Skip to content

LLM Wiki is a cross-platform desktop application that turns your documents into an organized, interlinked knowledge base — automatically. Instead of traditional RAG (retrieve-and-answer from scratch every time), the LLM incrementally builds and maintains a persistent wiki from your sources。

TypeScript

Release 0.4.10

nash_suea16fda6ea80
diff --git a/src-tauri/src/commands/file_sync.rs b/src-tauri/src/commands/file_sync.rs
index 0bfd75d8..c34750dc 100644
--- a/src-tauri/src/commands/file_sync.rs
+++ b/src-tauri/src/commands/file_sync.rs
@@ -104,6 +104,13 @@ pub struct FileChangeQueue {
tasks: Vec<FileChangeTask>,
}
+#[derive(Debug, Default, Clone, Serialize, Deserialize)]
+#[serde(rename_all = "camelCase")]
+pub struct FileChangeRescanResult {
+ queue: FileChangeQueue,
+ changed_tasks: Vec<FileChangeTask>,
+}
+
#[derive(Debug, Clone, Serialize, Deserialize)]
#[serde(rename_all = "camelCase")]
pub struct SourceWatchConfig {
@@ -324,16 +331,19 @@ pub fn rescan_project_files(
project_id: String,
project_path: String,
source_watch_config: Option<SourceWatchConfig>,
-) -> Result<FileChangeQueue, String> {
+) -> Result<FileChangeRescanResult, String> {
run_guarded("rescan_project_files", || {
let root = PathBuf::from(project_path);
let source_watch_config = normalize_source_watch_config(source_watch_config);
ensure_sync_dir(&root)?;
enqueue_rescan_changes(&root, &project_id, &source_watch_config)?;
- process_queue(&app, &root, &project_id)?;
+ let changed_tasks = process_queue(&app, &root, &project_id)?;
let queue = with_queue_lock(&root, || read_queue(&root))?;
emit_queue(&app, &project_id, &queue);
- Ok(queue)
+ Ok(FileChangeRescanResult {
+ queue,
+ changed_tasks,
+ })
})
}
@@ -746,7 +756,7 @@ fn upsert_task(
});
}
-fn process_queue(app: &AppHandle, root: &Path, project_id: &str) -> Result<(), String> {
+fn process_queue(app: &AppHandle, root: &Path, project_id: &str) -> Result<Vec<FileChangeTask>, String> {
process_queue_inner(
root,
project_id,
@@ -760,8 +770,9 @@ fn process_queue_inner(
project_id: &str,
mut on_queue: impl FnMut(&FileChangeQueue),
mut on_changed: impl FnMut(Vec<FileChangeTask>),
-) -> Result<(), String> {
+) -> Result<Vec<FileChangeTask>, String> {
let mut changed_tasks = Vec::<FileChangeTask>::new();
+ let mut all_changed_tasks = Vec::<FileChangeTask>::new();
let mut processed_since_emit = 0_usize;
let mut emitted_processing = false;
loop {
@@ -796,7 +807,7 @@ fn process_queue_inner(
};
on_changed(changed_tasks);
on_queue(&queue);
- return Ok(());
+ return Ok(all_changed_tasks);
};
if !emitted_processing {
emitted_processing = true;
@@ -835,6 +846,7 @@ fn process_queue_inner(
}
}
current.updated_at = now_ms();
+ all_changed_tasks.push(task.clone());
changed_tasks.push(task.clone());
processed_since_emit += 1;
if processed_since_emit >= QUEUE_EMIT_EVERY {
diff --git a/src/commands/file-sync.ts b/src/commands/file-sync.ts
index b902150e..05822491 100644
--- a/src/commands/file-sync.ts
+++ b/src/commands/file-sync.ts
@@ -27,6 +27,11 @@ export interface FileChangeQueue {
tasks: FileChangeTask[]
}
+export interface FileChangeRescanResult {
+ queue: FileChangeQueue
+ changedTasks: FileChangeTask[]
+}
+
export interface FileSyncPayload {
projectId: string
tasks: FileChangeTask[]
@@ -52,8 +57,8 @@ export function rescanProjectFiles(
projectId: string,
projectPath: string,
sourceWatchConfig?: SourceWatchConfig,
-): Promise<FileChangeQueue> {
- return invoke<FileChangeQueue>("rescan_project_files", {
+): Promise<FileChangeRescanResult> {
+ return invoke<FileChangeRescanResult>("rescan_project_files", {
projectId,
projectPath,
sourceWatchConfig: normalizeSourceWatchConfig(sourceWatchConfig),
diff --git a/src/components/layout/activity-panel.tsx b/src/components/layout/activity-panel.tsx
index 0a639de3..7f27ac41 100644
--- a/src/components/layout/activity-panel.tsx
+++ b/src/components/layout/activity-panel.tsx
@@ -96,8 +96,8 @@ export function ActivityPanel() {
const handleFileSyncRescan = useCallback(() => {
if (!project) return
rescanProjectFiles(project.id, normalizePath(project.path), useWikiStore.getState().sourceWatchConfig)
- .then((queue) => {
- setFileSyncTasks(queue.tasks)
+ .then((result) => {
+ setFileSyncTasks(result.queue.tasks)
useFileSyncStore.getState().setLastError(null)
})
.catch((err) => useFileSyncStore.getState().setLastError(String(err)))
diff --git a/src/components/settings/sections/embedding-section.tsx b/src/components/settings/sections/embedding-section.tsx
index 35f66f8d..2b521cd9 100644
--- a/src/components/settings/sections/embedding-section.tsx
+++ b/src/components/settings/sections/embedding-section.tsx
@@ -23,6 +23,14 @@ type ReindexState =
| { kind: "running"; done: number; total: number }
| { kind: "done"; count: number }
+function parsePositiveInteger(value: string): number | undefined {
+ const trimmed = value.trim()
+ if (trimmed === "") return undefined
+ const n = Number(trimmed)
+ if (!Number.isFinite(n) || n <= 0) return undefined
+ return Math.floor(n)
+}
+
export function EmbeddingSection({ draft, setDraft }: Props) {
const { t } = useTranslation()
const project = useWikiStore((s) => s.project)
@@ -113,6 +121,9 @@ export function EmbeddingSection({ draft, setDraft }: Props) {
onChange={(e) => setDraft("embeddingEndpoint", e.target.value)}
placeholder="http://127.0.0.1:1234/v1/embeddings"
/>
+ <p className="text-xs text-muted-foreground">
+ {t("settings.sections.embedding.endpointHint")}
+ </p>
</div>
<div className="space-y-2">
@@ -130,10 +141,27 @@ export function EmbeddingSection({ draft, setDraft }: Props) {
<Input
value={draft.embeddingModel}
onChange={(e) => setDraft("embeddingModel", e.target.value)}
- placeholder="e.g. text-embedding-qwen3-embedding-0.6b"
+ placeholder="e.g. text-embedding-qwen3-embedding-0.6b or gemini-embedding-001"
/>
</div>
+ <div className="space-y-2">
+ <Label>{t("settings.sections.embedding.outputDimensionality")}</Label>
+ <Input
+ type="number"
+ min={1}
+ step={1}
+ value={draft.embeddingOutputDimensionality ?? ""}
+ onChange={(e) => {
+ setDraft("embeddingOutputDimensionality", parsePositiveInteger(e.target.value))
+ }}
+ placeholder="768"
+ />
+ <p className="text-xs text-muted-foreground">
+ {t("settings.sections.embedding.outputDimensionalityHint")}
+ </p>
+ </div>
+
<div className="space-y-3 rounded-md border p-3">
<div className="text-sm font-medium">
{t("settings.sections.embedding.chunking")}
diff --git a/src/components/settings/sections/llm-provider-section.tsx b/src/components/settings/sections/llm-provider-section.tsx
index 005a163d..516aa657 100644
--- a/src/components/settings/sections/llm-provider-section.tsx
+++ b/src/components/settings/sections/llm-provider-section.tsx
@@ -187,7 +187,7 @@ function PresetRow({
: "border-muted-foreground/30 bg-muted-foreground/20 hover:bg-muted-foreground/30"
}`}
title={isActive ? t("settings.sections.llm.toggleOff") : t("settings.sections.llm.toggleOn")}
- aria-label={isActive ? "Deactivate" : "Activate"}
+ aria-label={isActive ? t("settings.sections.llm.deactivate") : t("settings.sections.llm.activate")}
>
<span
className={`inline-block h-3.5 w-3.5 rounded-full bg-white shadow-sm ring-1 ring-black/10 transition-transform ${
@@ -202,7 +202,7 @@ function PresetRow({
<div className="space-y-4 border-t bg-background/50 px-4 py-3">
{preset.provider === "custom" && (
<div className="space-y-2">
- <Label>API Mode</Label>
+ <Label>{t("settings.sections.llm.apiMode")}</Label>
<div className="flex flex-wrap gap-2">
{(
[
@@ -254,7 +254,7 @@ function PresetRow({
{needsApiKey && (
<div className="space-y-2">
- <Label>API Key</Label>
+ <Label>{t("settings.apiKey")}</Label>
<Input
type="password"
value={apiKey}
@@ -269,7 +269,7 @@ function PresetRow({
)}
<div className="space-y-2">
- <Label>Model</Label>
+ <Label>{t("settings.model")}</Label>
<ModelPicker
value={model}
suggestions={preset.suggestedModels ?? []}
@@ -279,7 +279,7 @@ function PresetRow({
</div>
<div className="space-y-2">
- <Label>Context window</Label>
+ <Label>{t("settings.sections.llm.contextWindow")}</Label>
<ContextSizeSelector
value={context}
onChange={(v) => onChange({ maxContextSize: v })}
@@ -303,19 +303,20 @@ function ReasoningControls({
value: ReasoningConfig
onChange: (value: ReasoningConfig) => void
}) {
+ const { t } = useTranslation()
const modes: { value: ReasoningMode; label: string }[] = [
- { value: "auto", label: "Auto" },
- { value: "off", label: "Off" },
- { value: "low", label: "Low" },
- { value: "medium", label: "Medium" },
- { value: "high", label: "High" },
- { value: "max", label: "Max" },
- { value: "custom", label: "Custom" },
+ { value: "auto", label: t("settings.sections.llm.reasoning.auto") },
+ { value: "off", label: t("settings.sections.llm.reasoning.off") },
+ { value: "low", label: t("settings.sections.llm.reasoning.low") },
+ { value: "medium", label: t("settings.sections.llm.reasoning.medium") },
+ { value: "high", label: t("settings.sections.llm.reasoning.high") },
+ { value: "max", label: t("settings.sections.llm.reasoning.max") },
+ { value: "custom", label: t("settings.sections.llm.reasoning.custom") },
]
return (
<div className="space-y-2">
- <Label>Reasoning / thinking</Label>
+ <Label>{t("settings.sections.llm.reasoning.title")}</Label>
<div className="flex flex-wrap gap-1.5">
{modes.map((m) => {
const active = value.mode === m.value
@@ -352,11 +353,13 @@ function ReasoningControls({
}}
placeholder="1024"
/>
- <span className="text-xs text-muted-foreground">thinking budget tokens</span>
+ <span className="text-xs text-muted-foreground">
+ {t("settings.sections.llm.reasoning.budgetTokens")}
+ </span>
</div>
)}
<p className="text-xs text-muted-foreground">
- Structured tasks such as ingest may override this to Off to avoid reasoning-only responses.
+ {t("settings.sections.llm.reasoning.hint")}
</p>
</div>
)
@@ -389,7 +392,7 @@ function EndpointField({ value, mode, placeholder, onChange }: EndpointFieldProp
return (
<div className="space-y-1.5">
- <Label>Endpoint</Label>
+ <Label>{t("settings.sections.llm.endpoint")}</Label>
<Input
value={value}
onChange={(e) => onChange(e.target.value)}
@@ -447,6 +450,7 @@ interface ModelPickerProps {
* `suggestedModels` render the input alone.
*/
function ModelPicker({ value, suggestions, placeholder, onChange }: ModelPickerProps) {
+ const { t } = useTranslation()
const hasSuggestions = suggestions.length > 0
const isCustom = hasSuggestions && value.length > 0 && !suggestions.includes(value)
@@ -466,7 +470,7 @@ function ModelPicker({ value, suggestions, placeholder, onChange }: ModelPickerP
? "border-primary bg-primary text-primary-foreground"
: "border-border bg-background hover:bg-accent hover:text-accent-foreground"
}`}
- title={`Use ${m}`}
+ title={t("settings.sections.llm.useModel", { model: m })}
>
{m}
</button>
@@ -480,9 +484,11 @@ function ModelPicker({ value, suggestions, placeholder, onChange }: ModelPickerP
? "border-primary/60 bg-primary/10 text-primary"
: "border-dashed border-muted-foreground/40 text-muted-foreground hover:bg-accent hover:text-accent-foreground"
}`}
- title="Type a custom model id"
+ title={t("settings.sections.llm.typeCustomModel")}
>
- {isCustom ? `Custom: ${value}` : "Custom…"}
+ {isCustom
+ ? t("settings.sections.llm.customModelBadge", { model: value })
+ : t("settings.sections.llm.customModel")}
</button>
</div>
)}
diff --git a/src/components/settings/sections/web-search-section.tsx b/src/components/settings/sections/web-search-section.tsx
index 5dba3638..efc1c174 100644
--- a/src/components/settings/sections/web-search-section.tsx
+++ b/src/components/settings/sections/web-search-section.tsx
@@ -96,7 +96,7 @@ export function WebSearchSection() {
type="button"
onClick={() => setExpanded((prev) => ({ ...prev, [provider.id]: !prev[provider.id] }))}
className="shrink-0 rounded p-0.5 text-muted-foreground hover:bg-accent"
- title={isExpanded ? "Collapse" : "Expand"}
+ title={isExpanded ? t("settings.sections.webSearch.collapse") : t("settings.sections.webSearch.expand")}
>
{isExpanded ? <ChevronDown className="h-4 w-4" /> : <ChevronRight className="h-4 w-4" />}
</button>
@@ -110,16 +110,18 @@ export function WebSearchSection() {
<span className="truncate text-sm font-medium">{provider.label}</span>
{hasConfig && !isActive && (
<span className="shrink-0 rounded-full bg-muted px-1.5 py-0.5 text-[10px] text-muted-foreground">
- Configured
+ {t("settings.sections.webSearch.configuredBadge")}
</span>
)}
{isActive && (
<span className="shrink-0 rounded-full bg-primary/20 px-1.5 py-0.5 text-[10px] font-medium text-primary">
- Active
+ {t("settings.sections.webSearch.activeBadge")}
</span>
)}
{savedId === provider.id && (
- <span className="shrink-0 text-[10px] text-emerald-600">Saved</span>
+ <span className="shrink-0 text-[10px] text-emerald-600">
+ {t("settings.sections.webSearch.savedBadge")}
+ </span>
)}
</div>
<div className="mt-0.5 truncate text-xs text-muted-foreground">
@@ -135,7 +137,7 @@ export function WebSearchSection() {
? "border-primary bg-primary"
: "border-muted-foreground/30 bg-muted-foreground/20 hover:bg-muted-foreground/30"
}`}
- aria-label={isActive ? "Deactivate" : "Activate"}
+ aria-label={isActive ? t("settings.sections.webSearch.deactivate") : t("settings.sections.webSearch.activate")}
>
<span
className={`inline-block h-3.5 w-3.5 rounded-full bg-white shadow-sm ring-1 ring-black/10 transition-transform ${
@@ -149,7 +151,7 @@ export function WebSearchSection() {
<div className="space-y-4 border-t bg-background/50 px-4 py-3">
{provider.needsApiKey ? (
<div className="space-y-2">
- <Label>API Key</Label>
+ <Label>{t("settings.apiKey")}</Label>
<Input
type="password"
value={override?.apiKey ?? ""}
@@ -159,14 +161,14 @@ export function WebSearchSection() {
</div>
) : (
<div className="space-y-2">
- <Label>Instance URL</Label>
+ <Label>{t("settings.sections.webSearch.instanceUrl")}</Label>
<Input
value={override?.searXngUrl ?? resolvedConfig.searXngUrl ?? ""}
onChange={(e) => updateProvider("searxng", { searXngUrl: e.target.value })}
placeholder={provider.urlPlaceholder}
/>
<p className="text-xs text-muted-foreground">
- The instance must allow JSON search responses (`format=json`).
+ {t("settings.sections.webSearch.searxngJsonHint")}
</p>
</div>
)}
@@ -201,6 +203,7 @@ function SearXngCategoryPicker({
value: string[]
onChange: (value: string[]) => void
}) {
+ const { t } = useTranslation()
const selected = value.length > 0 ? value : ["general"]
function toggle(category: string) {
@@ -212,7 +215,7 @@ function SearXngCategoryPicker({
return (
<div className="space-y-2">
- <Label>Search categories</Label>
+ <Label>{t("settings.sections.webSearch.searchCategories")}</Label>
<div className="flex flex-wrap gap-1.5">
{SEARXNG_CATEGORY_OPTIONS.map((category) => (
<button
@@ -231,7 +234,7 @@ function SearXngCategoryPicker({
))}
</div>
<p className="text-xs text-muted-foreground">
- Categories are sent as the SearXNG `categories` parameter.
+ {t("settings.sections.webSearch.searxngCategoriesHint")}
</p>
</div>
)
@@ -244,11 +247,12 @@ function SerpApiEnginePicker({
value: string
onChange: (value: string) => void
}) {
+ const { t } = useTranslation()
const isCustom = value.length > 0 && !SERPAPI_ENGINE_OPTIONS.some((e) => e.value === value)
return (
<div className="space-y-2">
- <Label>Search engine / category</Label>
+ <Label>{t("settings.sections.webSearch.searchEngine")}</Label>
<div className="flex flex-wrap gap-1.5">
{SERPAPI_ENGINE_OPTIONS.map((engine) => (
<button
@@ -269,11 +273,11 @@ function SerpApiEnginePicker({
<Input
value={value}
onChange={(e) => onChange(e.target.value)}
- placeholder="Custom SerpApi engine, e.g. google_finance"
+ placeholder={t("settings.sections.webSearch.customSerpApiPlaceholder")}
/>
{isCustom && (
<p className="text-xs text-muted-foreground">
- Custom engine will be sent as the SerpApi `engine` parameter.
+ {t("settings.sections.webSearch.customSerpApiHint")}
</p>
)}
</div>
diff --git a/src/components/settings/settings-types.ts b/src/components/settings/settings-types.ts
index f3e568c2..3f73525c 100644
--- a/src/components/settings/settings-types.ts
+++ b/src/components/settings/settings-types.ts
@@ -23,6 +23,8 @@ export interface SettingsDraft {
embeddingEndpoint: string
embeddingApiKey: string
embeddingModel: string
+ /** Optional Gemini native output_dimensionality. Empty = provider default. */
+ embeddingOutputDimensionality: number | undefined
/** Target characters per chunk. Empty = use chunker default (1000). */
embeddingMaxChunkChars: number | undefined
/** Overlap characters between adjacent chunks. Empty = default (200). */
diff --git a/src/components/settings/settings-view.tsx b/src/components/settings/settings-view.tsx
index f5ab4a3f..d1af73ed 100644
--- a/src/components/settings/settings-view.tsx
+++ b/src/components/settings/settings-view.tsx
@@ -110,6 +110,7 @@ function initialDraft(
embeddingEndpoint: embed.endpoint,
embeddingApiKey: embed.apiKey,
embeddingModel: embed.model,
+ embeddingOutputDimensionality: embed.outputDimensionality,
embeddingMaxChunkChars: embed.maxChunkChars,
embeddingOverlapChunkChars: embed.overlapChunkChars,
multimodalEnabled: multimodal.enabled,
@@ -263,6 +264,7 @@ export function SettingsView() {
endpoint: draft.embeddingEndpoint,
apiKey: draft.embeddingApiKey,
model: draft.embeddingModel,
+ outputDimensionality: draft.embeddingOutputDimensionality,
maxChunkChars: draft.embeddingMaxChunkChars,
overlapChunkChars: draft.embeddingOverlapChunkChars,
}
diff --git a/src/components/sources/sources-view.tsx b/src/components/sources/sources-view.tsx
index c4730f56..e4605fc3 100644
--- a/src/components/sources/sources-view.tsx
+++ b/src/components/sources/sources-view.tsx
@@ -3,13 +3,14 @@ import { open } from "@tauri-apps/plugin-dialog"
import { Plus, FileText, RefreshCw, BookOpen, Trash2, Folder, ChevronRight, ChevronDown } from "lucide-react"
import { Button } from "@/components/ui/button"
import { ScrollArea } from "@/components/ui/scroll-area"
+import { Tooltip, TooltipContent, TooltipProvider, TooltipTrigger } from "@/components/ui/tooltip"
import { useWikiStore } from "@/stores/wiki-store"
import { listDirectory, readFile } from "@/commands/fs"
import type { FileNode } from "@/types/wiki"
import { useTranslation } from "react-i18next"
import { normalizePath } from "@/lib/path-utils"
import { decideDeleteClick } from "@/lib/sources-tree-delete"
-import { rescanProjectFiles } from "@/commands/file-sync"
+import { rescanProjectFileSync } from "@/lib/project-file-sync"
import {
deleteSourceFile,
deleteSourceFolder,
@@ -33,6 +34,7 @@ export function SourcesView() {
const [sources, setSources] = useState<FileNode[]>([])
const [importing, setImporting] = useState(false)
const [ingestingPath, setIngestingPath] = useState<string | null>(null)
+ const [refreshing, setRefreshing] = useState(false)
const [refreshError, setRefreshError] = useState<string | null>(null)
/**
* Path of the source-tree node currently in "click again to
@@ -77,15 +79,18 @@ export function SourcesView() {
}, [loadSources, dataVersion])
async function handleRefreshSources() {
- if (!project) return
- const pp = normalizePath(project.path)
+ if (!project || refreshing) return
+ setRefreshing(true)
try {
- await rescanProjectFiles(project.id, pp, useWikiStore.getState().sourceWatchConfig)
+ await rescanProjectFileSync(project, useWikiStore.getState().sourceWatchConfig)
+ setRefreshError(null)
} catch (err) {
console.warn("[sources] failed to rescan project files:", err)
setRefreshError(String(err))
+ } finally {
+ await loadSources()
+ setRefreshing(false)
}
- await loadSources()
}
async function handleImport() {
@@ -252,13 +257,29 @@ export function SourcesView() {
}
return (
- <div className="flex h-full flex-col">
+ <TooltipProvider delay={300}>
+ <div className="flex h-full flex-col">
<div className="flex items-center justify-between border-b px-4 py-3">
<h2 className="text-sm font-semibold">{t("sources.title")}</h2>
<div className="flex gap-1">
- <Button variant="ghost" size="icon" onClick={handleRefreshSources} title={t("sources.refresh")}>
- <RefreshCw className="h-4 w-4" />
- </Button>
+ <Tooltip>
+ <TooltipTrigger
+ render={
+ <Button
+ variant="ghost"
+ size="icon"
+ onClick={handleRefreshSources}
+ disabled={refreshing}
+ aria-label={t("sources.refreshFolder")}
+ />
+ }
+ >
+ <RefreshCw className={`h-4 w-4 ${refreshing ? "animate-spin" : ""}`} />
+ </TooltipTrigger>
+ <TooltipContent side="bottom" align="end" className="max-w-80 whitespace-normal leading-relaxed">
+ {t("sources.refreshFolderTooltip")}
+ </TooltipContent>
+ </Tooltip>
<Button size="sm" onClick={handleImport} disabled={importing}>
<Plus className="mr-1 h-4 w-4" />
{importing ? t("sources.importing") : t("sources.import")}
@@ -310,10 +331,30 @@ export function SourcesView() {
)}
</ScrollArea>
- <div className="border-t px-4 py-2 text-xs text-muted-foreground">
- {t("sources.sourceCount", { count: countFiles(sources) })}
+ <div className="flex items-center justify-between gap-2 border-t px-4 py-2 text-xs text-muted-foreground">
+ <span>{t("sources.sourceCount", { count: countFiles(sources) })}</span>
+ <Tooltip>
+ <TooltipTrigger
+ render={
+ <Button
+ variant="ghost"
+ size="sm"
+ onClick={handleRefreshSources}
+ disabled={!project || refreshing}
+ className="h-7 px-2 text-xs"
+ />
+ }
+ >
+ <RefreshCw className={`mr-1.5 h-3.5 w-3.5 ${refreshing ? "animate-spin" : ""}`} />
+ {refreshing ? t("sources.refreshingFolder") : t("sources.refreshFolder")}
+ </TooltipTrigger>
+ <TooltipContent side="top" align="end" className="max-w-80 whitespace-normal leading-relaxed">
+ {t("sources.refreshFolderTooltip")}
+ </TooltipContent>
+ </Tooltip>
+ </div>
</div>
- </div>
+ </TooltipProvider>
)
}
diff --git a/src/i18n/en.json b/src/i18n/en.json
index 5795ca4b..80c4f436 100644
--- a/src/i18n/en.json
+++ b/src/i18n/en.json
@@ -80,6 +80,10 @@
"importSourceFiles": "Import Source Files",
"importSourceFolder": "Import Source Folder",
"refresh": "Refresh",
+ "refreshFolder": "Refresh folder",
+ "refreshingFolder": "Refreshing...",
+ "refreshFolderHint": "Rescan the source folder and process new, changed, or deleted files.",
+ "refreshFolderTooltip": "Rescan the project source folder with the same rules as automatic monitoring. New or modified allowed files are queued for ingest; deleted sources trigger the normal wiki and index cleanup; then the file tree is rebuilt.",
"refreshFailed": "Failed to refresh sources: {{error}}",
"ingest": "Ingest",
"delete": "Delete source and related wiki pages",
@@ -194,6 +198,11 @@
"collapse": "Collapse",
"toggleOff": "Click to deactivate",
"toggleOn": "Click to activate (turns off other providers)",
+ "deactivate": "Deactivate",
+ "activate": "Activate",
+ "apiMode": "API Mode",
+ "endpoint": "Endpoint",
+ "contextWindow": "Context window",
"wireOpenAi": "OpenAI-compat",
"wireAnthropic": "Anthropic-compat",
"apiKeyPlaceholderCustom": "Leave empty if no key (local models)",
@@ -201,18 +210,37 @@
"savedBadge": "saved",
"configuredBadge": "configured",
"activeBadge": "● active",
+ "useModel": "Use {{model}}",
+ "typeCustomModel": "Type a custom model id",
+ "customModel": "Custom…",
+ "customModelBadge": "Custom: {{model}}",
"endpointPreviewWillUse": "Will resolve to",
- "endpointPreviewAutoApply": "(applied automatically when you leave the input)"
+ "endpointPreviewAutoApply": "(applied automatically when you leave the input)",
+ "reasoning": {
+ "title": "Reasoning / thinking",
+ "auto": "Auto",
+ "off": "Off",
+ "low": "Low",
+ "medium": "Medium",
+ "high": "High",
+ "max": "Max",
+ "custom": "Custom",
+ "budgetTokens": "thinking budget tokens",
+ "hint": "Structured tasks such as ingest may override this to Off to avoid reasoning-only responses."
+ }
},
"embedding": {
"title": "Vector Embeddings",
- "description": "Enable semantic search against your wiki. Works with any OpenAI-compatible /v1/embeddings endpoint. Local models (LM Studio, Ollama embedding) don't need an API key.",
+ "description": "Enable semantic search against your wiki. Works with OpenAI-compatible /v1/embeddings endpoints and Google Gemini native embedContent endpoints. Local models (LM Studio, Ollama embedding) don't need an API key.",
"enableLabel": "Enable vector search",
"enableHint": "When off, search uses token matching only.",
"endpoint": "Endpoint",
+ "endpointHint": "OpenAI-compatible example: http://127.0.0.1:1234/v1/embeddings. Gemini example: https://generativelanguage.googleapis.com/v1beta with model gemini-embedding-001.",
"apiKey": "API Key (optional)",
"apiKeyPlaceholder": "Leave empty for local models",
"model": "Model",
+ "outputDimensionality": "Output dimensionality (Gemini only)",
+ "outputDimensionalityHint": "Optional. Sent only for Gemini native embedContent endpoints as `output_dimensionality`; OpenAI-compatible endpoints ignore this field. Leave empty for the model default. Rebuild embeddings after changing model or dimension.",
"chunking": "Chunking",
"maxChunkChars": "Max chunk size (characters)",
"maxChunkCharsHint": "Target size of each chunk. Lower this (e.g. 500) when your endpoint's context is small (llama.cpp n_ctx=512, mxbai-embed-large, etc.).",
@@ -257,7 +285,21 @@
},
"webSearch": {
"title": "Web Search",
- "description": "Used by Deep Research to pull fresh external context."
+ "description": "Used by Deep Research to pull fresh external context.",
+ "expand": "Expand",
+ "collapse": "Collapse",
+ "activate": "Activate",
+ "deactivate": "Deactivate",
+ "configuredBadge": "Configured",
+ "activeBadge": "Active",
+ "savedBadge": "Saved",
+ "instanceUrl": "Instance URL",
+ "searxngJsonHint": "The instance must allow JSON search responses (`format=json`).",
+ "searchCategories": "Search categories",
+ "searxngCategoriesHint": "Categories are sent as the SearXNG `categories` parameter.",
+ "searchEngine": "Search engine / category",
+ "customSerpApiPlaceholder": "Custom SerpApi engine, e.g. google_finance",
+ "customSerpApiHint": "Custom engine will be sent as the SerpApi `engine` parameter."
},
"network": {
"title": "Network",
diff --git a/src/i18n/zh.json b/src/i18n/zh.json
index 6ea4f97f..fce47791 100644
--- a/src/i18n/zh.json
+++ b/src/i18n/zh.json
@@ -80,6 +80,10 @@
"importSourceFiles": "导入源文件",
"importSourceFolder": "导入源文件夹",
"refresh": "刷新",
+ "refreshFolder": "刷新文件夹",
+ "refreshingFolder": "刷新中...",
+ "refreshFolderHint": "重新扫描资料文件夹,并处理新增、修改或删除的文件。",
+ "refreshFolderTooltip": "按自动监控相同规则重新扫描当前项目的资料文件夹。允许的新增或修改文件会加入提取队列;已删除的资料会走正常的 Wiki 与索引清理流程;随后重建文件树。",
"refreshFailed": "刷新原始资料失败:{{error}}",
"ingest": "提取到 Wiki",
"delete": "删除资料及相关 Wiki 页面",
@@ -194,6 +198,11 @@
"collapse": "收起",
"toggleOff": "点击关闭",
"toggleOn": "点击启用(会关闭其他 provider)",
+ "deactivate": "停用",
+ "activate": "启用",
+ "apiMode": "API 模式",
+ "endpoint": "Endpoint",
+ "contextWindow": "上下文窗口",
"wireOpenAi": "OpenAI 兼容",
"wireAnthropic": "Anthropic 兼容",
"apiKeyPlaceholderCustom": "无 Key 可留空(本地模型)",
@@ -201,18 +210,37 @@
"savedBadge": "已保存",
"configuredBadge": "已配置",
"activeBadge": "● 活跃",
+ "useModel": "使用 {{model}}",
+ "typeCustomModel": "输入自定义模型 ID",
+ "customModel": "自定义…",
+ "customModelBadge": "自定义:{{model}}",
"endpointPreviewWillUse": "将使用",
- "endpointPreviewAutoApply": "(离开输入框时自动套用)"
+ "endpointPreviewAutoApply": "(离开输入框时自动套用)",
+ "reasoning": {
+ "title": "Reasoning / thinking",
+ "auto": "自动",
+ "off": "关闭",
+ "low": "低",
+ "medium": "中",
+ "high": "高",
+ "max": "最大",
+ "custom": "自定义",
+ "budgetTokens": "thinking token 预算",
+ "hint": "结构化任务(如 ingest)可能会自动关闭 reasoning,避免模型只输出思考过程而没有正文。"
+ }
},
"embedding": {
"title": "向量嵌入",
- "description": "启用语义搜索。对任何 OpenAI 兼容的 /v1/embeddings endpoint 工作。本地模型(如 LM Studio、Ollama embedding)不需要 API Key。",
+ "description": "启用语义搜索。支持 OpenAI 兼容的 /v1/embeddings endpoint,也支持 Google Gemini 原生 embedContent endpoint。本地模型(如 LM Studio、Ollama embedding)不需要 API Key。",
"enableLabel": "启用向量搜索",
"enableHint": "关掉后搜索只走 token 匹配。",
"endpoint": "Endpoint",
+ "endpointHint": "OpenAI 兼容示例:http://127.0.0.1:1234/v1/embeddings。Gemini 示例:https://generativelanguage.googleapis.com/v1beta,模型填 gemini-embedding-001。",
"apiKey": "API Key(可选)",
"apiKeyPlaceholder": "本地模型留空即可",
"model": "Model",
+ "outputDimensionality": "输出维度(仅 Gemini)",
+ "outputDimensionalityHint": "可选。仅对 Gemini 原生 embedContent endpoint 作为 `output_dimensionality` 发送;OpenAI 兼容 endpoint 会忽略此项。留空则使用模型默认值。修改模型或维度后需要重建嵌入。",
"chunking": "切分参数",
"maxChunkChars": "每块最大字符数",
"maxChunkCharsHint": "单个 chunk 的目标长度。endpoint 上下文较小时(llama.cpp n_ctx=512、mxbai-embed-large 等)调低(如 500)。",
@@ -257,7 +285,21 @@
},
"webSearch": {
"title": "网页搜索",
- "description": "Deep Research 会用它拉取最新的外部上下文。"
+ "description": "Deep Research 会用它拉取最新的外部上下文。",
+ "expand": "展开",
+ "collapse": "收起",
+ "activate": "启用",
+ "deactivate": "停用",
+ "configuredBadge": "已配置",
+ "activeBadge": "活跃",
+ "savedBadge": "已保存",
+ "instanceUrl": "实例 URL",
+ "searxngJsonHint": "实例必须允许 JSON 搜索响应(`format=json`)。",
+ "searchCategories": "搜索分类",
+ "searxngCategoriesHint": "分类会作为 SearXNG 的 `categories` 参数发送。",
+ "searchEngine": "搜索引擎 / 分类",
+ "customSerpApiPlaceholder": "自定义 SerpApi engine,例如 google_finance",
+ "customSerpApiHint": "自定义 engine 会作为 SerpApi 的 `engine` 参数发送。"
},
"network": {
"title": "网络",
diff --git a/src/lib/changelog.ts b/src/lib/changelog.ts
index 90908cbd..b64a71be 100644
--- a/src/lib/changelog.ts
+++ b/src/lib/changelog.ts
@@ -30,10 +30,12 @@ export const CHANGELOG: ChangelogEntry[] = [
date: "2026-05-14",
highlights: {
en: [
- "Source folder monitoring now has configurable file type and exclusion rules. Existing projects should review Settings → Source Watch after upgrading because automatic ingest now defaults to common document formats only.",
+ "Added configurable source folder monitoring, manual source-folder refresh, and Gemini native embeddings support.",
+ "Fixed source sync, embedding provider compatibility, and settings localization issues.",
],
zh: [
- "资料文件夹监控新增文件类型和排除规则配置。升级后建议到「设置 → 资料文件夹监控」检查配置;自动提取现在默认只纳入常见文档格式。",
+ "新增可配置的资料文件夹监控、手动刷新资料文件夹,以及 Gemini 原生向量嵌入支持。",
+ "修复资料同步、向量 provider 兼容性和设置页本地化相关问题。",
],
},
},
diff --git a/src/lib/embedding.test.ts b/src/lib/embedding.test.ts
index fd085594..be8dda87 100644
--- a/src/lib/embedding.test.ts
+++ b/src/lib/embedding.test.ts
@@ -36,6 +36,7 @@ vi.mock("@/commands/fs", () => ({
import {
searchByEmbedding,
+ fetchEmbedding,
embedPage,
embedAllPages,
getLastEmbeddingError,
@@ -227,6 +228,389 @@ describe("searchByEmbedding — aggregation", () => {
})
})
+describe("fetchEmbedding — provider wire formats", () => {
+ it("does not route OpenAI text-embedding models through Gemini", async () => {
+ mockHttpFetch.mockResolvedValueOnce(okResponse([0.1, 0.2]))
+
+ const out = await fetchEmbedding("hi", {
+ enabled: true,
+ endpoint: "https://api.openai.com/v1/embeddings",
+ apiKey: "sk-test",
+ model: "text-embedding-3-small",
+ })
+
+ expect(out).toEqual([0.1, 0.2])
+ const [url, opts] = mockHttpFetch.mock.calls[0]
+ expect(url).toBe("https://api.openai.com/v1/embeddings")
+ expect((opts?.headers as Record<string, string>).Authorization).toBe("Bearer sk-test")
+ expect((opts?.headers as Record<string, string>)["x-goog-api-key"]).toBeUndefined()
+ expect(JSON.parse(String(opts?.body))).toEqual({
+ model: "text-embedding-3-small",
+ input: "hi",
+ })
+ })
+
+ it("does not route LM Studio text-embedding models through Gemini", async () => {
+ mockHttpFetch.mockResolvedValueOnce(okResponse([0.3, 0.4]))
+
+ const out = await fetchEmbedding("hi", {
+ enabled: true,
+ endpoint: "http://127.0.0.1:1234/v1/embeddings",
+ apiKey: "",
+ model: "text-embedding-qwen3-embedding-0.6b",
+ })
+
+ expect(out).toEqual([0.3, 0.4])
+ const [url, opts] = mockHttpFetch.mock.calls[0]
+ expect(url).toBe("http://127.0.0.1:1234/v1/embeddings")
+ expect((opts?.headers as Record<string, string>).Authorization).toBeUndefined()
+ expect((opts?.headers as Record<string, string>)["x-goog-api-key"]).toBeUndefined()
+ expect(JSON.parse(String(opts?.body))).toEqual({
+ model: "text-embedding-qwen3-embedding-0.6b",
+ input: "hi",
+ })
+ })
+
+ it("supports Gemini native embedContent endpoint and response shape", async () => {
+ mockHttpFetch.mockResolvedValueOnce(
+ new Response(JSON.stringify({ embedding: { values: [0.1, 0.2, 0.3] } }), {
+ status: 200,
+ headers: { "Content-Type": "application/json" },
+ }),
+ )
+
+ const out = await fetchEmbedding("hello", {
+ enabled: true,
+ endpoint: "https://generativelanguage.googleapis.com/v1beta",
+ apiKey: "g-key",
+ model: "gemini-embedding-001",
+ })
+
+ expect(out).toEqual([0.1, 0.2, 0.3])
+ expect(mockHttpFetch).toHaveBeenCalledTimes(1)
+ const [url, opts] = mockHttpFetch.mock.calls[0]
+ expect(url).toBe("https://generativelanguage.googleapis.com/v1beta/models/gemini-embedding-001:embedContent")
+ expect((opts?.headers as Record<string, string>)["x-goog-api-key"]).toBe("g-key")
+ expect((opts?.headers as Record<string, string>).Authorization).toBeUndefined()
+ expect(JSON.parse(String(opts?.body))).toEqual({
+ model: "models/gemini-embedding-001",
+ content: { parts: [{ text: "hello" }] },
+ })
+ })
+
+ it("accepts a full Gemini model embedContent endpoint", async () => {
+ mockHttpFetch.mockResolvedValueOnce(
+ new Response(JSON.stringify({ embedding: { values: [1, 2] } }), {
+ status: 200,
+ headers: { "Content-Type": "application/json" },
+ }),
+ )
+
+ const out = await fetchEmbedding("hello", {
+ enabled: true,
+ endpoint: "https://generativelanguage.googleapis.com/v1beta/models/text-embedding-004:embedContent",
+ apiKey: "g-key",
+ model: "text-embedding-004",
+ })
+
+ expect(out).toEqual([1, 2])
+ expect(mockHttpFetch.mock.calls[0][0]).toBe(
+ "https://generativelanguage.googleapis.com/v1beta/models/text-embedding-004:embedContent",
+ )
+ })
+
+ it("preserves non-key query parameters on pasted Gemini endpoints", async () => {
+ mockHttpFetch.mockResolvedValueOnce(
+ new Response(JSON.stringify({ embedding: { values: [1] } }), {
+ status: 200,
+ headers: { "Content-Type": "application/json" },
+ }),
+ )
+
+ await fetchEmbedding("hello", {
+ enabled: true,
+ endpoint: "https://generativelanguage.googleapis.com/v1beta/models/gemini-embedding-2:embedContent?key=url-key&other=foo",
+ apiKey: "header-key",
+ model: "gemini-embedding-2",
+ })
+
+ expect(mockHttpFetch.mock.calls[0][0]).toBe(
+ "https://generativelanguage.googleapis.com/v1beta/models/gemini-embedding-2:embedContent?other=foo",
+ )
+ })
+
+ it("routes custom embedContent proxy endpoints through Gemini format", async () => {
+ mockHttpFetch.mockResolvedValueOnce(
+ new Response(JSON.stringify({ embedding: { values: [0.7] } }), {
+ status: 200,
+ headers: { "Content-Type": "application/json" },
+ }),
+ )
+
+ const out = await fetchEmbedding("hello", {
+ enabled: true,
+ endpoint: "https://proxy.example.com/google/models/gemini-embedding-2:embedContent",
+ apiKey: "g-key",
+ model: "gemini-embedding-2",
+ })
+
+ expect(out).toEqual([0.7])
+ const [url, opts] = mockHttpFetch.mock.calls[0]
+ expect(url).toBe("https://proxy.example.com/google/models/gemini-embedding-2:embedContent")
+ expect((opts?.headers as Record<string, string>)["x-goog-api-key"]).toBe("g-key")
+ expect(JSON.parse(String(opts?.body))).toEqual({
+ model: "models/gemini-embedding-2",
+ content: { parts: [{ text: "hello" }] },
+ })
+ })
+
+ it("trims trailing slashes before building Gemini embedContent endpoint", async () => {
+ mockHttpFetch.mockResolvedValueOnce(
+ new Response(JSON.stringify({ embedding: { values: [0.8] } }), {
+ status: 200,
+ headers: { "Content-Type": "application/json" },
+ }),
+ )
+
+ await fetchEmbedding("hello", {
+ enabled: true,
+ endpoint: "https://generativelanguage.googleapis.com/v1beta///",
+ apiKey: "g-key",
+ model: "gemini-embedding-2",
+ })
+
+ expect(mockHttpFetch.mock.calls[0][0]).toBe(
+ "https://generativelanguage.googleapis.com/v1beta/models/gemini-embedding-2:embedContent",
+ )
+ })
+
+ it("sends Gemini output_dimensionality when configured", async () => {
+ mockHttpFetch.mockResolvedValueOnce(
+ new Response(JSON.stringify({ embedding: { values: [0.1, 0.2] } }), {
+ status: 200,
+ headers: { "Content-Type": "application/json" },
+ }),
+ )
+
+ const out = await fetchEmbedding("What is the meaning of life?", {
+ enabled: true,
+ endpoint: "https://generativelanguage.googleapis.com/v1beta",
+ apiKey: "g-key",
+ model: "gemini-embedding-2",
+ outputDimensionality: 768,
+ })
+
+ expect(out).toEqual([0.1, 0.2])
+ const [, opts] = mockHttpFetch.mock.calls[0]
+ expect(JSON.parse(String(opts?.body))).toEqual({
+ model: "models/gemini-embedding-2",
+ content: { parts: [{ text: "What is the meaning of life?" }] },
+ output_dimensionality: 768,
+ })
+ })
+
+ it("omits invalid Gemini output_dimensionality values", async () => {
+ const values = [0, -1, Number.NaN]
+ for (const value of values) {
+ mockHttpFetch.mockReset()
+ mockHttpFetch.mockResolvedValueOnce(
+ new Response(JSON.stringify({ embedding: { values: [0.1] } }), {
+ status: 200,
+ headers: { "Content-Type": "application/json" },
+ }),
+ )
+
+ await fetchEmbedding("hello", {
+ enabled: true,
+ endpoint: "https://generativelanguage.googleapis.com/v1beta",
+ apiKey: "g-key",
+ model: "gemini-embedding-2",
+ outputDimensionality: value,
+ })
+
+ expect(JSON.parse(String(mockHttpFetch.mock.calls[0][1]?.body))).toEqual({
+ model: "models/gemini-embedding-2",
+ content: { parts: [{ text: "hello" }] },
+ })
+ }
+ })
+
+ it("floors fractional Gemini output_dimensionality values", async () => {
+ mockHttpFetch.mockResolvedValueOnce(
+ new Response(JSON.stringify({ embedding: { values: [0.1] } }), {
+ status: 200,
+ headers: { "Content-Type": "application/json" },
+ }),
+ )
+
+ await fetchEmbedding("hello", {
+ enabled: true,
+ endpoint: "https://generativelanguage.googleapis.com/v1beta",
+ apiKey: "g-key",
+ model: "gemini-embedding-2",
+ outputDimensionality: 1.5,
+ })
+
+ expect(JSON.parse(String(mockHttpFetch.mock.calls[0][1]?.body))).toEqual({
+ model: "models/gemini-embedding-2",
+ content: { parts: [{ text: "hello" }] },
+ output_dimensionality: 1,
+ })
+ })
+
+ it("normalizes a pasted Gemini batch endpoint to single embedContent", async () => {
+ mockHttpFetch.mockResolvedValueOnce(
+ new Response(JSON.stringify({ embedding: { values: [0.4] } }), {
+ status: 200,
+ headers: { "Content-Type": "application/json" },
+ }),
+ )
+
+ const out = await fetchEmbedding("hello", {
+ enabled: true,
+ endpoint: "https://generativelanguage.googleapis.com/v1beta/models/gemini-embedding-2:batchEmbedContents",
+ apiKey: "g-key",
+ model: "gemini-embedding-2",
+ })
+
+ expect(out).toEqual([0.4])
+ expect(mockHttpFetch.mock.calls[0][0]).toBe(
+ "https://generativelanguage.googleapis.com/v1beta/models/gemini-embedding-2:embedContent",
+ )
+ })
+
+ it("strips pasted Gemini key query parameters from the endpoint", async () => {
+ mockHttpFetch.mockResolvedValueOnce(
+ new Response(JSON.stringify({ embedding: { values: [0.5] } }), {
+ status: 200,
+ headers: { "Content-Type": "application/json" },
+ }),
+ )
+
+ const out = await fetchEmbedding("hello", {
+ enabled: true,
+ endpoint: "https://generativelanguage.googleapis.com/v1beta/models/gemini-embedding-2:embedContent?key=url-key",
+ apiKey: "header-key",
+ model: "gemini-embedding-2",
+ })
+
+ expect(out).toEqual([0.5])
+ const [url, opts] = mockHttpFetch.mock.calls[0]
+ expect(url).toBe("https://generativelanguage.googleapis.com/v1beta/models/gemini-embedding-2:embedContent")
+ expect((opts?.headers as Record<string, string>)["x-goog-api-key"]).toBe("header-key")
+ })
+
+ it("rejects empty Gemini embedding vectors", async () => {
+ mockHttpFetch.mockResolvedValueOnce(
+ new Response(JSON.stringify({ embedding: { values: [] } }), {
+ status: 200,
+ headers: { "Content-Type": "application/json" },
+ }),
+ )
+
+ const out = await fetchEmbedding("hello", {
+ enabled: true,
+ endpoint: "https://generativelanguage.googleapis.com/v1beta",
+ apiKey: "g-key",
+ model: "gemini-embedding-2",
+ })
+
+ expect(out).toBeNull()
+ expect(getLastEmbeddingError()).toContain("missing embedding.values")
+ })
+
+ it("rejects Gemini embedding vectors with NaN or Infinity values", async () => {
+ const values = [[Number.NaN], [Number.POSITIVE_INFINITY]]
+ for (const vector of values) {
+ mockHttpFetch.mockReset()
+ mockHttpFetch.mockResolvedValueOnce(
+ new Response(JSON.stringify({ embedding: { values: vector } }), {
+ status: 200,
+ headers: { "Content-Type": "application/json" },
+ }),
+ )
+
+ const out = await fetchEmbedding("hello", {
+ enabled: true,
+ endpoint: "https://generativelanguage.googleapis.com/v1beta",
+ apiKey: "g-key",
+ model: "gemini-embedding-2",
+ })
+
+ expect(out).toBeNull()
+ expect(getLastEmbeddingError()).toContain("missing embedding.values")
+ }
+ })
+
+ it("rejects Gemini embedding vectors with mixed value types", async () => {
+ mockHttpFetch.mockResolvedValueOnce(
+ new Response(JSON.stringify({ embedding: { values: [1, "x"] } }), {
+ status: 200,
+ headers: { "Content-Type": "application/json" },
+ }),
+ )
+
+ const out = await fetchEmbedding("hello", {
+ enabled: true,
+ endpoint: "https://generativelanguage.googleapis.com/v1beta",
+ apiKey: "g-key",
+ model: "gemini-embedding-2",
+ })
+
+ expect(out).toBeNull()
+ expect(getLastEmbeddingError()).toContain("missing embedding.values")
+ })
+
+ it("surfaces Gemini auth errors with HTTP status", async () => {
+ mockHttpFetch.mockResolvedValueOnce(
+ new Response(JSON.stringify({ error: { message: "API key not valid" } }), {
+ status: 403,
+ statusText: "Forbidden",
+ }),
+ )
+
+ const out = await fetchEmbedding("hello", {
+ enabled: true,
+ endpoint: "https://generativelanguage.googleapis.com/v1beta",
+ apiKey: "bad-key",
+ model: "gemini-embedding-2",
+ })
+
+ expect(out).toBeNull()
+ expect(getLastEmbeddingError()).toContain("API 403 Forbidden")
+ expect(getLastEmbeddingError()).toContain("API key not valid")
+ })
+
+ it("auto-halves Gemini requests after an oversize error", async () => {
+ mockHttpFetch
+ .mockResolvedValueOnce(
+ new Response(JSON.stringify({ error: { message: "input length exceeds context" } }), {
+ status: 400,
+ statusText: "Bad Request",
+ }),
+ )
+ .mockResolvedValueOnce(
+ new Response(JSON.stringify({ embedding: { values: [0.1] } }), {
+ status: 200,
+ headers: { "Content-Type": "application/json" },
+ }),
+ )
+
+ const out = await fetchEmbedding("a".repeat(200), {
+ enabled: true,
+ endpoint: "https://generativelanguage.googleapis.com/v1beta",
+ apiKey: "g-key",
+ model: "gemini-embedding-2",
+ })
+
+ expect(out).toEqual([0.1])
+ const firstBody = JSON.parse(String(mockHttpFetch.mock.calls[0][1]?.body))
+ const secondBody = JSON.parse(String(mockHttpFetch.mock.calls[1][1]?.body))
+ expect(firstBody.content.parts[0].text).toHaveLength(200)
+ expect(secondBody.content.parts[0].text).toHaveLength(100)
+ })
+})
+
// ── fetchEmbedding auto-halve — via embedPage/searchByEmbedding ─────
describe("fetchEmbedding (via searchByEmbedding) — auto-halve", () => {
diff --git a/src/lib/embedding.ts b/src/lib/embedding.ts
index 8b12746b..b918008b 100644
--- a/src/lib/embedding.ts
+++ b/src/lib/embedding.ts
@@ -84,8 +84,16 @@ export async function fetchEmbedding(
): Promise<number[] | null> {
if (!cfg.endpoint) return null
+ const isGoogleNative = isGoogleEmbeddingConfig(cfg)
+ const endpoint = isGoogleNative ? googleEmbeddingEndpoint(cfg) : cfg.endpoint
const headers: Record<string, string> = { "Content-Type": "application/json" }
- if (cfg.apiKey) headers.Authorization = `Bearer ${cfg.apiKey}`
+ if (cfg.apiKey) {
+ if (isGoogleNative) {
+ headers["x-goog-api-key"] = cfg.apiKey
+ } else {
+ headers.Authorization = `Bearer ${cfg.apiKey}`
+ }
+ }
let current = text
let attempts = 0
@@ -93,20 +101,27 @@ export async function fetchEmbedding(
attempts++
try {
const httpFetch = await getHttpFetch()
- const resp = await httpFetch(cfg.endpoint, {
+ const resp = await httpFetch(endpoint, {
method: "POST",
headers,
- body: JSON.stringify({ model: cfg.model, input: current }),
+ body: JSON.stringify(
+ isGoogleNative
+ ? googleEmbeddingBody(cfg.model, current, cfg.outputDimensionality)
+ : { model: cfg.model, input: current },
+ ),
})
if (resp.ok) {
const data = await resp.json()
- const embedding = data?.data?.[0]?.embedding ?? null
- if (embedding) {
+ const embedding = isGoogleNative
+ ? data?.embedding?.values ?? null
+ : data?.data?.[0]?.embedding ?? null
+ if (isNonEmptyNumberArray(embedding)) {
lastEmbeddingError = null
return embedding
}
- lastEmbeddingError = `Embedding response missing data[0].embedding (got ${JSON.stringify(data).slice(0, 200)})`
+ const expectedShape = isGoogleNative ? "embedding.values" : "data[0].embedding"
+ lastEmbeddingError = `Embedding response missing ${expectedShape} (got ${JSON.stringify(data).slice(0, 200)})`
console.warn(`[Embedding] ${lastEmbeddingError}`)
return null
}
@@ -139,12 +154,12 @@ export async function fetchEmbedding(
}
// Non-oversize definitive failure (auth, rate limit, server down, …).
- lastEmbeddingError = `API ${resp.status} ${resp.statusText}${bodyText ? ` — ${bodyText.slice(0, 200)}` : ""} at ${cfg.endpoint}`
+ lastEmbeddingError = `API ${resp.status} ${resp.statusText}${bodyText ? ` — ${bodyText.slice(0, 200)}` : ""} at ${endpoint}`
console.warn(`[Embedding] ${lastEmbeddingError}`)
return null
} catch (err) {
if (isFetchNetworkError(err)) {
- lastEmbeddingError = `Network error reaching ${cfg.endpoint}. Check endpoint URL, API key, and connectivity.`
+ lastEmbeddingError = `Network error reaching ${endpoint}. Check endpoint URL, API key, and connectivity.`
} else {
lastEmbeddingError = err instanceof Error ? err.message : String(err)
}
@@ -160,6 +175,68 @@ export async function fetchEmbedding(
return null
}
+function isNonEmptyNumberArray(value: unknown): value is number[] {
+ return Array.isArray(value)
+ && value.length > 0
+ && value.every((item) => typeof item === "number" && Number.isFinite(item))
+}
+
+function isGoogleEmbeddingConfig(cfg: EmbeddingConfig): boolean {
+ const endpoint = cfg.endpoint.toLowerCase()
+ return endpoint.includes("generativelanguage.googleapis.com")
+ || /:embedcontent(\?|$)/i.test(endpoint)
+}
+
+function googleEmbeddingEndpoint(cfg: EmbeddingConfig): string {
+ const raw = stripGoogleApiKeyQuery(cfg.endpoint.trim()).replace(/\/+$/, "")
+ if (/:batchEmbedContents(\?|$)/i.test(raw)) {
+ return raw.replace(/:batchEmbedContents/i, ":embedContent")
+ }
+ if (/:embedContent(\?|$)/i.test(raw)) return raw
+
+ const modelPath = googleModelPath(cfg.model)
+ if (/\/models\/[^/?]+$/i.test(raw)) {
+ return `${raw}:embedContent`
+ }
+ return `${raw}/models/${encodeURIComponent(modelPath.replace(/^models\//, ""))}:embedContent`
+}
+
+function stripGoogleApiKeyQuery(endpoint: string): string {
+ if (!endpoint.includes("?")) return endpoint
+ try {
+ const url = new URL(endpoint)
+ url.searchParams.delete("key")
+ return url.toString()
+ } catch {
+ return endpoint.replace(/([?&])key=[^&]*&?/i, (_, prefix: string) => prefix === "?" ? "?" : "&")
+ .replace(/[?&]$/, "")
+ .replace("?&", "?")
+ }
+}
+
+function googleModelPath(model: string): string {
+ const trimmed = model.trim()
+ if (trimmed.startsWith("models/")) return trimmed
+ return `models/${trimmed}`
+}
+
+function googleEmbeddingBody(
+ model: string,
+ text: string,
+ outputDimensionality?: number,
+): Record<string, unknown> {
+ const body: Record<string, unknown> = {
+ model: googleModelPath(model),
+ content: {
+ parts: [{ text }],
+ },
+ }
+ if (typeof outputDimensionality === "number" && Number.isFinite(outputDimensionality) && outputDimensionality > 0) {
+ body.output_dimensionality = Math.floor(outputDimensionality)
+ }
+ return body
+}
+
// ── LanceDB v2 operations (via Rust Tauri commands) ──────────────────────
interface ChunkUpsertInput {
diff --git a/src/lib/project-file-sync.test.ts b/src/lib/project-file-sync.test.ts
index 6cc514ab..9ef51492 100644
--- a/src/lib/project-file-sync.test.ts
+++ b/src/lib/project-file-sync.test.ts
@@ -12,6 +12,42 @@ const mocks = vi.hoisted(() => {
}),
emit: (event: string, payload: unknown) => listeners[event]?.({ payload }),
stopProjectFileWatcher: vi.fn(async () => undefined),
+ rescanProjectFiles: vi.fn(async (projectId: string): Promise<{
+ queue: {
+ version: number
+ tasks: Array<{
+ id: string
+ projectId: string
+ path: string
+ kind: "created" | "modified" | "deleted"
+ status: "pending" | "processing" | "done" | "failed" | "superseded"
+ createdAt: number
+ updatedAt: number
+ retryCount: number
+ needsRerun: boolean
+ }>
+ }
+ changedTasks: Array<{
+ id: string
+ projectId: string
+ path: string
+ kind: "created" | "modified" | "deleted"
+ status: "pending" | "processing" | "done" | "failed" | "superseded"
+ createdAt: number
+ updatedAt: number
+ retryCount: number
+ needsRerun: boolean
+ }>
+ }> => {
+ void projectId
+ return {
+ queue: {
+ version: 1,
+ tasks: [],
+ },
+ changedTasks: [],
+ }
+ }),
startProjectFileWatcher: vi.fn(() => new Promise((resolve) => {
resolvers.push(resolve)
})),
@@ -58,6 +94,7 @@ vi.mock("@tauri-apps/api/event", () => ({
}))
vi.mock("@/commands/file-sync", () => ({
+ rescanProjectFiles: mocks.rescanProjectFiles,
startProjectFileWatcher: mocks.startProjectFileWatcher,
stopProjectFileWatcher: mocks.stopProjectFileWatcher,
}))
@@ -92,6 +129,16 @@ describe("project file sync", () => {
vi.useRealTimers()
vi.clearAllMocks()
mocks.clearResolvers()
+ mocks.rescanProjectFiles.mockImplementation(async (projectId: string) => {
+ void projectId
+ return {
+ queue: {
+ version: 1,
+ tasks: [],
+ },
+ changedTasks: [],
+ }
+ })
mocks.listDirectory.mockImplementation(async (_path?: string) => [])
mocks.readFile.mockImplementation(async (_path?: string) => "")
mocks.getFileSize.mockImplementation(async (_path?: string) => 1024)
@@ -235,6 +282,150 @@ describe("project file sync", () => {
expect(mocks.enqueueBatch).not.toHaveBeenCalled()
})
+ it("manual rescan uses the same source ingest flow when the watcher is stopped", async () => {
+ const { rescanProjectFileSync } = await import("@/lib/project-file-sync")
+ const { useWikiStore } = await import("@/stores/wiki-store")
+
+ const project = { id: "A", name: "A", path: "/tmp/a" }
+ useWikiStore.getState().setProject(project)
+ mocks.rescanProjectFiles.mockImplementation(async (projectId: string) => {
+ return {
+ queue: { version: 1, tasks: [] },
+ changedTasks: [
+ {
+ id: "t1",
+ projectId,
+ path: "raw/sources/manual.pdf",
+ kind: "created",
+ status: "done",
+ createdAt: 1,
+ updatedAt: 1,
+ retryCount: 0,
+ needsRerun: false,
+ },
+ ],
+ }
+ })
+
+ await rescanProjectFileSync(project)
+
+ expect(mocks.rescanProjectFiles).toHaveBeenCalledWith(
+ "A",
+ "/tmp/a",
+ expect.objectContaining({ enabled: true, autoIngest: true }),
+ )
+ expect(mocks.enqueueBatch).toHaveBeenCalledWith("A", [
+ { sourcePath: "raw/sources/manual.pdf", folderContext: "" },
+ ])
+ })
+
+ it("manual rescan uses returned changed tasks while the watcher is running", async () => {
+ const { startProjectFileSync, rescanProjectFileSync } = await import("@/lib/project-file-sync")
+ const { useWikiStore } = await import("@/stores/wiki-store")
+
+ const project = { id: "A", name: "A", path: "/tmp/a" }
+ useWikiStore.getState().setProject(project)
+ void startProjectFileSync(project)
+ await vi.waitFor(() => {
+ expect(mocks.listen).toHaveBeenCalledTimes(2)
+ })
+
+ mocks.rescanProjectFiles.mockImplementation(async (projectId: string) => ({
+ queue: { version: 1, tasks: [] },
+ changedTasks: [
+ {
+ id: "t1",
+ projectId,
+ path: "raw/sources/watcher-running.pdf",
+ kind: "created",
+ status: "done",
+ createdAt: 1,
+ updatedAt: 1,
+ retryCount: 0,
+ needsRerun: false,
+ },
+ ],
+ }))
+
+ await rescanProjectFileSync(project)
+
+ expect(mocks.enqueueBatch).toHaveBeenCalledWith("A", [
+ { sourcePath: "raw/sources/watcher-running.pdf", folderContext: "" },
+ ])
+ })
+
+ it("manual rescan ignores changed tasks after project switch", async () => {
+ const { rescanProjectFileSync } = await import("@/lib/project-file-sync")
+ const { useWikiStore } = await import("@/stores/wiki-store")
+
+ const projectA = { id: "A", name: "A", path: "/tmp/a" }
+ const projectB = { id: "B", name: "B", path: "/tmp/b" }
+ useWikiStore.getState().setProject(projectA)
+ mocks.rescanProjectFiles.mockImplementation(async (projectId: string) => {
+ useWikiStore.getState().setProject(projectB)
+ return {
+ queue: { version: 1, tasks: [] },
+ changedTasks: [
+ {
+ id: "t1",
+ projectId,
+ path: "raw/sources/stale.pdf",
+ kind: "created",
+ status: "done",
+ createdAt: 1,
+ updatedAt: 1,
+ retryCount: 0,
+ needsRerun: false,
+ },
+ ],
+ }
+ })
+
+ await rescanProjectFileSync(projectA)
+
+ expect(mocks.enqueueBatch).not.toHaveBeenCalled()
+ })
+
+ it("manual rescan refreshes the file tree when no files changed", async () => {
+ const { rescanProjectFileSync } = await import("@/lib/project-file-sync")
+ const { useWikiStore } = await import("@/stores/wiki-store")
+
+ const project = { id: "A", name: "A", path: "/tmp/a" }
+ useWikiStore.getState().setProject(project)
+
+ await rescanProjectFileSync(project)
+
+ expect(mocks.listDirectory).toHaveBeenCalledWith("/tmp/a")
+ })
+
+ it("manual rescan respects auto ingest disabled", async () => {
+ const { rescanProjectFileSync } = await import("@/lib/project-file-sync")
+ const { useWikiStore } = await import("@/stores/wiki-store")
+
+ const project = { id: "A", name: "A", path: "/tmp/a" }
+ useWikiStore.getState().setProject(project)
+ mocks.rescanProjectFiles.mockImplementation(async (projectId: string) => ({
+ queue: { version: 1, tasks: [] },
+ changedTasks: [
+ {
+ id: "t1",
+ projectId,
+ path: "raw/sources/manual.pdf",
+ kind: "created",
+ status: "done",
+ createdAt: 1,
+ updatedAt: 1,
+ retryCount: 0,
+ needsRerun: false,
+ },
+ ],
+ }))
+
+ await rescanProjectFileSync(project, { autoIngest: false } as never)
+
+ expect(mocks.enqueueBatch).not.toHaveBeenCalled()
+ })
+
it("removes an externally deleted raw source from every wiki page sources field", async () => {
vi.useFakeTimers()
const { startProjectFileSync } = await import("@/lib/project-file-sync")
diff --git a/src/lib/project-file-sync.ts b/src/lib/project-file-sync.ts
index dad7445d..a31b2848 100644
--- a/src/lib/project-file-sync.ts
+++ b/src/lib/project-file-sync.ts
@@ -1,6 +1,7 @@
import { listen, type UnlistenFn } from "@tauri-apps/api/event"
import { readFile, listDirectory } from "@/commands/fs"
import {
+ rescanProjectFiles,
startProjectFileWatcher,
stopProjectFileWatcher,
type FileSyncPayload,
@@ -87,6 +88,26 @@ export async function stopProjectFileSync(): Promise<void> {
}
}
+export async function rescanProjectFileSync(
+ project: WikiProject,
+ sourceWatchConfig?: SourceWatchConfig,
+): Promise<void> {
+ const config = normalizeSourceWatchConfig(sourceWatchConfig ?? useWikiStore.getState().sourceWatchConfig)
+ activeSourceWatchConfig = config
+
+ const result = await rescanProjectFiles(project.id, normalizePath(project.path), config)
+ if (useWikiStore.getState().project?.id !== project.id) return
+ useFileSyncStore.getState().setTasks(result.queue.tasks)
+
+ if (useWikiStore.getState().project?.id !== project.id) return
+ if (result.changedTasks.length > 0) {
+ const paths = [...new Set(result.changedTasks.map((task) => task.path))]
+ await processFileChangeBatch(project, paths, result.changedTasks)
+ } else {
+ await refreshAfterFileChanges(project, [])
+ }
+}
+
function scheduleRefreshAfterFileChanges(tasks: FileChangeTask[]): void {
for (const task of tasks) {
pendingRefreshPaths.add(task.path)
@@ -147,7 +168,7 @@ async function refreshAfterFileChanges(project: WikiProject, relativePaths: stri
}
async function enqueueRawSourceChanges(project: WikiProject, tasks: FileChangeTask[]): Promise<void> {
- const config = normalizeSourceWatchConfig(useWikiStore.getState().sourceWatchConfig ?? activeSourceWatchConfig)
+ const config = normalizeSourceWatchConfig(activeSourceWatchConfig)
if (!config.enabled || !config.autoIngest) return
const candidates = tasks
diff --git a/src/stores/wiki-store.ts b/src/stores/wiki-store.ts
index b1be2ed0..b19a5894 100644
--- a/src/stores/wiki-store.ts
+++ b/src/stores/wiki-store.ts
@@ -75,6 +75,8 @@ interface EmbeddingConfig {
endpoint: string // e.g. "http://127.0.0.1:1234/v1/embeddings"
apiKey: string
model: string // e.g. "text-embedding-qwen3-embedding-0.6b"
+ /** Optional Gemini native `output_dimensionality` value. Ignored by OpenAI-compatible endpoints. */
+ outputDimensionality?: number
/**
* Chunking knobs (Phase 1 RAG). Undefined values fall back to the
* chunker's built-in defaults in `src/lib/text-chunker.ts`: