From 8b23384840b5eb18b3e35a9dc4596ad456a03d8e Mon Sep 17 00:00:00 2001
From: Luis Nassif <lfcnassif@gmail.com>
Date: Sun, 28 Apr 2024 16:35:29 -0300
Subject: [PATCH] '#1823: redirect warmless console messages to log

---
 .../resources/scripts/tasks/WhisperProcess.py |  4 +-
 .../transcript/Wav2Vec2TranscriptTask.java    |  2 +-
 .../transcript/WhisperTranscriptTask.java     | 40 +++++++++++++++++++
 3 files changed, 43 insertions(+), 3 deletions(-)

diff --git a/iped-app/resources/scripts/tasks/WhisperProcess.py b/iped-app/resources/scripts/tasks/WhisperProcess.py
index f0f68c5ffe..8e2b6456b1 100644
--- a/iped-app/resources/scripts/tasks/WhisperProcess.py
+++ b/iped-app/resources/scripts/tasks/WhisperProcess.py
@@ -37,7 +37,7 @@ def main():
         deviceNum = 0
     
     try:
-        model = whisperx.load_model(modelName, device=deviceId, device_index=deviceNum, threads=threads, compute_type=compute_type)
+        model = whisperx.load_model(modelName, device=deviceId, device_index=deviceNum, threads=threads, compute_type=compute_type, language=language)
 
     except Exception as e:
         if deviceId != 'cpu':
@@ -45,7 +45,7 @@ def main():
             print('FAILED to load model on GPU, fallbacking to CPU!', file=sys.stderr)
             deviceId = 'cpu'
             compute_type = 'int8'
-            model = whisperx.load_model(modelName, device=deviceId, device_index=deviceNum, threads=threads, compute_type=compute_type)
+            model = whisperx.load_model(modelName, device=deviceId, device_index=deviceNum, threads=threads, compute_type=compute_type, language=language)
         else:
             raise e
     
diff --git a/iped-engine/src/main/java/iped/engine/task/transcript/Wav2Vec2TranscriptTask.java b/iped-engine/src/main/java/iped/engine/task/transcript/Wav2Vec2TranscriptTask.java
index 8cd05ae366..84a92dca8a 100644
--- a/iped-engine/src/main/java/iped/engine/task/transcript/Wav2Vec2TranscriptTask.java
+++ b/iped-engine/src/main/java/iped/engine/task/transcript/Wav2Vec2TranscriptTask.java
@@ -40,7 +40,7 @@ public class Wav2Vec2TranscriptTask extends AbstractTranscriptTask {
 
     private static LinkedBlockingDeque<Server> deque = new LinkedBlockingDeque<>();
 
-    private static volatile Level logLevel = Level.forName("MSG", 250);
+    protected static volatile Level logLevel = Level.forName("MSG", 250);
 
     static class Server {
         Process process;
diff --git a/iped-engine/src/main/java/iped/engine/task/transcript/WhisperTranscriptTask.java b/iped-engine/src/main/java/iped/engine/task/transcript/WhisperTranscriptTask.java
index d4d05744d0..6a77455cca 100644
--- a/iped-engine/src/main/java/iped/engine/task/transcript/WhisperTranscriptTask.java
+++ b/iped-engine/src/main/java/iped/engine/task/transcript/WhisperTranscriptTask.java
@@ -3,7 +3,10 @@
 import java.io.BufferedReader;
 import java.io.File;
 import java.io.IOException;
+import java.io.InputStream;
 import java.io.InputStreamReader;
+import java.util.Arrays;
+import java.util.List;
 
 import org.apache.commons.lang3.SystemUtils;
 import org.apache.logging.log4j.LogManager;
@@ -96,4 +99,41 @@ protected TextAndScore transcribeAudio(File tmpFile) throws Exception {
         return transcribeWavPart(tmpFile);
     }
 
+    @Override
+    protected void logInputStream(InputStream is) {
+        List<String> ignoreMsgs = Arrays.asList(
+                "With dispatcher enabled, this function is no-op. You can remove the function call.",
+                "torchvision is not available - cannot save figures",
+                "Lightning automatically upgraded your loaded checkpoint from",
+                "Model was trained with pyannote.audio 0.0.1, yours is",
+                "Model was trained with torch 1.10.0+cu102, yours is");
+        Thread t = new Thread() {
+            public void run() {
+                byte[] buf = new byte[1024];
+                int read = 0;
+                try {
+                    while ((read = is.read(buf)) != -1) {
+                        String msg = new String(buf, 0, read).trim();
+                        boolean ignore = false;
+                        for (String i : ignoreMsgs) {
+                            if (msg.contains(i)) {
+                                ignore = true;
+                                break;
+                            }
+                        }
+                        if (ignore) {
+                            logger.warn(msg);
+                        } else {
+                            logger.log(logLevel, msg);
+                        }
+                    }
+                } catch (IOException e) {
+                    e.printStackTrace();
+                }
+            }
+        };
+        t.setDaemon(true);
+        t.start();
+    }
+
 }