Initial CutScript release - Open-source AI-powered text-based video editor

CutScript is a local-first, Descript-like video editor where you edit video by editing text.
Delete a word from the transcript and it's cut from the video.

Features:
- Word-level transcription with WhisperX
- Text-based video editing with undo/redo
- AI filler word removal (Ollama/OpenAI/Claude)
- AI clip creation for shorts
- Waveform timeline with virtualized transcript
- FFmpeg stream-copy (fast) and re-encode (4K) export
- Caption burn-in and sidecar SRT generation
- Studio Sound audio enhancement (DeepFilterNet)
- Keyboard shortcuts (J/K/L, Space, Delete, Ctrl+Z/S/E)
- Encrypted API key storage
- Project save/load (.aive files)

Architecture:
- Electron + React + Tailwind (frontend)
- FastAPI + Python (backend)
- WhisperX for transcription
- FFmpeg for video processing
- Multi-provider AI support

Performance optimizations:
- RAF-throttled time updates
- Zustand selectors for granular subscriptions
- Dual-canvas waveform rendering
- Virtualized transcript with react-virtuoso

Built on top of DataAnts-AI/VideoTranscriber, completely rewritten as a desktop application.

License: MIT
This commit is contained in:
Your Name
2026-03-03 06:31:04 -05:00
parent d1e1fedcae
commit 33cca5f552
73 changed files with 7463 additions and 3906 deletions

310
frontend/src/App.tsx Normal file
View File

@ -0,0 +1,310 @@
import { useEffect, useState, useRef } from 'react';
import { useEditorStore } from './store/editorStore';
import VideoPlayer from './components/VideoPlayer';
import TranscriptEditor from './components/TranscriptEditor';
import WaveformTimeline from './components/WaveformTimeline';
import AIPanel from './components/AIPanel';
import ExportDialog from './components/ExportDialog';
import SettingsPanel from './components/SettingsPanel';
import { useKeyboardShortcuts } from './hooks/useKeyboardShortcuts';
import {
Film,
FolderOpen,
Settings,
Sparkles,
Download,
Loader2,
FolderSearch,
FileInput,
} from 'lucide-react';
const IS_ELECTRON = !!window.electronAPI;
type Panel = 'ai' | 'settings' | 'export' | null;
export default function App() {
const {
videoPath,
words,
isTranscribing,
transcriptionProgress,
loadVideo,
setBackendUrl,
setTranscription,
setTranscribing,
backendUrl,
} = useEditorStore();
const [activePanel, setActivePanel] = useState<Panel>(null);
const [manualPath, setManualPath] = useState('');
const [whisperModel, setWhisperModel] = useState('base');
const fileInputRef = useRef<HTMLInputElement>(null);
useKeyboardShortcuts();
useEffect(() => {
if (IS_ELECTRON) {
window.electronAPI!.getBackendUrl().then(setBackendUrl);
}
}, [setBackendUrl]);
const handleLoadProject = async () => {
if (!IS_ELECTRON) return;
try {
const projectPath = await window.electronAPI!.openProject();
if (!projectPath) return;
const content = await window.electronAPI!.readFile(projectPath);
const data = JSON.parse(content);
useEditorStore.getState().loadProject(data);
} catch (err) {
console.error('Failed to load project:', err);
alert(`Failed to load project: ${err}`);
}
};
const handleOpenFile = async () => {
if (IS_ELECTRON) {
const path = await window.electronAPI!.openFile();
if (path) {
loadVideo(path);
await transcribeVideo(path);
}
} else {
// Browser: use the manual path input
const path = manualPath.trim();
if (path) {
loadVideo(path);
await transcribeVideo(path);
}
}
};
const handleManualSubmit = async (e: React.FormEvent) => {
e.preventDefault();
const path = manualPath.trim();
if (!path) return;
loadVideo(path);
await transcribeVideo(path);
};
const transcribeVideo = async (path: string) => {
setTranscribing(true, 0);
try {
const res = await fetch(`${backendUrl}/transcribe`, {
method: 'POST',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({ file_path: path, model: whisperModel }),
});
if (!res.ok) throw new Error(`Transcription failed: ${res.statusText}`);
const data = await res.json();
setTranscription(data);
} catch (err) {
console.error('Transcription error:', err);
alert(`Transcription failed. Check the console for details.\n\n${err}`);
} finally {
setTranscribing(false);
}
};
const togglePanel = (panel: Panel) =>
setActivePanel((prev) => (prev === panel ? null : panel));
if (!videoPath) {
return (
<div className="h-screen flex flex-col items-center justify-center gap-8 bg-editor-bg px-6">
<div className="flex flex-col items-center gap-3">
<Film className="w-14 h-14 text-editor-accent opacity-80" />
<h1 className="text-3xl font-semibold tracking-tight">CutScript</h1>
<p className="text-editor-text-muted text-sm max-w-sm text-center">
Open-source text-based video editing powered by AI.
</p>
</div>
{/* Whisper model selector */}
<div className="flex items-center gap-3">
<label className="text-xs text-editor-text-muted whitespace-nowrap">Whisper model:</label>
<select
value={whisperModel}
onChange={(e) => setWhisperModel(e.target.value)}
className="px-3 py-1.5 bg-editor-surface border border-editor-border rounded-lg text-xs text-editor-text focus:outline-none focus:border-editor-accent"
>
<option value="tiny">tiny (~75 MB, fastest)</option>
<option value="base">base (~140 MB, fast)</option>
<option value="small">small (~460 MB, good)</option>
<option value="medium">medium (~1.5 GB, better)</option>
<option value="large">large (~2.9 GB, best)</option>
</select>
</div>
{IS_ELECTRON ? (
<div className="flex flex-col items-center gap-3">
<button
onClick={handleOpenFile}
className="flex items-center gap-2 px-6 py-3 bg-editor-accent hover:bg-editor-accent-hover rounded-lg text-white font-medium transition-colors"
>
<FolderOpen className="w-5 h-5" />
Open Video File
</button>
<button
onClick={handleLoadProject}
className="flex items-center gap-2 px-4 py-2 text-sm text-editor-text-muted hover:text-editor-text hover:bg-editor-surface rounded-lg transition-colors"
>
<FileInput className="w-4 h-4" />
Load Project (.aive)
</button>
</div>
) : (
/* Browser: manual path input */
<div className="w-full max-w-lg space-y-3">
<div className="flex items-center gap-2 px-3 py-1.5 bg-editor-warning/10 border border-editor-warning/30 rounded-lg">
<span className="text-editor-warning text-xs">
Running in browser paste the full path to your video file below.
</span>
</div>
<form onSubmit={handleManualSubmit} className="flex gap-2">
<div className="flex-1 relative">
<FolderSearch className="absolute left-3 top-1/2 -translate-y-1/2 w-4 h-4 text-editor-text-muted pointer-events-none" />
<input
ref={fileInputRef}
type="text"
value={manualPath}
onChange={(e) => setManualPath(e.target.value)}
placeholder="C:\Videos\my-video.mp4"
className="w-full pl-9 pr-3 py-2.5 bg-editor-surface border border-editor-border rounded-lg text-sm text-editor-text placeholder:text-editor-text-muted/40 focus:outline-none focus:border-editor-accent"
autoFocus
/>
</div>
<button
type="submit"
disabled={!manualPath.trim()}
className="flex items-center gap-2 px-5 py-2.5 bg-editor-accent hover:bg-editor-accent-hover disabled:opacity-40 rounded-lg text-sm text-white font-medium transition-colors whitespace-nowrap"
>
<Film className="w-4 h-4" />
Load &amp; Transcribe
</button>
</form>
<p className="text-[11px] text-editor-text-muted text-center">
Supported: MP4, AVI, MOV, MKV, WebM, M4A
</p>
</div>
)}
</div>
);
}
return (
<div className="h-screen flex flex-col bg-editor-bg overflow-hidden">
{/* Top bar */}
<header className="h-12 flex items-center justify-between px-4 border-b border-editor-border shrink-0">
<div className="flex items-center gap-3">
<Film className="w-5 h-5 text-editor-accent" />
<span className="text-sm font-medium truncate max-w-[300px]">
{videoPath.split(/[\\/]/).pop()}
</span>
</div>
<div className="flex items-center gap-1">
<ToolbarButton
icon={<FolderOpen className="w-4 h-4" />}
label="Open"
onClick={IS_ELECTRON ? handleOpenFile : () => useEditorStore.getState().reset()}
/>
<ToolbarButton
icon={<Sparkles className="w-4 h-4" />}
label="AI"
active={activePanel === 'ai'}
onClick={() => togglePanel('ai')}
disabled={words.length === 0}
/>
<ToolbarButton
icon={<Download className="w-4 h-4" />}
label="Export"
active={activePanel === 'export'}
onClick={() => togglePanel('export')}
disabled={words.length === 0}
/>
<ToolbarButton
icon={<Settings className="w-4 h-4" />}
label="Settings"
active={activePanel === 'settings'}
onClick={() => togglePanel('settings')}
/>
</div>
</header>
{/* Main content */}
<div className="flex-1 flex overflow-hidden">
{/* Left: video + transcript */}
<div className="flex-1 flex flex-col min-w-0">
<div className="flex-1 flex min-h-0">
{/* Video player */}
<div className="w-1/2 p-3 flex items-center justify-center bg-black/20">
<VideoPlayer />
</div>
{/* Transcript */}
<div className="w-1/2 border-l border-editor-border flex flex-col min-h-0">
{isTranscribing ? (
<div className="flex-1 flex flex-col items-center justify-center gap-4">
<Loader2 className="w-8 h-8 text-editor-accent animate-spin" />
<p className="text-sm text-editor-text-muted">
Transcribing... {Math.round(transcriptionProgress)}%
</p>
</div>
) : words.length > 0 ? (
<TranscriptEditor />
) : (
<div className="flex-1 flex items-center justify-center text-editor-text-muted text-sm">
No transcript yet
</div>
)}
</div>
</div>
{/* Waveform timeline */}
<div className="h-32 border-t border-editor-border shrink-0">
<WaveformTimeline />
</div>
</div>
{/* Right panel (AI / Export / Settings) */}
{activePanel && (
<div className="w-80 border-l border-editor-border overflow-y-auto shrink-0">
{activePanel === 'ai' && <AIPanel />}
{activePanel === 'export' && <ExportDialog />}
{activePanel === 'settings' && <SettingsPanel />}
</div>
)}
</div>
</div>
);
}
function ToolbarButton({
icon,
label,
active,
onClick,
disabled,
}: {
icon: React.ReactNode;
label: string;
active?: boolean;
onClick: () => void;
disabled?: boolean;
}) {
return (
<button
onClick={onClick}
disabled={disabled}
title={label}
className={`flex items-center gap-1.5 px-3 py-1.5 rounded-md text-xs font-medium transition-colors ${
active
? 'bg-editor-accent text-white'
: 'text-editor-text-muted hover:text-editor-text hover:bg-editor-surface'
} ${disabled ? 'opacity-40 cursor-not-allowed' : ''}`}
>
{icon}
{label}
</button>
);
}

View File

@ -0,0 +1,332 @@
import { useCallback, useState } from 'react';
import { useEditorStore } from '../store/editorStore';
import { useAIStore } from '../store/aiStore';
import { Sparkles, Scissors, Film, Loader2, Check, X, Play, Download } from 'lucide-react';
import type { ClipSuggestion } from '../types/project';
export default function AIPanel() {
const { words, videoPath, backendUrl, deleteWordRange, setCurrentTime } = useEditorStore();
const {
defaultProvider,
providers,
customFillerWords,
fillerResult,
clipSuggestions,
isProcessing,
processingMessage,
setCustomFillerWords,
setFillerResult,
setClipSuggestions,
setProcessing,
} = useAIStore();
const [activeTab, setActiveTab] = useState<'filler' | 'clips'>('filler');
const detectFillers = useCallback(async () => {
if (words.length === 0) return;
setProcessing(true, 'Detecting filler words...');
try {
const config = providers[defaultProvider];
const transcript = words.map((w) => w.word).join(' ');
const res = await fetch(`${backendUrl}/ai/filler-removal`, {
method: 'POST',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({
transcript,
words: words.map((w, i) => ({ index: i, word: w.word })),
provider: defaultProvider,
model: config.model,
api_key: config.apiKey || undefined,
base_url: config.baseUrl || undefined,
custom_filler_words: customFillerWords || undefined,
}),
});
if (!res.ok) throw new Error('Filler detection failed');
const data = await res.json();
setFillerResult(data);
} catch (err) {
console.error(err);
} finally {
setProcessing(false);
}
}, [words, backendUrl, defaultProvider, providers, customFillerWords, setProcessing, setFillerResult]);
const createClips = useCallback(async () => {
if (words.length === 0) return;
setProcessing(true, 'Finding best clip segments...');
try {
const config = providers[defaultProvider];
const transcript = words.map((w) => w.word).join(' ');
const res = await fetch(`${backendUrl}/ai/create-clip`, {
method: 'POST',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({
transcript,
words: words.map((w, i) => ({
index: i,
word: w.word,
start: w.start,
end: w.end,
})),
provider: defaultProvider,
model: config.model,
api_key: config.apiKey || undefined,
base_url: config.baseUrl || undefined,
target_duration: 60,
}),
});
if (!res.ok) throw new Error('Clip creation failed');
const data = await res.json();
setClipSuggestions(data.clips || []);
} catch (err) {
console.error(err);
} finally {
setProcessing(false);
}
}, [words, backendUrl, defaultProvider, providers, setProcessing, setClipSuggestions]);
const applyFillerDeletions = useCallback(() => {
if (!fillerResult) return;
const sorted = [...fillerResult.fillerWords].sort((a, b) => b.index - a.index);
for (const fw of sorted) {
deleteWordRange(fw.index, fw.index);
}
setFillerResult(null);
}, [fillerResult, deleteWordRange, setFillerResult]);
const handlePreviewClip = useCallback(
(clip: ClipSuggestion) => {
setCurrentTime(clip.startTime);
const video = document.querySelector('video');
if (video) {
video.currentTime = clip.startTime;
video.play();
}
},
[setCurrentTime],
);
const [exportingClipIndex, setExportingClipIndex] = useState<number | null>(null);
const handleExportClip = useCallback(
async (clip: ClipSuggestion, index: number) => {
if (!videoPath) return;
setExportingClipIndex(index);
try {
const safeName = clip.title.replace(/[^a-zA-Z0-9_-]/g, '_').substring(0, 40);
const dirSep = videoPath.lastIndexOf('\\') >= 0 ? '\\' : '/';
const dir = videoPath.substring(0, videoPath.lastIndexOf(dirSep));
const outputPath = `${dir}${dirSep}${safeName}_clip.mp4`;
const res = await fetch(`${backendUrl}/export`, {
method: 'POST',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({
input_path: videoPath,
output_path: outputPath,
keep_segments: [{ start: clip.startTime, end: clip.endTime }],
mode: 'fast',
format: 'mp4',
}),
});
if (!res.ok) throw new Error('Export failed');
const data = await res.json();
alert(`Clip exported to: ${data.output_path}`);
} catch (err) {
console.error(err);
alert('Failed to export clip. Check console for details.');
} finally {
setExportingClipIndex(null);
}
},
[videoPath, backendUrl],
);
return (
<div className="flex flex-col h-full">
<div className="flex border-b border-editor-border shrink-0">
<TabButton
active={activeTab === 'filler'}
onClick={() => setActiveTab('filler')}
icon={<Scissors className="w-3.5 h-3.5" />}
label="Filler Words"
/>
<TabButton
active={activeTab === 'clips'}
onClick={() => setActiveTab('clips')}
icon={<Film className="w-3.5 h-3.5" />}
label="Create Clips"
/>
</div>
<div className="flex-1 overflow-y-auto p-4">
{activeTab === 'filler' && (
<div className="space-y-4">
<p className="text-xs text-editor-text-muted">
Use AI to detect and remove filler words like "um", "uh", "like", "you know" from
your transcript.
</p>
<div className="space-y-1.5">
<label className="text-[11px] text-editor-text-muted font-medium">
Custom filler words (comma-separated)
</label>
<input
type="text"
value={customFillerWords}
onChange={(e) => setCustomFillerWords(e.target.value)}
placeholder="e.g. okay, alright, anyway"
className="w-full px-2.5 py-1.5 text-xs bg-editor-surface border border-editor-border rounded focus:border-editor-accent focus:outline-none"
/>
</div>
<button
onClick={detectFillers}
disabled={isProcessing || words.length === 0}
className="w-full flex items-center justify-center gap-2 px-4 py-2.5 bg-editor-accent hover:bg-editor-accent-hover disabled:opacity-50 rounded-lg text-sm font-medium transition-colors"
>
{isProcessing ? (
<>
<Loader2 className="w-4 h-4 animate-spin" />
{processingMessage}
</>
) : (
<>
<Sparkles className="w-4 h-4" />
Detect Filler Words
</>
)}
</button>
{fillerResult && fillerResult.fillerWords.length > 0 && (
<div className="space-y-3">
<div className="flex items-center justify-between">
<span className="text-xs font-medium">
Found {fillerResult.fillerWords.length} filler words
</span>
<div className="flex gap-1">
<button
onClick={applyFillerDeletions}
className="flex items-center gap-1 px-2 py-1 text-xs bg-editor-success/20 text-editor-success rounded hover:bg-editor-success/30"
>
<Check className="w-3 h-3" /> Apply All
</button>
<button
onClick={() => setFillerResult(null)}
className="flex items-center gap-1 px-2 py-1 text-xs bg-editor-border text-editor-text-muted rounded hover:bg-editor-surface"
>
<X className="w-3 h-3" /> Dismiss
</button>
</div>
</div>
<div className="space-y-1 max-h-64 overflow-y-auto">
{fillerResult.fillerWords.map((fw) => (
<div
key={fw.index}
className="flex items-center justify-between px-2 py-1.5 bg-editor-word-filler rounded text-xs"
>
<span>
<strong>"{fw.word}"</strong>
<span className="text-editor-text-muted ml-1"> {fw.reason}</span>
</span>
</div>
))}
</div>
</div>
)}
{fillerResult && fillerResult.fillerWords.length === 0 && (
<p className="text-xs text-editor-success">No filler words detected.</p>
)}
</div>
)}
{activeTab === 'clips' && (
<div className="space-y-4">
<p className="text-xs text-editor-text-muted">
AI analyzes your transcript and suggests the most engaging segments for a
YouTube Short or social media clip.
</p>
<button
onClick={createClips}
disabled={isProcessing || words.length === 0}
className="w-full flex items-center justify-center gap-2 px-4 py-2.5 bg-editor-accent hover:bg-editor-accent-hover disabled:opacity-50 rounded-lg text-sm font-medium transition-colors"
>
{isProcessing ? (
<>
<Loader2 className="w-4 h-4 animate-spin" />
{processingMessage}
</>
) : (
<>
<Film className="w-4 h-4" />
Find Best Clips
</>
)}
</button>
{clipSuggestions.length > 0 && (
<div className="space-y-3">
{clipSuggestions.map((clip, i) => (
<div key={i} className="p-3 bg-editor-surface rounded-lg space-y-2">
<div className="flex items-center justify-between">
<span className="text-xs font-semibold">{clip.title}</span>
<span className="text-[10px] text-editor-text-muted">
{Math.round(clip.endTime - clip.startTime)}s
</span>
</div>
<p className="text-[11px] text-editor-text-muted">{clip.reason}</p>
<div className="flex gap-2">
<button
onClick={() => handlePreviewClip(clip)}
className="flex-1 flex items-center justify-center gap-1 px-2 py-1.5 text-xs bg-editor-accent/20 text-editor-accent rounded hover:bg-editor-accent/30 transition-colors"
>
<Play className="w-3 h-3" /> Preview
</button>
<button
onClick={() => handleExportClip(clip, i)}
disabled={exportingClipIndex === i}
className="flex-1 flex items-center justify-center gap-1 px-2 py-1.5 text-xs bg-editor-success/20 text-editor-success rounded hover:bg-editor-success/30 disabled:opacity-50 transition-colors"
>
{exportingClipIndex === i ? (
<Loader2 className="w-3 h-3 animate-spin" />
) : (
<Download className="w-3 h-3" />
)}
Export
</button>
</div>
</div>
))}
</div>
)}
</div>
)}
</div>
</div>
);
}
function TabButton({
active,
onClick,
icon,
label,
}: {
active: boolean;
onClick: () => void;
icon: React.ReactNode;
label: string;
}) {
return (
<button
onClick={onClick}
className={`flex-1 flex items-center justify-center gap-1.5 px-3 py-2.5 text-xs font-medium transition-colors border-b-2 ${
active
? 'border-editor-accent text-editor-accent'
: 'border-transparent text-editor-text-muted hover:text-editor-text'
}`}
>
{icon}
{label}
</button>
);
}

View File

@ -0,0 +1,229 @@
import { useState, useCallback, useMemo } from 'react';
import { useEditorStore } from '../store/editorStore';
import { Download, Loader2, Zap, Cog, Info } from 'lucide-react';
import type { ExportOptions } from '../types/project';
export default function ExportDialog() {
const { videoPath, words, deletedRanges, isExporting, exportProgress, backendUrl, setExporting, getKeepSegments } =
useEditorStore();
const hasCuts = deletedRanges.length > 0;
const [options, setOptions] = useState<Omit<ExportOptions, 'outputPath'>>({
mode: 'fast',
resolution: '1080p',
format: 'mp4',
enhanceAudio: false,
captions: 'none',
});
const handleExport = useCallback(async () => {
if (!videoPath) return;
const outputPath = await window.electronAPI?.saveFile({
defaultPath: videoPath.replace(/\.[^.]+$/, '_edited.mp4'),
filters: [
{ name: 'MP4', extensions: ['mp4'] },
{ name: 'MOV', extensions: ['mov'] },
{ name: 'WebM', extensions: ['webm'] },
],
});
if (!outputPath) return;
setExporting(true, 0);
try {
const keepSegments = getKeepSegments();
const deletedSet = new Set<number>();
for (const range of deletedRanges) {
for (const idx of range.wordIndices) deletedSet.add(idx);
}
const res = await fetch(`${backendUrl}/export`, {
method: 'POST',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({
input_path: videoPath,
output_path: outputPath,
keep_segments: keepSegments,
words: options.captions !== 'none' ? words : undefined,
deleted_indices: options.captions !== 'none' ? [...deletedSet] : undefined,
...options,
}),
});
if (!res.ok) throw new Error(`Export failed: ${res.statusText}`);
setExporting(false, 100);
} catch (err) {
console.error('Export error:', err);
setExporting(false);
}
}, [videoPath, options, backendUrl, setExporting, getKeepSegments]);
return (
<div className="p-4 space-y-5">
<h3 className="text-sm font-semibold">Export Video</h3>
{/* Mode */}
<fieldset className="space-y-2">
<legend className="text-xs text-editor-text-muted font-medium">Export Mode</legend>
<div className="grid grid-cols-2 gap-2">
<ModeCard
active={options.mode === 'fast'}
onClick={() => setOptions((o) => ({ ...o, mode: 'fast' }))}
icon={<Zap className="w-4 h-4" />}
title="Fast"
desc="Stream copy, seconds"
/>
<ModeCard
active={options.mode === 'reencode'}
onClick={() => setOptions((o) => ({ ...o, mode: 'reencode' }))}
icon={<Cog className="w-4 h-4" />}
title="Re-encode"
desc="Custom quality, slower"
/>
</div>
</fieldset>
{/* Resolution (only for re-encode) */}
{options.mode === 'reencode' && (
<SelectField
label="Resolution"
value={options.resolution}
onChange={(v) => setOptions((o) => ({ ...o, resolution: v as ExportOptions['resolution'] }))}
options={[
{ value: '720p', label: '720p (HD)' },
{ value: '1080p', label: '1080p (Full HD)' },
{ value: '4k', label: '4K (Ultra HD)' },
]}
/>
)}
{/* Format */}
<SelectField
label="Format"
value={options.format}
onChange={(v) => setOptions((o) => ({ ...o, format: v as ExportOptions['format'] }))}
options={[
{ value: 'mp4', label: 'MP4 (H.264)' },
{ value: 'mov', label: 'MOV (QuickTime)' },
{ value: 'webm', label: 'WebM (VP9)' },
]}
/>
{/* Audio enhancement */}
<label className="flex items-center gap-2 cursor-pointer">
<input
type="checkbox"
checked={options.enhanceAudio}
onChange={(e) => setOptions((o) => ({ ...o, enhanceAudio: e.target.checked }))}
className="w-4 h-4 rounded bg-editor-surface border-editor-border accent-editor-accent"
/>
<span className="text-xs">Enhance audio (Studio Sound)</span>
</label>
{/* Captions */}
<SelectField
label="Captions"
value={options.captions}
onChange={(v) => setOptions((o) => ({ ...o, captions: v as ExportOptions['captions'] }))}
options={[
{ value: 'none', label: 'No captions' },
{ value: 'burn-in', label: 'Burn-in (permanent)' },
{ value: 'sidecar', label: 'Sidecar SRT file' },
]}
/>
{/* Export button */}
<button
onClick={handleExport}
disabled={isExporting || !videoPath}
className="w-full flex items-center justify-center gap-2 px-4 py-3 bg-editor-accent hover:bg-editor-accent-hover disabled:opacity-50 rounded-lg text-sm font-semibold transition-colors"
>
{isExporting ? (
<>
<Loader2 className="w-4 h-4 animate-spin" />
Exporting... {Math.round(exportProgress)}%
</>
) : (
<>
<Download className="w-4 h-4" />
Export
</>
)}
</button>
{options.mode === 'fast' && !hasCuts && (
<p className="text-[10px] text-editor-text-muted text-center">
Fast mode uses stream copy &mdash; no quality loss, exports in seconds.
</p>
)}
{options.mode === 'fast' && hasCuts && (
<div className="flex items-start gap-1.5 p-2 bg-editor-accent/10 rounded text-[10px] text-editor-accent">
<Info className="w-3.5 h-3.5 shrink-0 mt-0.5" />
<span>
Word-level cuts require re-encoding for frame-accurate output. Export will
automatically use re-encode mode. This takes longer but ensures your cuts are precise.
</span>
</div>
)}
</div>
);
}
function ModeCard({
active,
onClick,
icon,
title,
desc,
}: {
active: boolean;
onClick: () => void;
icon: React.ReactNode;
title: string;
desc: string;
}) {
return (
<button
onClick={onClick}
className={`flex flex-col items-center gap-1 p-3 rounded-lg border-2 transition-colors ${
active
? 'border-editor-accent bg-editor-accent/10'
: 'border-editor-border hover:border-editor-text-muted'
}`}
>
{icon}
<span className="text-xs font-medium">{title}</span>
<span className="text-[10px] text-editor-text-muted">{desc}</span>
</button>
);
}
function SelectField({
label,
value,
onChange,
options,
}: {
label: string;
value: string;
onChange: (value: string) => void;
options: Array<{ value: string; label: string }>;
}) {
return (
<div className="space-y-1">
<label className="text-xs text-editor-text-muted font-medium">{label}</label>
<select
value={value}
onChange={(e) => onChange(e.target.value)}
className="w-full px-3 py-2 bg-editor-surface border border-editor-border rounded-lg text-xs text-editor-text focus:outline-none focus:border-editor-accent"
>
{options.map((opt) => (
<option key={opt.value} value={opt.value}>
{opt.label}
</option>
))}
</select>
</div>
);
}

View File

@ -0,0 +1,192 @@
import { useAIStore } from '../store/aiStore';
import { useState, useEffect } from 'react';
import type { AIProvider } from '../types/project';
import { useEditorStore } from '../store/editorStore';
import { Bot, Cloud, Brain, RefreshCw } from 'lucide-react';
export default function SettingsPanel() {
const { providers, defaultProvider, setProviderConfig, setDefaultProvider } = useAIStore();
const { backendUrl } = useEditorStore();
const [ollamaModels, setOllamaModels] = useState<string[]>([]);
const [loadingModels, setLoadingModels] = useState(false);
const fetchOllamaModels = async () => {
setLoadingModels(true);
try {
const res = await fetch(`${backendUrl}/ai/ollama-models`);
if (res.ok) {
const data = await res.json();
setOllamaModels(data.models || []);
}
} catch {
setOllamaModels([]);
} finally {
setLoadingModels(false);
}
};
useEffect(() => {
fetchOllamaModels();
}, [backendUrl]);
const providerIcons: Record<AIProvider, React.ReactNode> = {
ollama: <Bot className="w-4 h-4" />,
openai: <Cloud className="w-4 h-4" />,
claude: <Brain className="w-4 h-4" />,
};
const providerLabels: Record<AIProvider, string> = {
ollama: 'Ollama (Local)',
openai: 'OpenAI',
claude: 'Claude (Anthropic)',
};
return (
<div className="p-4 space-y-6">
<h3 className="text-sm font-semibold">AI Settings</h3>
{/* Default provider selector */}
<div className="space-y-2">
<label className="text-xs text-editor-text-muted font-medium">Default AI Provider</label>
<div className="grid grid-cols-3 gap-1.5">
{(['ollama', 'openai', 'claude'] as AIProvider[]).map((p) => (
<button
key={p}
onClick={() => setDefaultProvider(p)}
className={`flex flex-col items-center gap-1 p-2 rounded-lg border transition-colors text-[10px] ${
defaultProvider === p
? 'border-editor-accent bg-editor-accent/10 text-editor-accent'
: 'border-editor-border text-editor-text-muted hover:text-editor-text'
}`}
>
{providerIcons[p]}
{p.charAt(0).toUpperCase() + p.slice(1)}
</button>
))}
</div>
</div>
{/* Ollama settings */}
<ProviderSection title="Ollama (Local)" icon={providerIcons.ollama}>
<InputField
label="Base URL"
value={providers.ollama.baseUrl || ''}
onChange={(v) => setProviderConfig('ollama', { baseUrl: v })}
placeholder="http://localhost:11434"
/>
<div className="space-y-1">
<div className="flex items-center justify-between">
<label className="text-xs text-editor-text-muted">Model</label>
<button
onClick={fetchOllamaModels}
disabled={loadingModels}
className="text-[10px] text-editor-accent hover:underline flex items-center gap-0.5"
>
<RefreshCw className={`w-2.5 h-2.5 ${loadingModels ? 'animate-spin' : ''}`} />
Refresh
</button>
</div>
{ollamaModels.length > 0 ? (
<select
value={providers.ollama.model}
onChange={(e) => setProviderConfig('ollama', { model: e.target.value })}
className="w-full px-3 py-2 bg-editor-surface border border-editor-border rounded-lg text-xs text-editor-text focus:outline-none focus:border-editor-accent"
>
{ollamaModels.map((m) => (
<option key={m} value={m}>{m}</option>
))}
</select>
) : (
<InputField
label=""
value={providers.ollama.model}
onChange={(v) => setProviderConfig('ollama', { model: v })}
placeholder="llama3"
/>
)}
</div>
</ProviderSection>
{/* OpenAI settings */}
<ProviderSection title="OpenAI" icon={providerIcons.openai}>
<InputField
label="API Key"
value={providers.openai.apiKey || ''}
onChange={(v) => setProviderConfig('openai', { apiKey: v })}
placeholder="sk-..."
type="password"
/>
<InputField
label="Model"
value={providers.openai.model}
onChange={(v) => setProviderConfig('openai', { model: v })}
placeholder="gpt-4o"
/>
</ProviderSection>
{/* Claude settings */}
<ProviderSection title="Claude (Anthropic)" icon={providerIcons.claude}>
<InputField
label="API Key"
value={providers.claude.apiKey || ''}
onChange={(v) => setProviderConfig('claude', { apiKey: v })}
placeholder="sk-ant-..."
type="password"
/>
<InputField
label="Model"
value={providers.claude.model}
onChange={(v) => setProviderConfig('claude', { model: v })}
placeholder="claude-sonnet-4-20250514"
/>
</ProviderSection>
</div>
);
}
function ProviderSection({
title,
icon,
children,
}: {
title: string;
icon: React.ReactNode;
children: React.ReactNode;
}) {
return (
<div className="space-y-3 p-3 bg-editor-surface rounded-lg">
<div className="flex items-center gap-2 text-xs font-medium">
{icon}
{title}
</div>
<div className="space-y-2">{children}</div>
</div>
);
}
function InputField({
label,
value,
onChange,
placeholder,
type = 'text',
}: {
label: string;
value: string;
onChange: (value: string) => void;
placeholder: string;
type?: string;
}) {
return (
<div className="space-y-1">
{label && <label className="text-xs text-editor-text-muted">{label}</label>}
<input
type={type}
value={value}
onChange={(e) => onChange(e.target.value)}
placeholder={placeholder}
className="w-full px-3 py-2 bg-editor-bg border border-editor-border rounded-lg text-xs text-editor-text placeholder:text-editor-text-muted/50 focus:outline-none focus:border-editor-accent"
/>
</div>
);
}

View File

@ -0,0 +1,204 @@
import { useCallback, useRef, useEffect, useMemo, useState } from 'react';
import { useEditorStore } from '../store/editorStore';
import { Virtuoso } from 'react-virtuoso';
import { Trash2, RotateCcw } from 'lucide-react';
export default function TranscriptEditor() {
const words = useEditorStore((s) => s.words);
const segments = useEditorStore((s) => s.segments);
const deletedRanges = useEditorStore((s) => s.deletedRanges);
const selectedWordIndices = useEditorStore((s) => s.selectedWordIndices);
const hoveredWordIndex = useEditorStore((s) => s.hoveredWordIndex);
const setSelectedWordIndices = useEditorStore((s) => s.setSelectedWordIndices);
const setHoveredWordIndex = useEditorStore((s) => s.setHoveredWordIndex);
const deleteSelectedWords = useEditorStore((s) => s.deleteSelectedWords);
const restoreRange = useEditorStore((s) => s.restoreRange);
const getWordAtTime = useEditorStore((s) => s.getWordAtTime);
const selectionStart = useRef<number | null>(null);
const wasDragging = useRef(false);
const virtuosoRef = useRef<any>(null);
const deletedSet = useMemo(() => {
const s = new Set<number>();
for (const range of deletedRanges) {
for (const idx of range.wordIndices) s.add(idx);
}
return s;
}, [deletedRanges]);
const selectedSet = useMemo(() => new Set(selectedWordIndices), [selectedWordIndices]);
const [activeWordIndex, setActiveWordIndex] = useState(-1);
useEffect(() => {
if (words.length === 0) return;
const interval = setInterval(() => {
const video = document.querySelector('video') as HTMLVideoElement | null;
if (!video) return;
const idx = getWordAtTime(video.currentTime);
setActiveWordIndex((prev) => (prev === idx ? prev : idx));
}, 250);
return () => clearInterval(interval);
}, [words, getWordAtTime]);
// Auto-scroll to active segment via Virtuoso
useEffect(() => {
if (activeWordIndex < 0 || segments.length === 0) return;
const segIdx = segments.findIndex((seg) => {
const start = seg.globalStartIndex ?? 0;
return activeWordIndex >= start && activeWordIndex < start + seg.words.length;
});
if (segIdx >= 0 && virtuosoRef.current) {
virtuosoRef.current.scrollIntoView({ index: segIdx, behavior: 'smooth', align: 'center' });
}
}, [activeWordIndex, segments]);
const handleWordMouseDown = useCallback(
(index: number, e: React.MouseEvent) => {
e.preventDefault();
wasDragging.current = false;
if (e.shiftKey && selectedWordIndices.length > 0) {
const first = selectedWordIndices[0];
const start = Math.min(first, index);
const end = Math.max(first, index);
const indices = [];
for (let i = start; i <= end; i++) indices.push(i);
setSelectedWordIndices(indices);
} else {
selectionStart.current = index;
setSelectedWordIndices([index]);
}
},
[selectedWordIndices, setSelectedWordIndices],
);
const handleWordMouseEnter = useCallback(
(index: number) => {
setHoveredWordIndex(index);
if (selectionStart.current !== null) {
wasDragging.current = true;
const start = Math.min(selectionStart.current, index);
const end = Math.max(selectionStart.current, index);
const indices = [];
for (let i = start; i <= end; i++) indices.push(i);
setSelectedWordIndices(indices);
}
},
[setHoveredWordIndex, setSelectedWordIndices],
);
const handleMouseUp = useCallback(() => {
selectionStart.current = null;
}, []);
const handleClickOutside = useCallback(
(e: React.MouseEvent) => {
if (wasDragging.current) {
wasDragging.current = false;
return;
}
if ((e.target as HTMLElement).dataset.wordIndex === undefined) {
setSelectedWordIndices([]);
}
},
[setSelectedWordIndices],
);
const getRangeForWord = useCallback(
(wordIndex: number) => deletedRanges.find((r) => r.wordIndices.includes(wordIndex)),
[deletedRanges],
);
const renderSegment = useCallback(
(index: number) => {
const segment = segments[index];
if (!segment) return null;
return (
<div className="mb-3 px-4">
{segment.speaker && (
<div className="text-xs text-editor-accent font-medium mb-1">
{segment.speaker}
</div>
)}
<p className="text-sm leading-relaxed flex flex-wrap">
{segment.words.map((word, localIndex) => {
const globalIndex = (segment.globalStartIndex ?? 0) + localIndex;
const isDeleted = deletedSet.has(globalIndex);
const isSelected = selectedSet.has(globalIndex);
const isActive = globalIndex === activeWordIndex;
const isHovered = globalIndex === hoveredWordIndex;
const deletedRange = isDeleted ? getRangeForWord(globalIndex) : null;
return (
<span
key={globalIndex}
id={`word-${globalIndex}`}
data-word-index={globalIndex}
onMouseDown={(e) => handleWordMouseDown(globalIndex, e)}
onMouseEnter={() => handleWordMouseEnter(globalIndex)}
onMouseLeave={() => setHoveredWordIndex(null)}
className={`
relative px-[2px] py-[1px] rounded cursor-pointer transition-colors
${isDeleted ? 'line-through text-editor-text-muted/40 bg-editor-word-deleted' : ''}
${isSelected && !isDeleted ? 'bg-editor-word-selected text-white' : ''}
${isActive && !isDeleted && !isSelected ? 'bg-editor-accent/20 text-editor-accent' : ''}
${isHovered && !isDeleted && !isSelected && !isActive ? 'bg-editor-word-hover' : ''}
`}
>
{word.word}{' '}
{isDeleted && isHovered && deletedRange && (
<button
onClick={(e) => {
e.stopPropagation();
restoreRange(deletedRange.id);
}}
className="absolute -top-5 left-1/2 -translate-x-1/2 flex items-center gap-0.5 px-1.5 py-0.5 bg-editor-surface border border-editor-border rounded text-[10px] text-editor-success whitespace-nowrap z-10"
>
<RotateCcw className="w-2.5 h-2.5" /> Restore
</button>
)}
</span>
);
})}
</p>
</div>
);
},
[segments, deletedSet, selectedSet, activeWordIndex, hoveredWordIndex, handleWordMouseDown, handleWordMouseEnter, setHoveredWordIndex, getRangeForWord, restoreRange],
);
return (
<div className="flex-1 flex flex-col min-h-0">
<div className="flex items-center gap-2 px-4 py-2 border-b border-editor-border shrink-0">
<span className="text-xs text-editor-text-muted flex-1">
{words.length} words &middot; {deletedRanges.length} cuts
</span>
{selectedWordIndices.length > 0 && (
<button
onClick={deleteSelectedWords}
className="flex items-center gap-1 px-2 py-1 text-xs bg-editor-danger/20 text-editor-danger rounded hover:bg-editor-danger/30 transition-colors"
>
<Trash2 className="w-3 h-3" />
Delete {selectedWordIndices.length} words
</button>
)}
</div>
<div
className="flex-1 min-h-0 select-none"
onMouseUp={handleMouseUp}
onClick={handleClickOutside}
>
<Virtuoso
ref={virtuosoRef}
totalCount={segments.length}
itemContent={renderSegment}
overscan={200}
className="h-full"
style={{ height: '100%' }}
/>
</div>
</div>
);
}

View File

@ -0,0 +1,133 @@
import { useRef, useCallback, useState, useEffect } from 'react';
import { useEditorStore } from '../store/editorStore';
import { useVideoSync } from '../hooks/useVideoSync';
import { Play, Pause, SkipBack, SkipForward, Volume2 } from 'lucide-react';
export default function VideoPlayer() {
const videoRef = useRef<HTMLVideoElement>(null);
const videoUrl = useEditorStore((s) => s.videoUrl);
const isPlaying = useEditorStore((s) => s.isPlaying);
const duration = useEditorStore((s) => s.duration);
const { seekTo, togglePlay } = useVideoSync(videoRef);
const [displayTime, setDisplayTime] = useState(0);
useEffect(() => {
const video = videoRef.current;
if (!video) return;
let raf = 0;
const tick = () => {
setDisplayTime(video.currentTime);
raf = requestAnimationFrame(tick);
};
raf = requestAnimationFrame(tick);
return () => cancelAnimationFrame(raf);
}, [videoUrl]);
const formatTime = (seconds: number) => {
const m = Math.floor(seconds / 60);
const s = Math.floor(seconds % 60);
return `${m}:${s.toString().padStart(2, '0')}`;
};
const handleProgressClick = useCallback(
(e: React.MouseEvent<HTMLDivElement>) => {
const rect = e.currentTarget.getBoundingClientRect();
const ratio = (e.clientX - rect.left) / rect.width;
seekTo(ratio * duration);
},
[seekTo, duration],
);
const skip = useCallback(
(delta: number) => {
const video = videoRef.current;
if (!video) return;
seekTo(Math.max(0, Math.min(duration, video.currentTime + delta)));
},
[seekTo, duration],
);
if (!videoUrl) {
return (
<div className="w-full h-full flex items-center justify-center text-editor-text-muted text-sm">
No video loaded
</div>
);
}
return (
<div className="w-full h-full flex flex-col">
<div className="flex-1 flex items-center justify-center bg-black rounded-lg overflow-hidden min-h-0">
<video
ref={videoRef}
src={videoUrl}
className="max-w-full max-h-full object-contain"
playsInline
onClick={togglePlay}
/>
</div>
<div className="pt-2 space-y-1.5 shrink-0">
<div
className="h-1.5 bg-editor-border rounded-full cursor-pointer group"
onClick={handleProgressClick}
>
<div
className="h-full bg-editor-accent rounded-full relative transition-all group-hover:h-2"
style={{ width: duration > 0 ? `${(displayTime / duration) * 100}%` : '0%' }}
>
<div className="absolute right-0 top-1/2 -translate-y-1/2 w-3 h-3 bg-white rounded-full opacity-0 group-hover:opacity-100 transition-opacity" />
</div>
</div>
<div className="flex items-center justify-between">
<div className="flex items-center gap-1">
<ControlButton onClick={() => skip(-5)} title="Back 5s">
<SkipBack className="w-4 h-4" />
</ControlButton>
<ControlButton onClick={togglePlay} title={isPlaying ? 'Pause' : 'Play'} primary>
{isPlaying ? <Pause className="w-5 h-5" /> : <Play className="w-5 h-5 ml-0.5" />}
</ControlButton>
<ControlButton onClick={() => skip(5)} title="Forward 5s">
<SkipForward className="w-4 h-4" />
</ControlButton>
</div>
<div className="flex items-center gap-3 text-xs text-editor-text-muted">
<Volume2 className="w-3.5 h-3.5" />
<span className="font-mono">
{formatTime(displayTime)} / {formatTime(duration)}
</span>
</div>
</div>
</div>
</div>
);
}
function ControlButton({
children,
onClick,
title,
primary,
}: {
children: React.ReactNode;
onClick: () => void;
title: string;
primary?: boolean;
}) {
return (
<button
onClick={onClick}
title={title}
className={`p-1.5 rounded-md transition-colors ${
primary
? 'bg-editor-accent/20 text-editor-accent hover:bg-editor-accent/30'
: 'text-editor-text-muted hover:text-editor-text hover:bg-editor-surface'
}`}
>
{children}
</button>
);
}

View File

@ -0,0 +1,220 @@
import { useRef, useEffect, useCallback, useState } from 'react';
import { useEditorStore } from '../store/editorStore';
import { ZoomIn, ZoomOut, AlertTriangle } from 'lucide-react';
export default function WaveformTimeline() {
const waveCanvasRef = useRef<HTMLCanvasElement>(null);
const headCanvasRef = useRef<HTMLCanvasElement>(null);
const containerRef = useRef<HTMLDivElement>(null);
const [audioError, setAudioError] = useState<string | null>(null);
const videoUrl = useEditorStore((s) => s.videoUrl);
const videoPath = useEditorStore((s) => s.videoPath);
const duration = useEditorStore((s) => s.duration);
const deletedRanges = useEditorStore((s) => s.deletedRanges);
const setCurrentTime = useEditorStore((s) => s.setCurrentTime);
const audioContextRef = useRef<AudioContext | null>(null);
const audioBufferRef = useRef<AudioBuffer | null>(null);
const zoomRef = useRef(1);
const rafRef = useRef(0);
useEffect(() => {
if (!videoUrl || !videoPath) return;
setAudioError(null);
const loadAudio = async () => {
try {
const ctx = new AudioContext();
audioContextRef.current = ctx;
const response = await fetch(videoUrl);
if (!response.ok) throw new Error(`HTTP ${response.status}`);
const arrayBuffer = await response.arrayBuffer();
const audioBuffer = await ctx.decodeAudioData(arrayBuffer);
audioBufferRef.current = audioBuffer;
drawStaticWaveform();
} catch (err) {
console.warn('Could not decode audio for waveform:', err);
setAudioError('Waveform unavailable — audio could not be decoded');
}
};
loadAudio();
return () => {
audioContextRef.current?.close();
};
}, [videoUrl, videoPath]);
const drawStaticWaveform = useCallback(() => {
const canvas = waveCanvasRef.current;
const buffer = audioBufferRef.current;
if (!canvas || !buffer) return;
const ctx = canvas.getContext('2d');
if (!ctx) return;
const dpr = window.devicePixelRatio || 1;
const rect = canvas.getBoundingClientRect();
canvas.width = rect.width * dpr;
canvas.height = rect.height * dpr;
ctx.scale(dpr, dpr);
const width = rect.width;
const height = rect.height;
const channelData = buffer.getChannelData(0);
const samplesPerPixel = Math.floor(channelData.length / width);
ctx.clearRect(0, 0, width, height);
for (const range of deletedRanges) {
const x1 = (range.start / buffer.duration) * width;
const x2 = (range.end / buffer.duration) * width;
ctx.fillStyle = 'rgba(239, 68, 68, 0.15)';
ctx.fillRect(x1, 0, x2 - x1, height);
}
const mid = height / 2;
ctx.beginPath();
ctx.strokeStyle = '#4a4d5e';
ctx.lineWidth = 1;
for (let x = 0; x < width; x++) {
const start = x * samplesPerPixel;
const end = Math.min(start + samplesPerPixel, channelData.length);
let min = 0;
let max = 0;
for (let i = start; i < end; i++) {
if (channelData[i] < min) min = channelData[i];
if (channelData[i] > max) max = channelData[i];
}
const yMin = mid + min * mid * 0.9;
const yMax = mid + max * mid * 0.9;
ctx.moveTo(x, yMin);
ctx.lineTo(x, yMax);
}
ctx.stroke();
}, [deletedRanges]);
// Redraw static layer when deletedRanges change
useEffect(() => {
drawStaticWaveform();
}, [drawStaticWaveform]);
// Lightweight RAF loop for playhead only -- reads video.currentTime directly,
// never triggers React re-renders
useEffect(() => {
const headCanvas = headCanvasRef.current;
const waveCanvas = waveCanvasRef.current;
if (!headCanvas || !waveCanvas) return;
const tick = () => {
const ctx = headCanvas.getContext('2d');
if (!ctx) { rafRef.current = requestAnimationFrame(tick); return; }
const buffer = audioBufferRef.current;
const video = document.querySelector('video') as HTMLVideoElement | null;
const dur = buffer?.duration ?? 0;
const dpr = window.devicePixelRatio || 1;
const rect = headCanvas.getBoundingClientRect();
if (headCanvas.width !== waveCanvas.width || headCanvas.height !== waveCanvas.height) {
headCanvas.width = rect.width * dpr;
headCanvas.height = rect.height * dpr;
}
ctx.setTransform(dpr, 0, 0, dpr, 0, 0);
const width = rect.width;
const height = rect.height;
ctx.clearRect(0, 0, width, height);
if (dur > 0 && video) {
const px = (video.currentTime / dur) * width;
ctx.beginPath();
ctx.strokeStyle = '#6366f1';
ctx.lineWidth = 2;
ctx.moveTo(px, 0);
ctx.lineTo(px, height);
ctx.stroke();
}
rafRef.current = requestAnimationFrame(tick);
};
rafRef.current = requestAnimationFrame(tick);
return () => cancelAnimationFrame(rafRef.current);
}, [videoUrl]);
useEffect(() => {
const observer = new ResizeObserver(() => {
drawStaticWaveform();
});
if (containerRef.current) observer.observe(containerRef.current);
return () => observer.disconnect();
}, [drawStaticWaveform]);
const handleClick = useCallback(
(e: React.MouseEvent<HTMLCanvasElement>) => {
if (!headCanvasRef.current || duration === 0) return;
const rect = headCanvasRef.current.getBoundingClientRect();
const ratio = (e.clientX - rect.left) / rect.width;
const newTime = ratio * duration;
setCurrentTime(newTime);
const video = document.querySelector('video');
if (video) video.currentTime = newTime;
},
[duration, setCurrentTime],
);
if (!videoUrl) {
return (
<div className="w-full h-full flex items-center justify-center text-editor-text-muted text-xs">
Load a video to see the waveform
</div>
);
}
return (
<div ref={containerRef} className="w-full h-full flex flex-col">
<div className="flex items-center justify-between px-3 py-1 shrink-0">
<span className="text-[10px] text-editor-text-muted font-medium uppercase tracking-wider">
Timeline
</span>
<div className="flex items-center gap-1">
<button
onClick={() => { zoomRef.current = Math.max(0.5, zoomRef.current - 0.5); drawStaticWaveform(); }}
className="p-0.5 text-editor-text-muted hover:text-editor-text"
title="Zoom out"
>
<ZoomOut className="w-3.5 h-3.5" />
</button>
<button
onClick={() => { zoomRef.current = Math.min(10, zoomRef.current + 0.5); drawStaticWaveform(); }}
className="p-0.5 text-editor-text-muted hover:text-editor-text"
title="Zoom in"
>
<ZoomIn className="w-3.5 h-3.5" />
</button>
</div>
</div>
{audioError ? (
<div className="flex-1 flex items-center justify-center gap-2 text-editor-text-muted text-xs">
<AlertTriangle className="w-4 h-4 text-yellow-500" />
<span>{audioError}</span>
</div>
) : (
<div className="flex-1 relative">
<canvas ref={waveCanvasRef} className="absolute inset-0 w-full h-full" />
<canvas
ref={headCanvasRef}
className="absolute inset-0 w-full h-full cursor-crosshair"
onClick={handleClick}
/>
</div>
)}
</div>
);
}

View File

@ -0,0 +1,236 @@
import { useEffect, useRef } from 'react';
import { useEditorStore } from '../store/editorStore';
export function useKeyboardShortcuts() {
const deleteSelectedWords = useEditorStore((s) => s.deleteSelectedWords);
const selectedWordIndices = useEditorStore((s) => s.selectedWordIndices);
const playbackRateRef = useRef(1);
useEffect(() => {
const getVideo = (): HTMLVideoElement | null => document.querySelector('video');
const handler = (e: KeyboardEvent) => {
const target = e.target as HTMLElement;
if (target.tagName === 'INPUT' || target.tagName === 'TEXTAREA' || target.tagName === 'SELECT') return;
const video = getVideo();
switch (true) {
// --- Undo / Redo ---
case e.key === 'z' && (e.ctrlKey || e.metaKey) && e.shiftKey: {
e.preventDefault();
useEditorStore.temporal.getState().redo();
return;
}
case e.key === 'z' && (e.ctrlKey || e.metaKey): {
e.preventDefault();
useEditorStore.temporal.getState().undo();
return;
}
// --- Delete / Backspace: delete selected words ---
case e.key === 'Delete' || e.key === 'Backspace': {
if (selectedWordIndices.length > 0) {
e.preventDefault();
deleteSelectedWords();
}
return;
}
// --- Space: play / pause ---
case e.key === ' ' && !e.ctrlKey: {
e.preventDefault();
if (video) {
if (video.paused) video.play();
else video.pause();
}
return;
}
// --- J: reverse / slow down ---
case e.key === 'j' || e.key === 'J': {
e.preventDefault();
if (video) {
playbackRateRef.current = Math.max(-2, playbackRateRef.current - 0.5);
if (playbackRateRef.current < 0) {
// HTML5 video doesn't support negative rates natively; step back
video.currentTime = Math.max(0, video.currentTime - 2);
} else {
video.playbackRate = playbackRateRef.current;
if (video.paused) video.play();
}
}
return;
}
// --- K: pause ---
case e.key === 'k' || e.key === 'K': {
e.preventDefault();
if (video) {
video.pause();
playbackRateRef.current = 1;
}
return;
}
// --- L: forward / speed up ---
case e.key === 'l' || e.key === 'L': {
e.preventDefault();
if (video) {
playbackRateRef.current = Math.min(4, playbackRateRef.current + 0.5);
video.playbackRate = Math.max(0.25, playbackRateRef.current);
if (video.paused) video.play();
}
return;
}
// --- Arrow Left: seek back 5s ---
case e.key === 'ArrowLeft' && !e.ctrlKey: {
e.preventDefault();
if (video) video.currentTime = Math.max(0, video.currentTime - 5);
return;
}
// --- Arrow Right: seek forward 5s ---
case e.key === 'ArrowRight' && !e.ctrlKey: {
e.preventDefault();
if (video) video.currentTime = Math.min(video.duration, video.currentTime + 5);
return;
}
// --- [ mark in-point (home) ---
case e.key === '[': {
e.preventDefault();
if (video) video.currentTime = 0;
return;
}
// --- ] mark out-point (end) ---
case e.key === ']': {
e.preventDefault();
if (video) video.currentTime = video.duration;
return;
}
// --- Ctrl+S: save project ---
case e.key === 's' && (e.ctrlKey || e.metaKey): {
e.preventDefault();
saveProject();
return;
}
// --- Ctrl+E: export ---
case e.key === 'e' && (e.ctrlKey || e.metaKey): {
e.preventDefault();
// Trigger export panel via DOM click
const exportBtn = document.querySelector('[title="Export"]') as HTMLButtonElement;
if (exportBtn) exportBtn.click();
return;
}
// --- ?: show shortcut cheatsheet ---
case e.key === '?' || (e.key === '/' && e.shiftKey): {
e.preventDefault();
toggleCheatsheet();
return;
}
default:
break;
}
};
window.addEventListener('keydown', handler);
return () => window.removeEventListener('keydown', handler);
}, [deleteSelectedWords, selectedWordIndices]);
}
async function saveProject() {
const state = useEditorStore.getState();
if (!state.videoPath || state.words.length === 0) return;
try {
const projectData = {
version: 1,
videoPath: state.videoPath,
words: state.words,
segments: state.segments,
deletedRanges: state.deletedRanges,
language: state.language,
createdAt: new Date().toISOString(),
modifiedAt: new Date().toISOString(),
};
const outputPath = await window.electronAPI?.saveFile({
defaultPath: state.videoPath.replace(/\.[^.]+$/, '.aive'),
filters: [{ name: 'CutScript Project', extensions: ['aive'] }],
});
if (outputPath) {
if (window.electronAPI?.writeFile) {
await window.electronAPI.writeFile(outputPath, JSON.stringify(projectData, null, 2));
} else {
const blob = new Blob([JSON.stringify(projectData, null, 2)], { type: 'application/json' });
const url = URL.createObjectURL(blob);
const a = document.createElement('a');
a.href = url;
a.download = outputPath.split(/[\\/]/).pop() || 'project.aive';
a.click();
URL.revokeObjectURL(url);
}
}
} catch (err) {
console.error('Failed to save project:', err);
}
}
let cheatsheetVisible = false;
function toggleCheatsheet() {
const existing = document.getElementById('keyboard-cheatsheet');
if (existing) {
existing.remove();
cheatsheetVisible = false;
return;
}
cheatsheetVisible = true;
const overlay = document.createElement('div');
overlay.id = 'keyboard-cheatsheet';
overlay.style.cssText =
'position:fixed;inset:0;z-index:9999;display:flex;align-items:center;justify-content:center;background:rgba(0,0,0,0.7);';
overlay.onclick = () => {
overlay.remove();
cheatsheetVisible = false;
};
const shortcuts = [
['Space', 'Play / Pause'],
['J', 'Reverse / Slow down'],
['K', 'Pause'],
['L', 'Forward / Speed up'],
['\u2190 / \u2192', 'Seek \u00b15 seconds'],
['Delete', 'Delete selected words'],
['Ctrl+Z', 'Undo'],
['Ctrl+Shift+Z', 'Redo'],
['Ctrl+S', 'Save project'],
['Ctrl+E', 'Export'],
['?', 'This cheatsheet'],
];
const rows = shortcuts
.map(
([key, desc]) =>
`<tr><td style="padding:6px 16px 6px 0;font-family:monospace;color:#818cf8;font-weight:600">${key}</td><td style="padding:6px 0;color:#e2e8f0">${desc}</td></tr>`,
)
.join('');
overlay.innerHTML = `<div style="background:#1a1d27;border:1px solid #2a2d3a;border-radius:12px;padding:24px 32px;max-width:400px;" onclick="event.stopPropagation()">
<h3 style="margin:0 0 16px;font-size:14px;font-weight:600;color:#e2e8f0">Keyboard Shortcuts</h3>
<table style="font-size:13px">${rows}</table>
<p style="margin:16px 0 0;font-size:11px;color:#94a3b8;text-align:center">Press ? or click outside to close</p>
</div>`;
document.body.appendChild(overlay);
}

View File

@ -0,0 +1,69 @@
import { useCallback, useRef, useEffect } from 'react';
import { useEditorStore } from '../store/editorStore';
export function useVideoSync(videoRef: React.RefObject<HTMLVideoElement | null>) {
const rafRef = useRef<number>(0);
const {
setCurrentTime,
setDuration,
setIsPlaying,
deletedRanges,
} = useEditorStore();
const seekTo = useCallback(
(time: number) => {
if (videoRef.current) {
videoRef.current.currentTime = time;
setCurrentTime(time);
}
},
[videoRef, setCurrentTime],
);
const togglePlay = useCallback(() => {
if (!videoRef.current) return;
if (videoRef.current.paused) {
videoRef.current.play();
} else {
videoRef.current.pause();
}
}, [videoRef]);
useEffect(() => {
const video = videoRef.current;
if (!video) return;
const onTimeUpdate = () => {
cancelAnimationFrame(rafRef.current);
rafRef.current = requestAnimationFrame(() => {
const t = video.currentTime;
for (const range of deletedRanges) {
if (t >= range.start && t < range.end) {
video.currentTime = range.end;
return;
}
}
setCurrentTime(t);
});
};
const onPlay = () => setIsPlaying(true);
const onPause = () => setIsPlaying(false);
const onLoadedMetadata = () => setDuration(video.duration);
video.addEventListener('timeupdate', onTimeUpdate);
video.addEventListener('play', onPlay);
video.addEventListener('pause', onPause);
video.addEventListener('loadedmetadata', onLoadedMetadata);
return () => {
video.removeEventListener('timeupdate', onTimeUpdate);
video.removeEventListener('play', onPlay);
video.removeEventListener('pause', onPause);
video.removeEventListener('loadedmetadata', onLoadedMetadata);
cancelAnimationFrame(rafRef.current);
};
}, [videoRef, deletedRanges, setCurrentTime, setIsPlaying, setDuration]);
return { seekTo, togglePlay };
}

37
frontend/src/index.css Normal file
View File

@ -0,0 +1,37 @@
@tailwind base;
@tailwind components;
@tailwind utilities;
* {
margin: 0;
padding: 0;
box-sizing: border-box;
}
body {
font-family: 'Inter', system-ui, -apple-system, sans-serif;
overflow: hidden;
user-select: none;
}
::-webkit-scrollbar {
width: 6px;
height: 6px;
}
::-webkit-scrollbar-track {
background: transparent;
}
::-webkit-scrollbar-thumb {
background: #2a2d3a;
border-radius: 3px;
}
::-webkit-scrollbar-thumb:hover {
background: #3a3d4a;
}
video::-webkit-media-controls {
display: none !important;
}

10
frontend/src/main.tsx Normal file
View File

@ -0,0 +1,10 @@
import React from 'react';
import ReactDOM from 'react-dom/client';
import App from './App';
import './index.css';
ReactDOM.createRoot(document.getElementById('root')!).render(
<React.StrictMode>
<App />
</React.StrictMode>,
);

View File

@ -0,0 +1,129 @@
import { create } from 'zustand';
import { persist } from 'zustand/middleware';
import type { AIProvider, AIProviderConfig, FillerWordResult, ClipSuggestion } from '../types/project';
const ENCRYPTED_KEY_PREFIX = 'aive_enc_';
interface AIState {
providers: Record<AIProvider, AIProviderConfig>;
defaultProvider: AIProvider;
customFillerWords: string;
fillerResult: FillerWordResult | null;
clipSuggestions: ClipSuggestion[];
isProcessing: boolean;
processingMessage: string;
_keysHydrated: boolean;
}
interface AIActions {
setProviderConfig: (provider: AIProvider, config: Partial<AIProviderConfig>) => void;
setDefaultProvider: (provider: AIProvider) => void;
setCustomFillerWords: (words: string) => void;
setFillerResult: (result: FillerWordResult | null) => void;
setClipSuggestions: (suggestions: ClipSuggestion[]) => void;
setProcessing: (active: boolean, message?: string) => void;
hydrateKeys: () => Promise<void>;
}
async function encryptAndStore(key: string, value: string): Promise<void> {
if (!value) {
localStorage.removeItem(ENCRYPTED_KEY_PREFIX + key);
return;
}
if (window.electronAPI) {
const encrypted = await window.electronAPI.encryptString(value);
localStorage.setItem(ENCRYPTED_KEY_PREFIX + key, encrypted);
} else {
localStorage.setItem(ENCRYPTED_KEY_PREFIX + key, btoa(value));
}
}
async function loadAndDecrypt(key: string): Promise<string> {
const stored = localStorage.getItem(ENCRYPTED_KEY_PREFIX + key);
if (!stored) return '';
if (window.electronAPI) {
try {
return await window.electronAPI.decryptString(stored);
} catch {
return '';
}
}
try {
return atob(stored);
} catch {
return '';
}
}
export const useAIStore = create<AIState & AIActions>()(
persist(
(set, get) => ({
providers: {
ollama: { provider: 'ollama', baseUrl: 'http://localhost:11434', model: 'llama3' },
openai: { provider: 'openai', apiKey: '', model: 'gpt-4o' },
claude: { provider: 'claude', apiKey: '', model: 'claude-sonnet-4-20250514' },
},
defaultProvider: 'ollama',
customFillerWords: '',
fillerResult: null,
clipSuggestions: [],
isProcessing: false,
processingMessage: '',
_keysHydrated: false,
setProviderConfig: (provider, config) => {
set((state) => ({
providers: {
...state.providers,
[provider]: { ...state.providers[provider], ...config },
},
}));
if (config.apiKey !== undefined) {
encryptAndStore(`${provider}_apiKey`, config.apiKey);
}
},
setDefaultProvider: (provider) => set({ defaultProvider: provider }),
setCustomFillerWords: (words) => set({ customFillerWords: words }),
setFillerResult: (result) => set({ fillerResult: result }),
setClipSuggestions: (suggestions) => set({ clipSuggestions: suggestions }),
setProcessing: (active, message) =>
set({ isProcessing: active, processingMessage: message ?? '' }),
hydrateKeys: async () => {
const [openaiKey, claudeKey] = await Promise.all([
loadAndDecrypt('openai_apiKey'),
loadAndDecrypt('claude_apiKey'),
]);
const state = get();
set({
providers: {
...state.providers,
openai: { ...state.providers.openai, apiKey: openaiKey },
claude: { ...state.providers.claude, apiKey: claudeKey },
},
_keysHydrated: true,
});
},
}),
{
name: 'aive-ai-settings',
partialize: (state) => ({
providers: {
ollama: { ...state.providers.ollama, apiKey: undefined },
openai: { ...state.providers.openai, apiKey: '' },
claude: { ...state.providers.claude, apiKey: '' },
},
defaultProvider: state.defaultProvider,
customFillerWords: state.customFillerWords,
}),
},
),
);
useAIStore.getState().hydrateKeys();

View File

@ -0,0 +1,232 @@
import { create } from 'zustand';
import { temporal } from 'zundo';
import type { Word, Segment, DeletedRange, TranscriptionResult } from '../types/project';
interface EditorState {
videoPath: string | null;
videoUrl: string | null;
words: Word[];
segments: Segment[];
deletedRanges: DeletedRange[];
language: string;
currentTime: number;
duration: number;
isPlaying: boolean;
selectedWordIndices: number[];
hoveredWordIndex: number | null;
isTranscribing: boolean;
transcriptionProgress: number;
isExporting: boolean;
exportProgress: number;
backendUrl: string;
}
interface EditorActions {
setBackendUrl: (url: string) => void;
loadVideo: (path: string) => void;
setTranscription: (result: TranscriptionResult) => void;
setCurrentTime: (time: number) => void;
setDuration: (duration: number) => void;
setIsPlaying: (playing: boolean) => void;
setSelectedWordIndices: (indices: number[]) => void;
setHoveredWordIndex: (index: number | null) => void;
deleteSelectedWords: () => void;
deleteWordRange: (startIndex: number, endIndex: number) => void;
restoreRange: (rangeId: string) => void;
setTranscribing: (active: boolean, progress?: number) => void;
setExporting: (active: boolean, progress?: number) => void;
getKeepSegments: () => Array<{ start: number; end: number }>;
getWordAtTime: (time: number) => number;
loadProject: (projectData: any) => void;
reset: () => void;
}
const initialState: EditorState = {
videoPath: null,
videoUrl: null,
words: [],
segments: [],
deletedRanges: [],
language: '',
currentTime: 0,
duration: 0,
isPlaying: false,
selectedWordIndices: [],
hoveredWordIndex: null,
isTranscribing: false,
transcriptionProgress: 0,
isExporting: false,
exportProgress: 0,
backendUrl: 'http://localhost:8642',
};
let nextRangeId = 1;
export const useEditorStore = create<EditorState & EditorActions>()(
temporal(
(set, get) => ({
...initialState,
setBackendUrl: (url) => set({ backendUrl: url }),
loadVideo: (path) => {
const backend = get().backendUrl;
const url = `${backend}/file?path=${encodeURIComponent(path)}`;
set({
...initialState,
backendUrl: backend,
videoPath: path,
videoUrl: url,
});
},
setTranscription: (result) => {
let globalIdx = 0;
const annotatedSegments = result.segments.map((seg) => {
const annotated = { ...seg, globalStartIndex: globalIdx };
globalIdx += seg.words.length;
return annotated;
});
set({
words: result.words,
segments: annotatedSegments,
language: result.language,
deletedRanges: [],
selectedWordIndices: [],
});
},
setCurrentTime: (time) => set({ currentTime: time }),
setDuration: (duration) => set({ duration }),
setIsPlaying: (playing) => set({ isPlaying: playing }),
setSelectedWordIndices: (indices) => set({ selectedWordIndices: indices }),
setHoveredWordIndex: (index) => set({ hoveredWordIndex: index }),
deleteSelectedWords: () => {
const { selectedWordIndices, words, deletedRanges } = get();
if (selectedWordIndices.length === 0) return;
const sorted = [...selectedWordIndices].sort((a, b) => a - b);
const startWord = words[sorted[0]];
const endWord = words[sorted[sorted.length - 1]];
const newRange: DeletedRange = {
id: `dr_${nextRangeId++}`,
start: startWord.start,
end: endWord.end,
wordIndices: sorted,
};
set({
deletedRanges: [...deletedRanges, newRange],
selectedWordIndices: [],
});
},
deleteWordRange: (startIndex, endIndex) => {
const { words, deletedRanges } = get();
const indices = [];
for (let i = startIndex; i <= endIndex; i++) indices.push(i);
const newRange: DeletedRange = {
id: `dr_${nextRangeId++}`,
start: words[startIndex].start,
end: words[endIndex].end,
wordIndices: indices,
};
set({ deletedRanges: [...deletedRanges, newRange] });
},
restoreRange: (rangeId) => {
const { deletedRanges } = get();
set({ deletedRanges: deletedRanges.filter((r) => r.id !== rangeId) });
},
setTranscribing: (active, progress) =>
set({
isTranscribing: active,
transcriptionProgress: progress ?? (active ? 0 : 100),
}),
setExporting: (active, progress) =>
set({
isExporting: active,
exportProgress: progress ?? (active ? 0 : 100),
}),
getKeepSegments: () => {
const { words, deletedRanges, duration } = get();
if (words.length === 0) return [{ start: 0, end: duration }];
const deletedSet = new Set<number>();
for (const range of deletedRanges) {
for (const idx of range.wordIndices) deletedSet.add(idx);
}
const segments: Array<{ start: number; end: number }> = [];
let segStart: number | null = null;
for (let i = 0; i < words.length; i++) {
if (!deletedSet.has(i)) {
if (segStart === null) segStart = words[i].start;
} else {
if (segStart !== null) {
segments.push({ start: segStart, end: words[i - 1].end });
segStart = null;
}
}
}
if (segStart !== null) {
segments.push({ start: segStart, end: words[words.length - 1].end });
}
return segments;
},
getWordAtTime: (time) => {
const { words } = get();
let lo = 0;
let hi = words.length - 1;
while (lo <= hi) {
const mid = (lo + hi) >>> 1;
if (words[mid].end < time) lo = mid + 1;
else if (words[mid].start > time) hi = mid - 1;
else return mid;
}
return lo < words.length ? lo : words.length - 1;
},
loadProject: (data) => {
const backend = get().backendUrl;
const url = `${backend}/file?path=${encodeURIComponent(data.videoPath)}`;
let globalIdx = 0;
const annotatedSegments = (data.segments || []).map((seg: Segment) => {
const annotated = { ...seg, globalStartIndex: globalIdx };
globalIdx += seg.words.length;
return annotated;
});
set({
...initialState,
backendUrl: backend,
videoPath: data.videoPath,
videoUrl: url,
words: data.words || [],
segments: annotatedSegments,
deletedRanges: data.deletedRanges || [],
language: data.language || '',
});
},
reset: () => set(initialState),
}),
{ limit: 100 },
),
);

View File

@ -0,0 +1,86 @@
export interface Word {
word: string;
start: number;
end: number;
confidence: number;
speaker?: string;
}
export interface Segment {
id: number;
start: number;
end: number;
text: string;
words: Word[];
speaker?: string;
globalStartIndex: number;
}
export interface TimeRange {
start: number;
end: number;
}
export interface DeletedRange extends TimeRange {
id: string;
wordIndices: number[];
}
export interface ProjectFile {
version: 1;
videoPath: string;
words: Word[];
segments: Segment[];
deletedRanges: DeletedRange[];
language: string;
createdAt: string;
modifiedAt: string;
}
export interface TranscriptionResult {
words: Word[];
segments: Segment[];
language: string;
}
export interface ExportOptions {
outputPath: string;
mode: 'fast' | 'reencode';
resolution: '720p' | '1080p' | '4k';
format: 'mp4' | 'mov' | 'webm';
enhanceAudio: boolean;
captions: 'none' | 'burn-in' | 'sidecar';
captionStyle?: CaptionStyle;
}
export interface CaptionStyle {
fontName: string;
fontSize: number;
fontColor: string;
backgroundColor: string;
position: 'bottom' | 'top' | 'center';
bold: boolean;
}
export type AIProvider = 'ollama' | 'openai' | 'claude';
export interface AIProviderConfig {
provider: AIProvider;
apiKey?: string;
baseUrl?: string;
model: string;
}
export interface FillerWordResult {
wordIndices: number[];
fillerWords: Array<{ index: number; word: string; reason: string }>;
}
export interface ClipSuggestion {
title: string;
startWordIndex: number;
endWordIndex: number;
startTime: number;
endTime: number;
reason: string;
}

16
frontend/src/vite-env.d.ts vendored Normal file
View File

@ -0,0 +1,16 @@
/// <reference types="vite/client" />
interface ElectronAPI {
openFile: (options?: Record<string, unknown>) => Promise<string | null>;
saveFile: (options?: Record<string, unknown>) => Promise<string | null>;
openProject: () => Promise<string | null>;
getBackendUrl: () => Promise<string>;
encryptString: (data: string) => Promise<string>;
decryptString: (encrypted: string) => Promise<string>;
readFile: (path: string) => Promise<string>;
writeFile: (path: string, content: string) => Promise<boolean>;
}
interface Window {
electronAPI?: ElectronAPI;
}