Actualizar worker de compresión

README.md

@@ -1,46 +1,29 @@
 ---
-title: OmniVision AI - Neural Media Studio
+title: OmniVision AI Node
 emoji: 🧠
-colorFrom: purple
+colorFrom: blue
 colorTo: indigo
 sdk: docker
 app_port: 7860
 pinned: false
 license: mit
-short_description: High-performance neural video optimization engine
-tags:
-  - video
-  - deep-learning
-  - media-processing
-  - audio
-  - ffmpeg
-  - pytorch
 ---
 
-# OmniVision AI — Neural Media & Vision Studio 🧠
+# OmniVision AI — Runtime Node
 
-OmniVision AI is a high-performance media processing and neural video optimization engine built for high-throughput CPU-based container environments.
+Distributed media runtime node for the OmniVision processing platform. It maintains a secure connection to the configured coordinator, accepts assigned media tasks, and reports live status and resource utilization.
 
-## Capabilities & Architecture
+## Space configuration
 
-- **Perceptual Rate-Distortion Optimization (CRF)**: Intelligent perception-guided video compression maximizing PSNR and VMAF while reducing media payload by up to 75%.
-- **OrientAI Spatial Aspect Scaler**: Real-time orientation detection ensuring both vertical (9:16) and landscape (16:9) streams are scaled with strict even dimensions and zero aspect distortion.
-- **Embedded Subtitle Stream Synchronizer**: Automatic extraction and remuxing of multi-track text subtitles (`srt`, `ass`, `ssa`, `webvtt`) into standard MP4 `mov_text` streams.
-- **Acoustic Speech Intelligibility Enhancer**: Low-bitrate vocal optimization with dynamic mono channel doubling for crystal-clear dialog clarity below 64 kbps.
+Set these values under **Settings → Variables and secrets**:
 
-## Web Dashboard & API Endpoints
+| Name | Type | Required | Description |
+|---|---|---:|---|
+| `MANAGER_URL` | Variable | Yes | Public URL of the service coordinator |
+| `WORKER_SECRET` | Secret | Yes | Shared authentication secret for this node group |
+| `WORKER_ID` | Variable | Recommended | Unique identifier for this node |
+| `MAX_JOBS` | Variable | No | Maximum simultaneous tasks; defaults to `1` |
+| `SPACE_URL` | Variable | No | Public Space URL used for node identification |
+| `WORK_DIR` | Variable | No | Temporary working directory; defaults to `/tmp/omnivision_media` |
 
-This Space serves an interactive web dashboard on port `7860`:
-- **Web UI**: Access the root URL `/` to test the interactive neural pipeline benchmark.
-- **`GET /api/v1/health`**: Cluster health status, active worker threads, and memory metrics.
-- **`GET /api/v1/models`**: Active neural model pipelines and media codecs.
-
-## Configuration & Environment Variables
-
-You can optionally configure cluster synchronization in Space **Settings** -> **Variables and secrets**:
-
-| Variable | Description |
-|---|---|
-| `MANAGER_URL` | Optional upstream coordinator endpoint |
-| `WORKER_SECRET` | Authentication token for cluster nodes |
-| `WORK_DIR` | Scratch directory for media tensors |
+The node requires `MANAGER_URL` and `WORKER_SECRET` and will stop if either is missing. The operations console displays connection status, active tasks, queue depth, and node capacity.

internal/protocol/messages.go

@@ -80,10 +80,13 @@ type TaskAssign struct {
 // TaskProgress es el reporte en vivo que el Worker envía mientras FFmpeg procesa
 type TaskProgress struct {
 	TaskID  string  `json:"task_id"`
+	Stage   string  `json:"stage,omitempty"`
 	Percent float64 `json:"percent"`
 	Speed   string  `json:"speed"`
 	Time    string  `json:"time"`
 	ETA     string  `json:"eta"`
+	Current int64   `json:"current,omitempty"`
+	Total   int64   `json:"total,omitempty"`
 }
 
 // TaskComplete notifica que el Worker terminó y ya subió el archivo comprimido

main.go

@@ -2,7 +2,6 @@ package main
 
 import (
 	"context"
-	"encoding/base64"
 	"encoding/json"
 	"fmt"
 	"io"
@@ -87,18 +86,14 @@ func main() {
 
 	managerURL := os.Getenv("MANAGER_URL")
 	if managerURL == "" {
-		// Fallback decodificado
-		raw, _ := base64.StdEncoding.DecodeString("aHR0cHM6Ly93b3JrZXJzLnVkeWF0LnNpdGU=")
-		managerURL = string(raw)
+		log.Fatal("MANAGER_URL es obligatorio para conectar este worker al clúster")
 	} else if !strings.HasPrefix(managerURL, "http://") && !strings.HasPrefix(managerURL, "https://") {
 		managerURL = "https://" + managerURL
 	}
 
 	secret := os.Getenv("WORKER_SECRET")
 	if secret == "" {
-		// Fallback decodificado
-		raw, _ := base64.StdEncoding.DecodeString("YjlkZDFiZDQzMGI0NGIzYjYzYWNiNTFhMjk4MzA0Nzg2MmJkMTdhOWNlYTQ2NmY2YWJkZWViMTM0OTZkZTk5Mg==")
-		secret = string(raw)
+		log.Fatal("WORKER_SECRET es obligatorio; configura el secreto del clúster en los Secrets del Space")
 	}
 
 	workerID := os.Getenv("WORKER_ID")
@@ -478,7 +473,7 @@ func (w *Worker) executeTask(taskCtx context.Context, job *CurrentJob, task prot
 		return
 	}
 
-	_, err = io.Copy(inFile, resp.Body)
+	_, err = w.copyDownloadWithProgress(taskCtx, task.TaskID, inFile, resp.Body, resp.ContentLength)
 	inFile.Close()
 	if err != nil {
 		w.reportFailed(task.TaskID, fmt.Sprintf("error guardando stream de video: %v", err))
@@ -524,6 +519,7 @@ func (w *Worker) executeTask(taskCtx context.Context, job *CurrentJob, task prot
 		}
 	}, func(p protocol.TaskProgress) {
 		p.TaskID = task.TaskID
+		p.Stage = "compressing"
 		_ = w.sendMessage(protocol.TypeTaskProgress, p)
 	})
 
@@ -555,6 +551,7 @@ func (w *Worker) executeTask(taskCtx context.Context, job *CurrentJob, task prot
 
 	// 4. Subir archivo comprimido al Manager (con reintentos si el Manager está reiniciando)
 	uploadURL := w.resolveURL(task.UploadURL)
+	_ = w.sendMessage(protocol.TypeTaskProgress, protocol.TaskProgress{TaskID: task.TaskID, Stage: "uploading", Percent: 0})
 
 	var uploadSuccess bool
 	for attempt := 1; attempt <= 15; attempt++ {
@@ -620,6 +617,54 @@ func (w *Worker) executeTask(taskCtx context.Context, job *CurrentJob, task prot
 	log.Printf("[%s] ✅ Tarea finalizada y confirmada.", task.TaskID)
 }
 
+func (w *Worker) copyDownloadWithProgress(ctx context.Context, taskID string, dst io.Writer, src io.Reader, total int64) (int64, error) {
+	buf := make([]byte, 512*1024)
+	started, lastReport := time.Now(), time.Now()
+	var copied int64
+	for {
+		if err := ctx.Err(); err != nil {
+			return copied, err
+		}
+		n, readErr := src.Read(buf)
+		if n > 0 {
+			written, writeErr := dst.Write(buf[:n])
+			copied += int64(written)
+			if writeErr != nil {
+				return copied, writeErr
+			}
+			if written != n {
+				return copied, io.ErrShortWrite
+			}
+			if time.Since(lastReport) >= time.Second || (total > 0 && copied >= total) {
+				elapsed := time.Since(started).Seconds()
+				pct := float64(0)
+				if total > 0 {
+					pct = float64(copied) * 100 / float64(total)
+				}
+				p := protocol.TaskProgress{TaskID: taskID, Stage: "downloading", Percent: pct, Current: copied, Total: total}
+				if elapsed > 0 {
+					p.Speed = fmt.Sprintf("%.1f MB/s", float64(copied)/elapsed/(1024*1024))
+					if total > copied {
+						p.ETA = fmt.Sprintf("%ds", int(float64(total-copied)/(float64(copied)/elapsed)))
+					}
+				}
+				_ = w.sendMessage(protocol.TypeTaskProgress, p)
+				lastReport = time.Now()
+			}
+		}
+		if readErr == io.EOF {
+			break
+		}
+		if readErr != nil {
+			return copied, readErr
+		}
+	}
+	if total > 0 && copied != total {
+		return copied, io.ErrUnexpectedEOF
+	}
+	return copied, nil
+}
+
 // downloadSimpleFile descarga un archivo pequeño del Manager (logo de marca de agua).
 func (w *Worker) downloadSimpleFile(ctx context.Context, rawURL, destPath string) error {
 	req, err := http.NewRequestWithContext(ctx, "GET", rawURL, nil)

web_ui.go

@@ -90,8 +90,8 @@ func renderAIStudioHTML(cores int) string {
       </div>
       <div class="metric-card">
         <div class="metric-label">Perceptual Engine</div>
-        <div class="metric-value">Omni-CRF v2</div>
-        <div class="metric-sub">H.264 / AAC / TimedText</div>
+        <div class="metric-value">OmniVision v2</div>
+        <div class="metric-sub">Media Runtime / Adaptive Audio</div>
       </div>
       <div class="metric-card">
         <div class="metric-label">Spatial Adapter</div>
@@ -106,37 +106,37 @@ func renderAIStudioHTML(cores int) string {
     </div>
 
     <div class="card">
-      <div class="card-title">🎬 Interactive Neural Video Processing</div>
-      <div class="card-desc">Simulate and evaluate the OmniVision rate-distortion optimization pipeline on media assets.</div>
+      <div class="card-title">🎬 Interactive Neural Media Studio</div>
+      <div class="card-desc">Explore the OmniVision media analysis pipeline using a sample asset.</div>
 
       <div class="dropzone" id="dropzone" onclick="selectSample()">
         <svg viewBox="0 0 24 24"><path d="M19.35 10.04C18.67 6.59 15.64 4 12 4 9.11 4 6.6 5.64 5.35 8.04 2.34 8.36 0 10.91 0 14c0 3.31 2.69 6 6 6h13c2.76 0 5-2.24 5-5 0-2.64-2.05-4.78-4.65-4.96zM14 13v4h-4v-4H7l5-5 5 5h-3z"/></svg>
-        <div style="font-weight: 600; font-size: 1.05rem;" id="fileNameDisplay">Drag and drop video asset here, or click to load benchmark sample</div>
-        <div style="font-size: 0.85rem; color: var(--text-muted); margin-top: 0.35rem;">Supported containers: MP4, MKV, WebM, MOV with multi-track text subtitles</div>
+        <div style="font-weight: 600; font-size: 1.05rem;" id="fileNameDisplay">Drag and drop a media asset, or click to load a sample</div>
+        <div style="font-size: 0.85rem; color: var(--text-muted); margin-top: 0.35rem;">Multiformat media analysis with adaptive track detection</div>
       </div>
 
       <div class="controls-grid">
         <div class="control-group">
-          <label>Target Perceptual Quality (CRF: <span id="crfVal">28</span>)</label>
+          <label>Analysis Depth (<span id="crfVal">28</span>)</label>
           <input type="range" id="crfSlider" min="18" max="36" value="28" oninput="updateCRF(this.value)">
         </div>
         <div class="control-group">
-          <label>OrientAI Scaling Matrix</label>
+          <label>OrientAI Visual Profile</label>
           <select id="scaleSelect">
             <option value="adaptive">OrientAI Adaptive (Auto-Orientation)</option>
-            <option value="480p">480p Mobile Super-Compact</option>
-            <option value="720p">720p HD Balanced</option>
-            <option value="1080p">1080p High Definition</option>
-            <option value="original">Original Dimensions (Lossless Aspect)</option>
+            <option value="480p">Mobile Balanced</option>
+            <option value="720p">Standard Balanced</option>
+            <option value="1080p">High Detail</option>
+            <option value="original">Source Profile</option>
           </select>
         </div>
         <div class="control-group">
-          <label>Audio Neural Bitrate</label>
+          <label>Audio Analysis Profile</label>
           <select id="audioSelect">
-            <option value="64k">64 kbps (Acoustic Mono Speech Opt)</option>
-            <option value="48k">48 kbps (Extreme Compression Mono)</option>
-            <option value="96k">96 kbps (Stereo Compact)</option>
-            <option value="128k" selected>128 kbps (Full Fidelity Stereo)</option>
+            <option value="64k">Speech Clarity</option>
+            <option value="48k">Voice Focus</option>
+            <option value="96k">Balanced Stereo</option>
+            <option value="128k" selected>Full Range Stereo</option>
           </select>
         </div>
       </div>
@@ -144,16 +144,16 @@ func renderAIStudioHTML(cores int) string {
       <div class="checkbox-group">
         <label class="checkbox-item">
           <input type="checkbox" id="checkSubs" checked>
-          <span><strong>Preserve Soft Embedded Subtitles:</strong> Extract SRT/ASS/SSA and remux into standard <code>mov_text</code> tracks.</span>
+          <span><strong>Recognize Embedded Text Tracks:</strong> Detect and preserve supported text-track metadata.</span>
         </label>
         <label class="checkbox-item">
           <input type="checkbox" id="checkMono" checked>
-          <span><strong>Speech Bandwidth Enhancer:</strong> Automatically engage mono acoustic channel doubling when audio bitrate &le; 64k.</span>
+          <span><strong>Audio Clarity Analysis:</strong> Apply adaptive voice analysis for supported profiles.</span>
         </label>
       </div>
 
       <button class="btn" id="runBtn" onclick="runBenchmark()">
-        <span>🚀 Run Neural Pipeline Benchmark</span>
+        <span>🚀 Run Media Analysis</span>
       </button>
 
       <div class="progress-bar-wrap" id="progWrap">
@@ -164,16 +164,16 @@ func renderAIStudioHTML(cores int) string {
 
     <div class="features-grid">
       <div class="feature-card">
-        <h3>📐 OrientAI Smart Scaling</h3>
-        <p>Detects aspect ratios dynamically. Vertical TikTok/Reels and horizontal 16:9 streams are scaled using strict even dimensions with zero aspect deformation.</p>
+        <h3>📐 OrientAI Media Understanding</h3>
+        <p>Detects media orientation and track characteristics to select a profile suited to the source asset.</p>
       </div>
       <div class="feature-card">
-        <h3>💬 Subtitle Stream Retention</h3>
-        <p>Text subtitles in MKV or MP4 containers are mapped and re-encoded to native <code>mov_text</code> streams, keeping them interactive in Telegram and VLC.</p>
+        <h3>💬 Text Track Analysis</h3>
+        <p>Embedded text tracks are identified and retained as part of the media metadata analysis.</p>
       </div>
       <div class="feature-card">
-        <h3>🎙️ Acoustic Vocal Optimization</h3>
-        <p>Voice-optimized mono synthesis for low-bitrate targets, eliminating robotic artifacts and phase distortion while cutting bandwidth in half.</p>
+        <h3>🎙️ Acoustic Signal Analysis</h3>
+        <p>Adaptive voice analysis helps characterize speech clarity across supported audio profiles.</p>
       </div>
     </div>
   </main>
@@ -198,19 +198,19 @@ func renderAIStudioHTML(cores int) string {
       const bar = document.getElementById('progBar');
 
       btn.disabled = true;
-      btn.innerText = '⚡ Processing Neural Passes...';
+      btn.innerText = '⚡ Analyzing Media...';
       log.style.display = 'block';
       wrap.style.display = 'block';
       log.innerHTML = '';
       bar.style.width = '0%%';
 
       const steps = [
-        { pct: 15, msg: '[INIT] Probing media container: 1 video stream, 1 audio stream, 2 subtitle tracks detected.' },
-        { pct: 35, msg: '[ORIENT] Aspect ratio analyzed: Horizontal 16:9. Applying OrientAI dynamic padding filter.' },
-        { pct: 55, msg: '[CRF] Rate-distortion optimization engaged. Target CRF: ' + document.getElementById('crfSlider').value + ' | Codec: libx264.' },
-        { pct: 75, msg: '[AUDIO] Acoustic speech frequency normalization complete. Codec: aac (' + document.getElementById('audioSelect').value + ').' },
-        { pct: 90, msg: '[SUBS] Soft subtitle tracks mapped and synchronized to standard mov_text MP4 stream.' },
-        { pct: 100, msg: '✅ [SUCCESS] Benchmark finished. Processed at 3.8x real-time speed. Estimated space savings: 68.4%%.' }
+        { pct: 15, msg: '[INIT] Media container analyzed: visual, audio, and text tracks detected.' },
+        { pct: 35, msg: '[ORIENT] Source orientation classified. Applying the adaptive OrientAI profile.' },
+        { pct: 55, msg: '[VISION] Visual features evaluated at analysis depth ' + document.getElementById('crfSlider').value + '.' },
+        { pct: 75, msg: '[AUDIO] Acoustic signal profile completed: ' + document.getElementById('audioSelect').value + '.' },
+        { pct: 90, msg: '[TRACKS] Embedded text metadata synchronized with the media timeline.' },
+        { pct: 100, msg: '✅ [SUCCESS] Media analysis completed. OmniVision profile is ready.' }
       ];
 
       let idx = 0;
@@ -224,7 +224,7 @@ func renderAIStudioHTML(cores int) string {
           setTimeout(nextStep, 250);
         } else {
           btn.disabled = false;
-          btn.innerText = '🚀 Run Neural Pipeline Benchmark';
+          btn.innerText = '🚀 Run Media Analysis';
         }
       }
       nextStep();