dsh-vision-recognizer - npm Package Compare versions

Comparing version 0.1.0 to 0.1.1

client/client.js

@@ -26,2 +26,3 @@ window.__ModuleLoader__.load({ id: "dsh-vision-recognizer", factory: (require) => {
   providerLabel: '供应商',
+  providerHint: '选择图片转译使用的视觉模型供应商。可在下方覆盖模型与接口地址。',
   apiKeyLabel: 'API Key',
@@ -32,9 +33,15 @@ apiKeyHint: '留空则读取该供应商的环境变量；本地 Ollama 无需 Key。',
   modelLabel: '模型',
+  modelHint: '视觉模型 ID。留空使用该供应商的默认模型。',
   modelPlaceholder: '留空使用供应商默认',
   baseUrlLabel: '接口地址',
+  baseUrlHint: 'OpenAI 兼容接口地址（…/chat/completions 会自动拼接）。留空使用该供应商的默认地址。',
   baseUrlPlaceholder: '留空使用供应商默认',
   maxTokensLabel: '最大输出 token',
+  maxTokensHint: '视觉模型单次输出的 token 上限，越大转译越详尽、也越贵。',
   timeoutLabel: '超时（毫秒）',
+  timeoutHint: '请求超时时间。本地/匿名端点会被强制限制在 20 秒内。',
   markerLabel: '转译标记',
+  markerHint: '转译文本的前缀标记，DeepSeek 靠它区分「这是图片转译的内容」。默认 [图片转译]。',
   autoOllamaLabel: '自动探测本地 Ollama',
+  autoOllamaHint: '启动时探测本机 http://localhost:11434 的 Ollama，检测到则作为兜底——图片不出本机、无需 Key。',
   save: '保存',
@@ -53,2 +60,3 @@ saving: '保存中…',
   providerLabel: 'Provider',
+  providerHint: 'Vision provider used to transcribe attached images. Model and endpoint can be overridden below.',
   apiKeyLabel: 'API key',
@@ -59,9 +67,15 @@ apiKeyHint: 'Leave blank to read the provider env var; local Ollama needs none.',
   modelLabel: 'Model',
+  modelHint: 'Vision model id. Leave blank for the provider default.',
   modelPlaceholder: 'Leave blank for the provider default',
   baseUrlLabel: 'Endpoint',
+  baseUrlHint: 'OpenAI-compatible endpoint (…/chat/completions is appended). Leave blank for the provider default.',
   baseUrlPlaceholder: 'Leave blank for the provider default',
   maxTokensLabel: 'Max output tokens',
+  maxTokensHint: 'Output token cap for the vision model; larger is more thorough but costs more.',
   timeoutLabel: 'Timeout (ms)',
+  timeoutHint: 'Request timeout. Local/anonymous endpoints are hard-capped at 20s.',
   markerLabel: 'Transcription marker',
+  markerHint: 'Text prefix prepended to each transcription, so DeepSeek can tell image transcriptions apart. Default [图片转译].',
   autoOllamaLabel: 'Auto-detect local Ollama',
+  autoOllamaHint: 'Probe local Ollama at http://localhost:11434 at startup and use it as a fallback — images never leave your machine and no key is needed.',
   save: 'Save',
@@ -97,2 +111,5 @@ saving: 'Saving…',
 .vrg-status.err{color:var(--dsw-alias-state-error-primary,#dc2626)}
+.vrg-tip{position:relative;display:inline-flex;align-items:center;justify-content:center;width:15px;height:15px;border-radius:99px;border:1px solid var(--dsw-alias-border-l2,#e5e7eb);color:var(--dsw-alias-label-tertiary,#9ca3af);font-size:10px;font-weight:600;line-height:1;cursor:help;margin-left:6px;vertical-align:middle;user-select:none}
+.vrg-tip .vrg-tiptext{position:absolute;bottom:calc(100% + 8px);left:50%;transform:translateX(-50%) translateY(-2px);background:var(--dsw-alias-bg-layer-1,#fff);border:1px solid var(--dsw-alias-border-l2,#e5e7eb);border-radius:8px;padding:8px 10px;font-size:12px;font-weight:400;line-height:1.5;color:var(--dsw-alias-label-primary,#1f2328);width:max-content;max-width:280px;white-space:normal;text-align:left;z-index:60;opacity:0;pointer-events:none;transition:opacity .12s ease,transform .12s ease;box-shadow:0 6px 20px rgba(0,0,0,.14)}
+.vrg-tip:hover .vrg-tiptext,.vrg-tip:focus-visible .vrg-tiptext{opacity:1;transform:translateX(-50%) translateY(0)}
 `
@@ -109,2 +126,9 @@
 
+/** A small "?" circle whose hover/focus reveals a tooltip explaining the field. */
+function HelpTip({ text }) {
+  return h('span', { className: 'vrg-tip', tabIndex: 0, 'aria-label': text },
+    '?',
+    h('span', { className: 'vrg-tiptext' }, text))
+}
+
 function VisionPanel({ t, locale }) {
@@ -168,3 +192,7 @@ const [data, setData] = useState(null)
     })
-      .then((res) => res.json().then((body) => ({ status: res.status, body })))
+      .then((res) => res.text().then((text) => {
+        let body = {}
+        try { body = text ? JSON.parse(text) : {} } catch { body = { error: text.slice(0, 200) } }
+        return { status: res.status, body }
+      }))
       .then(({ status: httpStatus, body }) => {
@@ -200,3 +228,3 @@ if (httpStatus === 200 && body.ok) {
     h('div', { className: 'vrg-field' },
-      h('label', { className: 'vrg-label' }, t('providerLabel')),
+      h('label', { className: 'vrg-label' }, t('providerLabel'), h(HelpTip, { text: t('providerHint') })),
       h('select', {
@@ -212,3 +240,4 @@ className: 'vrg-input',
         h('span', { className: 'vrg-badge' + (form.apiKeyConfigured ? ' on' : '') },
-          form.apiKeyConfigured ? t('apiKeyConfigured') : t('apiKeyUnset'))),
+          form.apiKeyConfigured ? t('apiKeyConfigured') : t('apiKeyUnset')),
+        h(HelpTip, { text: t('apiKeyHint') })),
       h('input', {
@@ -225,3 +254,3 @@ className: 'vrg-input',
       h('div', { className: 'vrg-field' },
-        h('label', { className: 'vrg-label' }, t('modelLabel')),
+        h('label', { className: 'vrg-label' }, t('modelLabel'), h(HelpTip, { text: t('modelHint') })),
         h('input', {
@@ -234,3 +263,3 @@ className: 'vrg-input',
       h('div', { className: 'vrg-field' },
-        h('label', { className: 'vrg-label' }, t('baseUrlLabel')),
+        h('label', { className: 'vrg-label' }, t('baseUrlLabel'), h(HelpTip, { text: t('baseUrlHint') })),
         h('input', {
@@ -244,3 +273,3 @@ className: 'vrg-input',
       h('div', { className: 'vrg-field' },
-        h('label', { className: 'vrg-label' }, t('maxTokensLabel')),
+        h('label', { className: 'vrg-label' }, t('maxTokensLabel'), h(HelpTip, { text: t('maxTokensHint') })),
         h('input', {
@@ -253,3 +282,3 @@ className: 'vrg-input',
       h('div', { className: 'vrg-field' },
-        h('label', { className: 'vrg-label' }, t('timeoutLabel')),
+        h('label', { className: 'vrg-label' }, t('timeoutLabel'), h(HelpTip, { text: t('timeoutHint') })),
         h('input', {
@@ -262,3 +291,3 @@ className: 'vrg-input',
     h('div', { className: 'vrg-field' },
-      h('label', { className: 'vrg-label' }, t('markerLabel')),
+      h('label', { className: 'vrg-label' }, t('markerLabel'), h(HelpTip, { text: t('markerHint') })),
       h('input', {
@@ -275,3 +304,4 @@ className: 'vrg-input',
       }),
-      t('autoOllamaLabel')),
+      t('autoOllamaLabel'),
+      h(HelpTip, { text: t('autoOllamaHint') })),
     form.provider === 'custom' && h('p', { className: 'vrg-hint' }, t('requiredHint')),
@@ -278,0 +308,0 @@ h('div', { className: 'vrg-actions' },

lib/index.js

@@ -157,12 +157,10 @@ /**
 
-    const disposers = [
-        hostCtx.webServer.register({
-            kind: 'exact',
-            path: '/dsh-vision-recognizer/config',
-            handler: async (request, response) => {
-                if (request.method !== 'GET') {
-                    response.writeHead(405, { allow: 'GET' });
-                    response.end();
-                    return;
-                }
+    // Single exact route; the handler dispatches by HTTP method. The webserver
+    // rejects duplicate (kind, path) registrations, so registering GET and POST
+    // as two routes would throw and leave the POST handler dead.
+    return hostCtx.webServer.register({
+        kind: 'exact',
+        path: '/dsh-vision-recognizer/config',
+        handler: async (request, response) => {
+            if (request.method === 'GET') {
                 sendJson(response, 200, {
@@ -172,13 +170,5 @@ providers: providerCatalog(),
                 });
-            },
-        }),
-        hostCtx.webServer.register({
-            kind: 'exact',
-            path: '/dsh-vision-recognizer/config',
-            handler: async (request, response) => {
-                if (request.method !== 'POST') {
-                    response.writeHead(405, { allow: 'POST' });
-                    response.end();
-                    return;
-                }
+                return;
+            }
+            if (request.method === 'POST') {
                 if (!sameOrigin(request)) {
@@ -197,8 +187,8 @@ sendJson(response, 403, { error: 'untrusted origin' });
                 }
-            },
-        }),
-    ];
-    return () => {
-        for (const dispose of disposers) dispose();
-    };
+                return;
+            }
+            response.writeHead(405, { allow: 'GET, POST' });
+            response.end();
+        },
+    });
 }
@@ -205,0 +195,0 @@

package.json

 {
   "name": "dsh-vision-recognizer",
-  "version": "0.1.0",
+  "version": "0.1.1",
   "description": "DeepSeek Harness image recognition plugin: keep DeepSeek as the conversation brain while a configurable vision model (15+ providers, OpenAI-compatible + Anthropic) transcribes attached images. Configurable from Settings → Plugins.",
@@ -5,0 +5,0 @@ "private": false,

New alerts

Long strings

Supply chain risk

Contains long string literals, which may be a sign of obfuscated or packed code.

Fixed alerts

Long strings

Supply chain risk

Contains long string literals, which may be a sign of obfuscated or packed code.

Improved metrics

Total package byte prevSize

69069

4.66%

Lines of code

1080

2.08%

No dependency changes