前的陈旧值导致校验误判;改用 FileManager.attributesOfItem) - ModelDownloadService: @MainActor @Observable 编排逐文件下载,聚合进度/速度, 支持下载全部/暂停/重试,以及旁路文件导入 - ModelStore: 新增 fileURL/localBytes/isComplete(可注入清单)/importModel(补 VL) - ModelManagementView: 分模型卡片(状态/进度/速度) + 下载全部/暂停 + NWPathMonitor 蜂窝提示 + 从文件导入(离线兜底) - MeView: 模型管理卡改 NavigationLink + 动态状态(已就绪/下载中/N就绪) 测试(Swift Testing): Manifest 清单/字节数、Store 路径/校验/导入、 DownloadState、FileDownloader(URLProtocol mock:下载/Range续传/大小校验) Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
60 lines
2.7 KiB
Swift
60 lines
2.7 KiB
Swift
import Foundation
|
|
|
|
/// 模型文件清单中的一项:相对模型目录的路径 + 预期字节数(用于总进度计算与下载后大小校验)。
|
|
struct ModelFile: Equatable, Sendable {
|
|
let path: String
|
|
let bytes: Int
|
|
}
|
|
|
|
/// 硬编码的模型文件清单与下载源。
|
|
/// 只列加载必需的功能文件,排除 README.md / .gitattributes(省下载)。
|
|
/// 字节数与服务器素材逐一核对一致,见
|
|
/// docs/superpowers/specs/2026-05-29-model-download-design.md 附录 A。
|
|
enum ModelManifest {
|
|
/// 自建 Caddy 静态服务(用户自建 HTTPS 反代)。
|
|
/// 备选纯 IP(需 App 端 ATS 例外): http://101.132.124.52:5244/
|
|
static let baseURL = URL(string: "https://file.myv0.com/")!
|
|
|
|
static func files(for kind: ModelKind) -> [ModelFile] {
|
|
switch kind {
|
|
case .llm:
|
|
return [
|
|
ModelFile(path: "config.json", bytes: 937),
|
|
ModelFile(path: "model.safetensors", bytes: 968_080_210),
|
|
ModelFile(path: "model.safetensors.index.json", bytes: 49_731),
|
|
ModelFile(path: "tokenizer.json", bytes: 11_422_654),
|
|
ModelFile(path: "tokenizer_config.json", bytes: 9_706),
|
|
ModelFile(path: "vocab.json", bytes: 2_776_833),
|
|
ModelFile(path: "merges.txt", bytes: 1_671_853),
|
|
ModelFile(path: "special_tokens_map.json", bytes: 613),
|
|
ModelFile(path: "added_tokens.json", bytes: 707),
|
|
]
|
|
case .vl:
|
|
return [
|
|
ModelFile(path: "config.json", bytes: 1_659),
|
|
ModelFile(path: "model.safetensors", bytes: 3_073_720_461),
|
|
ModelFile(path: "model.safetensors.index.json", bytes: 108_307),
|
|
ModelFile(path: "tokenizer.json", bytes: 11_421_896),
|
|
ModelFile(path: "tokenizer_config.json", bytes: 7_256),
|
|
ModelFile(path: "vocab.json", bytes: 2_776_833),
|
|
ModelFile(path: "merges.txt", bytes: 1_671_853),
|
|
ModelFile(path: "special_tokens_map.json", bytes: 613),
|
|
ModelFile(path: "added_tokens.json", bytes: 605),
|
|
ModelFile(path: "chat_template.json", bytes: 1_050),
|
|
ModelFile(path: "preprocessor_config.json", bytes: 350),
|
|
]
|
|
}
|
|
}
|
|
|
|
static func totalBytes(for kind: ModelKind) -> Int {
|
|
files(for: kind).reduce(0) { $0 + $1.bytes }
|
|
}
|
|
|
|
/// 单个文件的下载 URL = baseURL / <仓库名> / <相对路径>。
|
|
static func fileURL(for kind: ModelKind, file: ModelFile) -> URL {
|
|
baseURL
|
|
.appendingPathComponent(kind.rawValue, isDirectory: true)
|
|
.appendingPathComponent(file.path)
|
|
}
|
|
}
|