dsh-free-search 0.9.8 → 0.9.9

package/README.md

@@ -472,7 +472,7 @@ Search engine test:
 
 #### 离线自检断言（tools/）
 
-仓库自带 **19 个**零依赖断言脚本（合计 **528 项**），**不需要 API key、不联网、不需要测试框架**，Node ≥ 20 即可跑：
+仓库自带 **20 个**零依赖断言脚本（合计 **534 项**），**不需要 API key、不联网、不需要测试框架**，Node ≥ 20 即可跑：
 
 | 脚本 | 覆盖 | 运行（项数） |
 |---|---|---|
@@ -492,6 +492,7 @@ Search engine test:
 | `assert-killtree-win32.mjs` | Windows 子进程树终止（#74）：不再依赖 `taskkill` | `node tools/assert-killtree-win32.mjs`（8） |
 | `assert-killtree-portable.mjs` | 跨平台 `killTree` 便携版契约 | `node tools/assert-killtree-portable.mjs`（5） |
 | `assert-parallel-mcp.mjs` | Parallel MCP 引擎：25 s 超时/调用方取消覆盖**响应体读取**、共享 SSE 解析器（通知/多行帧）、6 类错误路径、REST 精确时间过滤 | `node tools/assert-parallel-mcp.mjs`（15） |
+| `assert-fetch-charset.mjs` | 抓取编码（#88）：`charset=gbk` / `gb2312` 按声明解码、无声明时嗅探 `<meta charset>` 与 http-equiv 形式、默认 UTF-8、未知 label 兜底 UTF-8 不抛错 | `node tools/assert-fetch-charset.mjs`（6） |
 | `assert-fetch-abort.mjs` | 抓取中断（#87）：压缩响应中途断连必须 reject（不再永久 pending）、明文断连同样 reject、abort 能打断停滞的 gzip 流、complete 响应与重定向仍正常 | `node tools/assert-fetch-abort.mjs`（6） |
 | `assert-bridge-access.mjs` | 桥接 Host 围栏（#84）：loopback 放行、remote peer 打 loopback Host 拒绝、`--trusted-host`（无端口匹配任意端口 / 带端口精确匹配）、`--public-url`、绑定地址、跨站与异源拒绝 | `node tools/assert-bridge-access.mjs`（11） |
 | `assert-provider-takeover.mjs` | provider 接管：旧属性 API 与 rc.2 的 `config.*` ref 两套兼容、`settings.mutate(web)` 单批写入、显式指定的 provider 不被抢占、`fetchProviderTakeover: false` 时保留显式 `http`、describe 噪声不误报、嵌套 config 变化与 web 重建分别成行、关闭 DSH 搜索开关后重新接管且不抢 `deepseek-search-api`、apply 重入/重复注册不再中断、重复日志去重 | `node tools/assert-provider-takeover.mjs`（16） |
@@ -987,7 +988,7 @@ Restart `dsh web` after switching to apply changes.
 
 #### Offline self-check assertions (`tools/`)
 
-**Nineteen** zero-dependency assertion scripts ship with the repo (**528 assertions** total): **no API keys, no network, no test framework**, Node ≥ 20:
+**Twenty** zero-dependency assertion scripts ship with the repo (**534 assertions** total): **no API keys, no network, no test framework**, Node ≥ 20:
 
 | Script | Covers | Run (count) |
 |---|---|---|
@@ -1007,6 +1008,7 @@ Restart `dsh web` after switching to apply changes.
 | `assert-killtree-win32.mjs` | Windows process-tree termination (#74): no longer relies on `taskkill` | `node tools/assert-killtree-win32.mjs` (8) |
 | `assert-killtree-portable.mjs` | portable `killTree` contract | `node tools/assert-killtree-portable.mjs` (5) |
 | `assert-parallel-mcp.mjs` | Parallel MCP engine: the 25 s timeout / caller abort covers the **body read**, shared SSE parser (notifications, multiline frames), six error paths, REST precise time filtering | `node tools/assert-parallel-mcp.mjs` (15) |
+| `assert-fetch-charset.mjs` | fetch encoding (#88): `charset=gbk` / `gb2312` decoded as declared, `<meta charset>` and the http-equiv form sniffed when the header omits it, UTF-8 by default, and an unknown label falls back to UTF-8 instead of throwing | `node tools/assert-fetch-charset.mjs` (6) |
 | `assert-fetch-abort.mjs` | fetch interruption (#87): a compressed response cut mid-body rejects instead of hanging forever, identity truncation rejects too, abort breaks a stalled gzip stream, and complete responses plus redirects still work | `node tools/assert-fetch-abort.mjs` (6) |
 | `assert-bridge-access.mjs` | bridge Host fence (#84): loopback allowed, a remote peer behind a loopback Host refused, `--trusted-host` (port-less matches any port, explicit port matches exactly), `--public-url`, the bind address, cross-site and foreign-Origin refused | `node tools/assert-bridge-access.mjs` (11) |
 | `assert-provider-takeover.mjs` | provider takeover: both the legacy property API and the rc.2 `config.*` refs, a single batched `settings.mutate(web)`, explicit foreign providers left untouched, an explicit `http` kept when `fetchProviderTakeover` is off, describe()-triggered events ignored when nothing changed, nested config changes and web rebuilds reported separately, re-takeover after DSH's search switch is turned off without fighting an explicit `deepseek-search-api`, apply re-entry / duplicate registrations tolerated, and repeated logs deduplicated | `node tools/assert-provider-takeover.mjs` (16) |

package/lib/client.js

@@ -426,7 +426,7 @@ window.__ModuleLoader__.load({
     };
     const tt = (lang) => I18N[lang === "en" ? "en" : "zh"];
     // 当前插件版本（与 lib/index.js 的 PLUGIN_VERSION 保持一致；跟随上游 v0.7.5）
-    const PLUGIN_VERSION = "0.9.8";
+    const PLUGIN_VERSION = "0.9.9";
     const PLUGIN_REPO_URL = "https://github.com/DDDMUC/dsh-free-search";
     // 右侧浮动目录（#85）：条目顺序即页面顺序，anchor 是页面里对应 field 的 id。
     const NAV_ITEMS = [

package/lib/fetch_provider.js

@@ -122,6 +122,33 @@ function browserDumpDom(browserPath, url, { timeoutMs = 25000, signal } = {}) {
   });
 }
 
+// #88：按声明的 charset 解码。GBK/GB2312/Big5/Shift_JIS 老站很常见，
+// 一律当 UTF-8 会得到整页 \uFFFD，而且不会报错、不会降级，模型直接吃一页乱码。
+function parseCharset(contentType) {
+  const match = /charset\s*=\s*"?([^";\s]+)"?/i.exec(String(contentType || ""));
+  return match ? match[1].trim().toLowerCase() : "";
+}
+
+function sniffHtmlCharset(buffer) {
+  const head = buffer.subarray(0, 2048).toString("latin1");
+  let match = /<meta[^>]+charset\s*=\s*["']?\s*([\w.:-]+)/i.exec(head);
+  if (match) return match[1].toLowerCase();
+  match = /<meta[^>]+content\s*=\s*["'][^"']*charset\s*=\s*([\w.:-]+)/i.exec(head);
+  return match ? match[1].toLowerCase() : "";
+}
+
+function decodeBody(buffer, contentType) {
+  let label = parseCharset(contentType);
+  if (!label && /html|xml/i.test(String(contentType || ""))) label = sniffHtmlCharset(buffer);
+  if (!label) label = "utf-8";
+  try {
+    return new TextDecoder(label, { fatal: false }).decode(buffer);
+  } catch {
+    // 无法识别的 label（或该 Node 没编进对应编码表）：退回 UTF-8，不抛错。
+    return buffer.toString("utf8");
+  }
+}
+
 /** GET 跟随重定向（≤5 跳），返回最终 URL / 状态 / 正文 / 内容类型；支持 gzip/br/deflate 解压。 */
 export function httpsGetFollow(url, { headers = {}, timeoutMs = 20000, signal } = {}, transport = https) {
   return new Promise((resolve, reject) => {
@@ -222,7 +249,7 @@ export function httpsGetFollow(url, { headers = {}, timeoutMs = 20000, signal }
             settle(true, {
               url: current,
               status: res.statusCode,
-              text: Buffer.concat(chunks).toString("utf8"),
+              text: decodeBody(Buffer.concat(chunks), res.headers["content-type"] || ""),
               contentType: res.headers["content-type"] || "",
             })
           );

package/lib/index.js

@@ -256,7 +256,7 @@ function routeEngines(query, { timeRange, disabledEngines = [], fallbackOrder =
 }
 
 // 当前插件版本（发布时与 package.json 同步；跟随上游 v0.7.5，不在 PR 里 bump）
-const PLUGIN_VERSION = "0.9.8";
+const PLUGIN_VERSION = "0.9.9";
 // 检查更新的 npm registry 元数据地址（dsh-free-search 是 npmjs 上的公开包）
 const NPM_REGISTRY_URL = "https://registry.npmjs.org/dsh-free-search/latest";
 const PLUGIN_NPM_URL = "https://www.npmjs.com/package/dsh-free-search";

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "dsh-free-search",
-  "version": "0.9.8",
+  "version": "0.9.9",
   "description": "Free web search for DeepSeek Harness: 27 engines (Bing/DuckDuckGo/AnySearch/SearXNG/Exa/Tavily/Keenable/Firecrawl/Parallel keyless; Zhihu global/site via MCP and Perplexity/SerpBase/DeepSeek (official + native tools/search)/You.com/Baidu/Kimi/Aliyun/Doubao with key; OpenAI/Gemini/Claude model built-in search, explicit-only) + auto smart routing + global engine enable/disable & fallback priority + editable zh/en language pools + multi mode (parallel full language-pool merge) + multi_search cross-source tool + failure-aware fallback (quota/auth cooldown, transient retry, fallbackOn) + time filtering + platform search + domain blocklist (blockedDomains) + web_fetch, with web settings UI.",
   "type": "module",
   "main": "lib/index.js",