Spaces:
Running on Zero
Running on Zero
| <html lang="en"> | |
| <head> | |
| <meta charset="UTF-8" /> | |
| <meta name="viewport" content="width=device-width, initial-scale=1.0" /> | |
| <title>Unlimited OCR – Baidu</title> | |
| <meta name="description" content="Unlimited OCR: One-shot long-horizon document parsing powered by Baidu's vision-language model." /> | |
| <link rel="preconnect" href="https://fonts.googleapis.com" /> | |
| <link rel="preconnect" href="https://fonts.gstatic.com" crossorigin /> | |
| <link href="https://fonts.googleapis.com/css2?family=JetBrains+Mono:wght@300;400;500;600&family=Inter:wght@300;400;500;600;700&display=swap" rel="stylesheet" /> | |
| <style> | |
| *, *::before, *::after { box-sizing: border-box; margin: 0; padding: 0; } | |
| :root { | |
| --bg-deep: #0a0c10; | |
| --bg-panel: #0f1117; | |
| --bg-surface: #161b22; | |
| --bg-card: #1c2230; | |
| --bg-hover: #21283a; | |
| --border: #2a3347; | |
| --border-soft: #1e2635; | |
| --accent: #ff6b35; | |
| --accent-glow: rgba(255,107,53,0.25); | |
| --teal: #00d4aa; | |
| --teal-dim: #008f73; | |
| --teal-glow: rgba(0,212,170,0.2); | |
| --blue: #4a9eff; | |
| --red: #ef4444; | |
| --text-primary: #e8eaf0; | |
| --text-secondary:#8892a4; | |
| --text-muted: #4a5568; | |
| --font-mono: 'JetBrains Mono', monospace; | |
| --font-ui: 'Inter', sans-serif; | |
| --radius: 8px; | |
| --radius-lg: 12px; | |
| } | |
| html, body { height: 100%; background: var(--bg-deep); color: var(--text-primary); font-family: var(--font-ui); overflow: hidden; } | |
| .app { display: flex; flex-direction: column; height: 100vh; } | |
| /* ── Header ── */ | |
| header { | |
| background: var(--bg-panel); | |
| border-bottom: 1px solid var(--border); | |
| padding: 0 20px; | |
| height: 52px; | |
| display: flex; | |
| align-items: center; | |
| gap: 16px; | |
| flex-shrink: 0; | |
| } | |
| .logo-area { display: flex; align-items: center; gap: 10px; } | |
| .logo-icon { | |
| width: 28px; height: 28px; | |
| background: linear-gradient(135deg, var(--accent) 0%, #ff9f5a 100%); | |
| border-radius: 6px; | |
| display: flex; align-items: center; justify-content: center; | |
| font-size: 14px; | |
| box-shadow: 0 0 12px var(--accent-glow); | |
| } | |
| .logo-text { font-size: 14px; font-weight: 600; letter-spacing: 0.01em; } | |
| .logo-text span { color: var(--accent); } | |
| .header-sep { width: 1px; height: 24px; background: var(--border); } | |
| .mode-group { | |
| display: flex; gap: 4px; | |
| background: var(--bg-deep); | |
| border: 1px solid var(--border); | |
| border-radius: var(--radius); | |
| padding: 3px; | |
| } | |
| .mode-btn { | |
| padding: 4px 14px; | |
| border-radius: 5px; | |
| font-family: var(--font-mono); | |
| font-size: 11px; font-weight: 500; | |
| cursor: pointer; border: none; | |
| transition: all 0.18s ease; | |
| color: var(--text-secondary); | |
| background: transparent; | |
| letter-spacing: 0.04em; | |
| } | |
| .mode-btn.active { background: var(--bg-surface); color: var(--text-primary); box-shadow: 0 1px 6px rgba(0,0,0,0.4); } | |
| .mode-btn:hover:not(.active) { color: var(--text-primary); } | |
| .ngram-group { display: flex; align-items: center; gap: 8px; } | |
| .ngram-label { font-family: var(--font-mono); font-size: 10px; color: var(--text-muted); letter-spacing: 0.06em; } | |
| .toggle-wrap { position: relative; width: 36px; height: 20px; } | |
| .toggle-wrap input { display: none; } | |
| .toggle-track { position: absolute; inset: 0; background: var(--border); border-radius: 10px; cursor: pointer; transition: background 0.2s; } | |
| .toggle-track::after { content: ''; position: absolute; top: 3px; left: 3px; width: 14px; height: 14px; background: var(--text-secondary); border-radius: 50%; transition: transform 0.2s, background 0.2s; } | |
| .toggle-wrap input:checked + .toggle-track { background: var(--teal-dim); } | |
| .toggle-wrap input:checked + .toggle-track::after { transform: translateX(16px); background: var(--teal); } | |
| .header-actions { margin-left: auto; display: flex; align-items: center; gap: 8px; } | |
| .btn { display: flex; align-items: center; gap: 6px; padding: 6px 16px; border-radius: var(--radius); font-family: var(--font-mono); font-size: 11px; font-weight: 600; letter-spacing: 0.06em; cursor: pointer; border: none; transition: all 0.18s ease; } | |
| .btn-start { background: var(--teal); color: #000; box-shadow: 0 0 14px var(--teal-glow); } | |
| .btn-start:hover:not(:disabled) { background: #00eebc; box-shadow: 0 0 22px var(--teal-glow); transform: translateY(-1px); } | |
| .btn-start:disabled { opacity: 0.4; cursor: not-allowed; transform: none; } | |
| .btn-stop { background: var(--red); color: #fff; box-shadow: 0 0 12px rgba(239,68,68,0.3); } | |
| .btn-stop:hover:not(:disabled) { background: #f87171; transform: translateY(-1px); } | |
| .btn-stop:disabled { opacity: 0.35; cursor: not-allowed; transform: none; } | |
| .btn-clear { background: var(--bg-surface); color: var(--text-secondary); border: 1px solid var(--border); } | |
| .btn-clear:hover { color: var(--text-primary); border-color: var(--text-muted); } | |
| .btn-copy { background: var(--bg-surface); color: var(--text-secondary); border: 1px solid var(--border); } | |
| .btn-copy:hover { color: var(--blue); border-color: var(--blue); } | |
| /* ── Status bar ── */ | |
| .status-bar { | |
| background: var(--bg-panel); | |
| border-bottom: 1px solid var(--border-soft); | |
| padding: 0 20px; height: 28px; | |
| display: flex; align-items: center; gap: 12px; | |
| flex-shrink: 0; | |
| } | |
| .status-dot { width: 6px; height: 6px; border-radius: 50%; background: var(--text-muted); transition: background 0.3s; } | |
| .status-dot.idle { background: var(--text-muted); } | |
| .status-dot.running { background: var(--teal); animation: pulse 1.2s infinite; } | |
| .status-dot.done { background: var(--teal); } | |
| .status-dot.error { background: var(--red); } | |
| @keyframes pulse { 0%,100% { opacity:1; } 50% { opacity:0.5; } } | |
| .status-text { font-family: var(--font-mono); font-size: 10px; color: var(--text-muted); letter-spacing: 0.05em; } | |
| .status-meta { margin-left: auto; display: flex; gap: 16px; } | |
| .meta-chip { font-family: var(--font-mono); font-size: 10px; color: var(--text-muted); } | |
| .meta-chip span { color: var(--text-secondary); } | |
| /* ── Progress bar ── */ | |
| .progress-bar-wrap { height: 2px; background: var(--bg-deep); flex-shrink: 0; overflow: hidden; } | |
| .progress-bar { height: 100%; background: linear-gradient(90deg, var(--teal) 0%, var(--blue) 100%); width: 0%; transition: width 0.3s ease; } | |
| .progress-bar.indeterminate { width: 40%; animation: slide 1.4s ease-in-out infinite; } | |
| @keyframes slide { 0% { transform: translateX(-100%); } 100% { transform: translateX(350%); } } | |
| /* ── Panels ── */ | |
| .panels { flex: 1; display: grid; grid-template-columns: 1fr 1fr; overflow: hidden; } | |
| .panel { display: flex; flex-direction: column; overflow: hidden; } | |
| .panel-left { border-right: 1px solid var(--border); } | |
| .panel-header { | |
| background: var(--bg-surface); | |
| border-bottom: 1px solid var(--border-soft); | |
| padding: 8px 16px; | |
| display: flex; align-items: center; gap: 10px; | |
| flex-shrink: 0; | |
| } | |
| .panel-label { font-family: var(--font-mono); font-size: 10px; font-weight: 600; letter-spacing: 0.08em; color: var(--text-muted); } | |
| .panel-tag { font-family: var(--font-mono); font-size: 9px; padding: 2px 7px; border-radius: 3px; background: var(--bg-deep); border: 1px solid var(--border); color: var(--text-muted); } | |
| .panel-tag.active { border-color: var(--accent); color: var(--accent); background: rgba(255,107,53,0.08); } | |
| .panel-actions { margin-left: auto; display: flex; gap: 6px; } | |
| .icon-btn { width: 24px; height: 24px; display: flex; align-items: center; justify-content: center; border-radius: 4px; background: transparent; border: 1px solid transparent; cursor: pointer; color: var(--text-muted); font-size: 12px; transition: all 0.15s; } | |
| .icon-btn:hover { background: var(--bg-hover); color: var(--text-primary); border-color: var(--border); } | |
| /* ── Drop zone ── */ | |
| .drop-zone { | |
| flex: 1; | |
| display: flex; flex-direction: column; align-items: center; justify-content: center; gap: 16px; | |
| padding: 32px; | |
| border: 2px dashed var(--border); | |
| margin: 16px; border-radius: var(--radius-lg); | |
| cursor: pointer; transition: all 0.2s; text-align: center; | |
| } | |
| .drop-zone.dragging { border-color: var(--accent); background: rgba(255,107,53,0.04); } | |
| .drop-icon { font-size: 40px; opacity: 0.3; filter: grayscale(1); } | |
| .drop-title { font-size: 14px; font-weight: 500; color: var(--text-secondary); } | |
| .drop-hint { font-size: 11px; color: var(--text-muted); font-family: var(--font-mono); } | |
| .drop-formats { display: flex; gap: 6px; } | |
| .format-chip { font-family: var(--font-mono); font-size: 9px; padding: 2px 8px; border-radius: 3px; background: var(--bg-deep); border: 1px solid var(--border); color: var(--text-muted); } | |
| #image-preview { width: 100%; height: 100%; object-fit: contain; background: #000; display: none; } | |
| #image-preview.visible { display: block; } | |
| /* PDF.js canvas viewer */ | |
| #pdf-canvas-wrap { | |
| display: none; | |
| flex-direction: column; | |
| align-items: center; | |
| gap: 10px; | |
| overflow-y: auto; | |
| padding: 12px; | |
| height: 100%; | |
| background: #525659; | |
| } | |
| #pdf-canvas-wrap.visible { display: flex; } | |
| #pdf-canvas-wrap canvas { | |
| display: block; | |
| width: 100%; | |
| max-width: 680px; | |
| box-shadow: 0 2px 10px rgba(0,0,0,0.6); | |
| border-radius: 2px; | |
| } | |
| /* ── Output ── */ | |
| .output-area { | |
| flex: 1; overflow-y: auto; | |
| padding: 16px 20px; | |
| font-family: var(--font-mono); font-size: 12px; line-height: 1.7; | |
| color: var(--text-primary); position: relative; | |
| } | |
| .output-area::-webkit-scrollbar { width: 6px; } | |
| .output-area::-webkit-scrollbar-thumb { background: var(--border); border-radius: 3px; } | |
| .output-placeholder { position: absolute; inset: 0; display: flex; flex-direction: column; align-items: center; justify-content: center; gap: 12px; color: var(--text-muted); } | |
| .output-placeholder svg { opacity: 0.2; } | |
| .output-placeholder p { font-size: 12px; font-family: var(--font-mono); } | |
| #ocr-text { white-space: pre-wrap; word-break: break-word; } | |
| .cursor { display: inline-block; width: 8px; height: 14px; background: var(--teal); vertical-align: middle; margin-left: 2px; border-radius: 1px; animation: blink 0.8s step-end infinite; } | |
| @keyframes blink { 0%,100% { opacity:1; } 50% { opacity:0; } } | |
| /* ── Prompt bar ── */ | |
| .prompt-bar { background: var(--bg-surface); border-top: 1px solid var(--border-soft); padding: 8px 16px; display: flex; align-items: center; gap: 10px; flex-shrink: 0; } | |
| .prompt-label { font-family: var(--font-mono); font-size: 10px; color: var(--text-muted); white-space: nowrap; } | |
| .prompt-input { flex: 1; background: var(--bg-deep); border: 1px solid var(--border); border-radius: 5px; padding: 4px 10px; font-family: var(--font-mono); font-size: 11px; color: var(--text-primary); outline: none; transition: border-color 0.2s; } | |
| .prompt-input:focus { border-color: var(--accent); } | |
| .token-counter { font-family: var(--font-mono); font-size: 10px; color: var(--teal); margin-left: auto; } | |
| /* ── Toast ── */ | |
| .toast { position: fixed; bottom: 20px; left: 50%; transform: translateX(-50%) translateY(60px); background: var(--bg-card); border: 1px solid var(--border); border-radius: var(--radius); padding: 8px 20px; font-family: var(--font-mono); font-size: 11px; color: var(--text-primary); opacity: 0; transition: all 0.3s ease; z-index: 100; white-space: nowrap; } | |
| .toast.show { transform: translateX(-50%) translateY(0); opacity: 1; } | |
| * { scrollbar-width: thin; scrollbar-color: var(--border) transparent; } | |
| </style> | |
| </head> | |
| <body> | |
| <div class="app"> | |
| <!-- Header --> | |
| <header> | |
| <div class="logo-area"> | |
| <div class="logo-icon">📄</div> | |
| <span class="logo-text">Unlimited<span>-OCR</span></span> | |
| </div> | |
| <div class="header-sep"></div> | |
| <div class="mode-group" role="group" aria-label="Inference mode"> | |
| <button class="mode-btn active" id="btn-long" onclick="setMode('long')">Long</button> | |
| <button class="mode-btn" id="btn-base" onclick="setMode('base')">Base</button> | |
| </div> | |
| <div class="header-sep"></div> | |
| <div class="ngram-group"> | |
| <span class="ngram-label">NGRAM</span> | |
| <label class="toggle-wrap"> | |
| <input type="checkbox" id="ngram-toggle" checked /> | |
| <div class="toggle-track"></div> | |
| </label> | |
| </div> | |
| <div class="header-actions"> | |
| <button class="btn btn-clear" onclick="clearAll()"> | |
| <svg width="12" height="12" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2"><polyline points="1 4 1 10 7 10"/><path d="M3.51 15a9 9 0 1 0 .49-3.51"/></svg> | |
| RESET | |
| </button> | |
| <button class="btn btn-copy" id="copy-btn" onclick="copyText()" disabled> | |
| <svg width="12" height="12" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2"><rect x="9" y="9" width="13" height="13" rx="2"/><path d="M5 15H4a2 2 0 0 1-2-2V4a2 2 0 0 1 2-2h9a2 2 0 0 1 2 2v1"/></svg> | |
| COPY | |
| </button> | |
| <button class="btn btn-start" id="start-btn" onclick="startOCR()" disabled> | |
| <svg width="12" height="12" viewBox="0 0 24 24" fill="currentColor"><polygon points="5 3 19 12 5 21 5 3"/></svg> | |
| START | |
| </button> | |
| <button class="btn btn-stop" id="stop-btn" onclick="stopOCR()" disabled> | |
| <svg width="12" height="12" viewBox="0 0 24 24" fill="currentColor"><rect x="6" y="6" width="12" height="12"/></svg> | |
| STOP | |
| </button> | |
| </div> | |
| </header> | |
| <!-- Status bar --> | |
| <div class="status-bar"> | |
| <div class="status-dot idle" id="status-dot"></div> | |
| <span class="status-text" id="status-text">IDLE · drop or click to load a document</span> | |
| <div class="status-meta"> | |
| <span class="meta-chip">MODEL <span>baidu/Unlimited-OCR</span></span> | |
| <span class="meta-chip">MODE <span id="mode-display">LONG</span></span> | |
| <span class="meta-chip" id="page-info" style="display:none">PAGES <span id="page-count">1</span></span> | |
| <span class="meta-chip token-counter" id="token-display" style="display:none"><span id="char-count">0</span> chars</span> | |
| </div> | |
| </div> | |
| <!-- Progress bar --> | |
| <div class="progress-bar-wrap"> | |
| <div class="progress-bar" id="progress-bar"></div> | |
| </div> | |
| <!-- Panels --> | |
| <div class="panels"> | |
| <!-- Left: input --> | |
| <div class="panel panel-left"> | |
| <div class="panel-header"> | |
| <span class="panel-label">INPUT</span> | |
| <span class="panel-tag active" id="input-type-tag">EMPTY</span> | |
| <div class="panel-actions"> | |
| <button class="icon-btn" title="Upload file" onclick="document.getElementById('file-input').click()"> | |
| <svg width="13" height="13" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2"><path d="M21 15v4a2 2 0 0 1-2 2H5a2 2 0 0 1-2-2v-4"/><polyline points="17 8 12 3 7 8"/><line x1="12" y1="3" x2="12" y2="15"/></svg> | |
| </button> | |
| <button class="icon-btn" title="Toggle zoom" id="zoom-btn" onclick="toggleZoom()" style="display:none"> | |
| <svg width="13" height="13" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2"><polyline points="15 3 21 3 21 9"/><polyline points="9 21 3 21 3 15"/><line x1="21" y1="3" x2="14" y2="10"/><line x1="3" y1="21" x2="10" y2="14"/></svg> | |
| </button> | |
| </div> | |
| </div> | |
| <div class="drop-zone" id="drop-zone" | |
| onclick="document.getElementById('file-input').click()" | |
| ondragover="onDragOver(event)" | |
| ondragleave="onDragLeave(event)" | |
| ondrop="onDrop(event)"> | |
| <div class="drop-icon">📄</div> | |
| <div class="drop-title">Drop your document here</div> | |
| <p class="drop-hint">or click anywhere to browse</p> | |
| <div class="drop-formats"> | |
| <span class="format-chip">PNG</span> | |
| <span class="format-chip">JPG</span> | |
| <span class="format-chip">WEBP</span> | |
| <span class="format-chip">PDF</span> | |
| <span class="format-chip">TIFF</span> | |
| </div> | |
| </div> | |
| <img id="image-preview" alt="Document preview" /> | |
| <div id="pdf-canvas-wrap"></div> | |
| <div class="prompt-bar"> | |
| <span class="prompt-label">PROMPT</span> | |
| <input class="prompt-input" id="prompt-input" value="document parsing." placeholder="e.g. document parsing." /> | |
| </div> | |
| </div> | |
| <!-- Right: output --> | |
| <div class="panel"> | |
| <div class="panel-header"> | |
| <span class="panel-label">OCR OUTPUT</span> | |
| <span class="panel-tag" id="output-tag">WAITING</span> | |
| <div class="panel-actions" style="margin-left:auto"> | |
| <button class="icon-btn" title="Scroll to top" onclick="scrollOutputTop()"> | |
| <svg width="13" height="13" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2"><polyline points="17 11 12 6 7 11"/><polyline points="17 18 12 13 7 18"/></svg> | |
| </button> | |
| <button class="icon-btn" title="Scroll to bottom" onclick="scrollOutputBottom()"> | |
| <svg width="13" height="13" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2"><polyline points="7 13 12 18 17 13"/><polyline points="7 6 12 11 17 6"/></svg> | |
| </button> | |
| </div> | |
| </div> | |
| <div class="output-area" id="output-area"> | |
| <div class="output-placeholder" id="output-placeholder"> | |
| <svg width="48" height="48" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1"><rect x="3" y="3" width="18" height="18" rx="2"/><line x1="3" y1="9" x2="21" y2="9"/><line x1="3" y1="15" x2="21" y2="15"/><line x1="9" y1="9" x2="9" y2="21"/></svg> | |
| <p>OCR output will appear here</p> | |
| <p style="color:var(--border);font-size:10px">Load a document and press START</p> | |
| </div> | |
| <div id="ocr-text"></div> | |
| </div> | |
| </div> | |
| </div> | |
| </div> | |
| <input type="file" id="file-input" style="display:none" accept="image/*,.pdf" onchange="handleFileInput(event)" /> | |
| <div class="toast" id="toast"></div> | |
| <script type="module"> | |
| import { Client, handle_file } from "https://cdn.jsdelivr.net/npm/@gradio/client/dist/index.min.js"; | |
| import * as pdfjsLib from "https://cdn.jsdelivr.net/npm/pdfjs-dist@4.4.168/build/pdf.min.mjs"; | |
| pdfjsLib.GlobalWorkerOptions.workerSrc = | |
| "https://cdn.jsdelivr.net/npm/pdfjs-dist@4.4.168/build/pdf.worker.min.mjs"; | |
| // ── State ───────────────────────────────────────────────────────────────────── | |
| let client = null; | |
| let currentMode = 'long'; | |
| let currentFile = null; | |
| let isPDF = false; | |
| let isRunning = false; | |
| let stopRequested = false; | |
| let charCount = 0; | |
| let zoomedOut = false; | |
| // ── Boot ────────────────────────────────────────────────────────────────────── | |
| async function boot() { | |
| try { | |
| client = await Client.connect(window.location.origin); | |
| setStatus('idle', 'READY · model loaded'); | |
| } catch (e) { | |
| setStatus('error', 'CONNECTION ERROR · ' + e.message); | |
| } | |
| } | |
| boot(); | |
| // ── Global handlers (called from onclick="…") ───────────────────────────────── | |
| window.setMode = (m) => { | |
| currentMode = m; | |
| document.getElementById('btn-long').classList.toggle('active', m === 'long'); | |
| document.getElementById('btn-base').classList.toggle('active', m === 'base'); | |
| document.getElementById('mode-display').textContent = m.toUpperCase(); | |
| }; | |
| window.handleFileInput = (e) => { if (e.target.files[0]) loadFile(e.target.files[0]); }; | |
| window.onDragOver = (e) => { | |
| e.preventDefault(); | |
| document.getElementById('drop-zone').classList.add('dragging'); | |
| }; | |
| window.onDragLeave = () => { | |
| document.getElementById('drop-zone').classList.remove('dragging'); | |
| }; | |
| window.onDrop = (e) => { | |
| e.preventDefault(); | |
| document.getElementById('drop-zone').classList.remove('dragging'); | |
| if (e.dataTransfer.files[0]) loadFile(e.dataTransfer.files[0]); | |
| }; | |
| window.startOCR = startOCR; | |
| window.stopOCR = stopOCR; | |
| window.clearAll = clearAll; | |
| window.copyText = copyText; | |
| window.toggleZoom = () => { | |
| zoomedOut = !zoomedOut; | |
| document.getElementById('image-preview').style.objectFit = zoomedOut ? 'cover' : 'contain'; | |
| }; | |
| window.scrollOutputTop = () => { document.getElementById('output-area').scrollTop = 0; }; | |
| window.scrollOutputBottom = () => { | |
| const a = document.getElementById('output-area'); | |
| a.scrollTop = a.scrollHeight; | |
| }; | |
| // ── File loading ────────────────────────────────────────────────────────────── | |
| async function loadFile(file) { | |
| currentFile = file; | |
| isPDF = file.type === 'application/pdf' || file.name.toLowerCase().endsWith('.pdf'); | |
| // Hide drop zone, reset previews | |
| document.getElementById('drop-zone').style.display = 'none'; | |
| document.getElementById('image-preview').classList.remove('visible'); | |
| document.getElementById('pdf-canvas-wrap').classList.remove('visible'); | |
| document.getElementById('zoom-btn').style.display = 'flex'; | |
| document.getElementById('input-type-tag').textContent = | |
| isPDF ? 'PDF' : file.name.split('.').pop().toUpperCase(); | |
| if (isPDF) { | |
| // Render PDF pages with PDF.js into canvases (blob: URLs are blocked in iframes on HTTPS) | |
| const wrap = document.getElementById('pdf-canvas-wrap'); | |
| wrap.innerHTML = '<div style="color:var(--teal);font-family:var(--font-mono);font-size:11px;padding:20px">Rendering PDF…</div>'; | |
| wrap.classList.add('visible'); | |
| document.getElementById('page-info').style.display = 'flex'; | |
| const buffer = await file.arrayBuffer(); | |
| const pdf = await pdfjsLib.getDocument({ data: buffer }).promise; | |
| const total = pdf.numPages; | |
| document.getElementById('page-count').textContent = total; | |
| wrap.innerHTML = ''; | |
| for (let i = 1; i <= total; i++) { | |
| const page = await pdf.getPage(i); | |
| const viewport = page.getViewport({ scale: 2.0 }); | |
| const canvas = document.createElement('canvas'); | |
| canvas.width = viewport.width; | |
| canvas.height = viewport.height; | |
| await page.render({ canvasContext: canvas.getContext('2d'), viewport }).promise; | |
| wrap.appendChild(canvas); | |
| } | |
| } else { | |
| const preview = document.getElementById('image-preview'); | |
| preview.src = URL.createObjectURL(file); | |
| preview.classList.add('visible'); | |
| } | |
| setStatus('idle', `READY · ${file.name}`); | |
| document.getElementById('start-btn').disabled = false; | |
| clearOutput(); | |
| } | |
| // ── OCR entry ───────────────────────────────────────────────────────────────── | |
| async function startOCR() { | |
| if (!currentFile || isRunning) return; | |
| if (!client) { showToast('Client not ready yet'); return; } | |
| isRunning = true; | |
| stopRequested = false; | |
| charCount = 0; | |
| clearOutput(); | |
| document.getElementById('output-placeholder').style.display = 'none'; | |
| document.getElementById('output-tag').textContent = 'RUNNING'; | |
| document.getElementById('start-btn').disabled = true; | |
| document.getElementById('stop-btn').disabled = false; | |
| setProgressIndeterminate(true); | |
| const prompt = document.getElementById('prompt-input').value || 'document parsing.'; | |
| // Long mode → gundam (640px crop, fast, ZeroGPU-friendly) | |
| // Base mode → base (1024px, more accurate) | |
| const mode = currentMode === 'long' ? 'gundam' : 'base'; | |
| const ocrDiv = document.getElementById('ocr-text'); | |
| try { | |
| if (isPDF) { | |
| await runPDF(ocrDiv, prompt, mode); | |
| } else { | |
| await runSingleImage(ocrDiv, prompt, mode); | |
| } | |
| } finally { | |
| isRunning = false; | |
| setProgressIndeterminate(false); | |
| document.getElementById('start-btn').disabled = false; | |
| document.getElementById('stop-btn').disabled = true; | |
| } | |
| } | |
| // ── Single image — streaming via client.submit() ──────────────────────────── | |
| async function runSingleImage(ocrDiv, prompt, mode) { | |
| ocrDiv.innerHTML = '<span class="cursor"></span>'; | |
| setStatus('running', 'IN GPU QUEUE…'); | |
| const submission = client.submit('/run_ocr', { | |
| image_path: handle_file(currentFile), | |
| mode, | |
| prompt, | |
| }); | |
| let finalText = ''; | |
| try { | |
| for await (const msg of submission) { | |
| if (stopRequested) { submission.cancel(); return finishStop(ocrDiv); } | |
| if (msg.type === 'data') { | |
| const chunk = msg.data?.[0]?.text ?? msg.data?.[0] ?? ''; | |
| finalText = chunk; | |
| ocrDiv.textContent = chunk; | |
| window.scrollOutputBottom(); | |
| } | |
| if (msg.type === 'status' && msg.stage === 'complete') break; | |
| } | |
| finishDone(finalText); | |
| } catch (err) { | |
| if (stopRequested) return finishStop(ocrDiv); | |
| showError(ocrDiv, err.message); | |
| } | |
| } | |
| // ── PDF: CPU explode → one 60-second GPU slot per page ─────────────────────── | |
| // | |
| // ZeroGPU design pattern: | |
| // • Short bursts = higher queue priority | |
| // • One page/call = multiple users can be served concurrently | |
| // • PDF extraction is CPU-only (no GPU wasted on I/O) | |
| // | |
| async function runPDF(ocrDiv, prompt, mode) { | |
| setStatus('running', 'EXTRACTING PDF PAGES (CPU)…'); | |
| ocrDiv.innerHTML = '<span class="cursor"></span>'; | |
| let pages; | |
| try { | |
| const res = await client.predict('/explode_pdf', { | |
| pdf_file: handle_file(currentFile), | |
| }); | |
| pages = res?.data?.[0]?.pages ?? []; | |
| } catch (err) { | |
| if (stopRequested) return finishStop(ocrDiv); | |
| return showError(ocrDiv, 'PDF extraction failed: ' + err.message); | |
| } | |
| if (!pages.length) return showError(ocrDiv, 'No pages found in PDF.'); | |
| const total = pages.length; | |
| document.getElementById('page-count').textContent = total; | |
| document.getElementById('page-info').style.display = 'flex'; | |
| ocrDiv.textContent = ''; | |
| let fullText = ''; | |
| for (let i = 0; i < total; i++) { | |
| if (stopRequested) return finishStop(ocrDiv); | |
| setStatus('running', `GPU QUEUE · PAGE ${i + 1} / ${total}`); | |
| setProgressValue(Math.round((i / total) * 100)); | |
| // Section header | |
| const sep = document.createElement('div'); | |
| sep.style.cssText = | |
| 'margin:10px 0 5px;font-size:9px;letter-spacing:.1em;' + | |
| 'color:var(--text-muted);border-bottom:1px solid var(--border-soft);padding-bottom:3px;'; | |
| sep.textContent = `── PAGE ${i + 1} / ${total} ──`; | |
| ocrDiv.appendChild(sep); | |
| // Waiting-for-GPU indicator | |
| const slot = document.createElement('div'); | |
| slot.style.cssText = 'color:var(--text-muted);font-size:11px;padding:3px 0;'; | |
| slot.innerHTML = '<span class="cursor"></span> waiting for GPU slot…'; | |
| ocrDiv.appendChild(slot); | |
| window.scrollOutputBottom(); | |
| // Stream each page token-by-token using client.submit() | |
| const pageSub = client.submit('/run_ocr', { | |
| image_path: { path: pages[i].path, orig_name: pages[i].orig_name }, | |
| mode, | |
| prompt, | |
| }); | |
| try { | |
| let pageText = ''; | |
| for await (const msg of pageSub) { | |
| if (stopRequested) { pageSub.cancel(); slot.remove(); return finishStop(ocrDiv); } | |
| if (msg.type === 'data') { | |
| pageText = msg.data?.[0]?.text ?? msg.data?.[0] ?? ''; | |
| slot.style.color = 'var(--text-primary)'; | |
| slot.innerHTML = ''; | |
| slot.textContent = pageText; | |
| window.scrollOutputBottom(); | |
| } | |
| if (msg.type === 'status' && msg.stage === 'complete') break; | |
| } | |
| fullText += (fullText ? '\n\n' : '') + pageText; | |
| charCount = fullText.length; | |
| document.getElementById('char-count').textContent = charCount.toLocaleString(); | |
| document.getElementById('token-display').style.display = 'flex'; | |
| if (!pageText) slot.textContent = '(no text detected)'; | |
| } catch (err) { | |
| if (stopRequested) return finishStop(ocrDiv); | |
| slot.innerHTML = `<span style="color:var(--red)">Page ${i + 1} error: ${err.message}</span>`; | |
| } | |
| } | |
| finishDone(fullText); | |
| } | |
| // ── Stop ────────────────────────────────────────────────────────────────────── | |
| function stopOCR() { stopRequested = true; } | |
| function finishStop(ocrDiv) { | |
| ocrDiv.querySelectorAll('.cursor').forEach(c => c.remove()); | |
| setStatus('idle', 'STOPPED by user'); | |
| document.getElementById('output-tag').textContent = 'STOPPED'; | |
| setProgressIndeterminate(false); | |
| isRunning = false; | |
| document.getElementById('start-btn').disabled = false; | |
| document.getElementById('stop-btn').disabled = true; | |
| } | |
| // ── Done / error ────────────────────────────────────────────────────────────── | |
| function finishDone(text) { | |
| charCount = text.length; | |
| setProgressValue(100); | |
| setStatus('done', `DONE · ${charCount.toLocaleString()} chars`); | |
| document.getElementById('output-tag').textContent = 'DONE'; | |
| document.getElementById('token-display').style.display = 'flex'; | |
| document.getElementById('char-count').textContent = charCount.toLocaleString(); | |
| document.getElementById('copy-btn').disabled = false; | |
| showToast('OCR complete'); | |
| } | |
| function showError(ocrDiv, msg) { | |
| ocrDiv.innerHTML = `<span style="color:var(--red)">ERROR: ${msg}</span>`; | |
| setStatus('error', 'ERROR · ' + msg); | |
| document.getElementById('output-tag').textContent = 'ERROR'; | |
| } | |
| // ── Utilities ───────────────────────────────────────────────────────────────── | |
| function clearOutput() { | |
| document.getElementById('ocr-text').innerHTML = ''; | |
| charCount = 0; | |
| document.getElementById('output-placeholder').style.display = 'flex'; | |
| document.getElementById('output-tag').textContent = 'WAITING'; | |
| document.getElementById('token-display').style.display = 'none'; | |
| document.getElementById('copy-btn').disabled = true; | |
| setProgressValue(0); | |
| } | |
| function clearAll() { | |
| if (isRunning) stopOCR(); | |
| currentFile = null; isPDF = false; | |
| document.getElementById('drop-zone').style.display = ''; | |
| const p = document.getElementById('image-preview'); | |
| p.classList.remove('visible'); p.src = ''; | |
| const v = document.getElementById('pdf-canvas-wrap'); | |
| v.classList.remove('visible'); v.innerHTML = ''; | |
| document.getElementById('zoom-btn').style.display = 'none'; | |
| document.getElementById('input-type-tag').textContent = 'EMPTY'; | |
| document.getElementById('start-btn').disabled = true; | |
| document.getElementById('page-info').style.display = 'none'; | |
| document.getElementById('file-input').value = ''; | |
| clearOutput(); | |
| setStatus('idle', 'IDLE · drop or click to load a document'); | |
| } | |
| function copyText() { | |
| navigator.clipboard.writeText(document.getElementById('ocr-text').innerText) | |
| .then(() => showToast('Copied to clipboard')); | |
| } | |
| function setStatus(type, msg) { | |
| document.getElementById('status-dot').className = 'status-dot ' + type; | |
| document.getElementById('status-text').textContent = msg; | |
| } | |
| function setProgressIndeterminate(on) { | |
| const bar = document.getElementById('progress-bar'); | |
| bar.className = 'progress-bar' + (on ? ' indeterminate' : ''); | |
| if (!on) bar.style.width = '0%'; | |
| } | |
| function setProgressValue(v) { | |
| const bar = document.getElementById('progress-bar'); | |
| bar.className = 'progress-bar'; | |
| bar.style.width = v + '%'; | |
| } | |
| let _toastTimer = null; | |
| function showToast(msg) { | |
| const t = document.getElementById('toast'); | |
| t.textContent = msg; | |
| t.classList.add('show'); | |
| clearTimeout(_toastTimer); | |
| _toastTimer = setTimeout(() => t.classList.remove('show'), 2600); | |
| } | |
| </script> | |
| </body> | |
| </html> | |