Update app.py

app.py

@@ -1,8 +1,8 @@
-import subprocess
+import easyocr
+import numpy as np
 
-# debug line: will show up in your Container logs
-res = subprocess.run(["which", "tesseract"], capture_output=True, text=True)
-print("🔍 TESSERACT PATH:", repr(res.stdout.strip()), "— stderr:", repr(res.stderr.strip()))
+# create a reader once
+ocr_reader = easyocr.Reader(["en"], gpu=False)
 import gradio as gr
 import torch
 from transformers import pipeline as hf_pipeline, AutoModelForSequenceClassification, AutoTokenizer
@@ -204,32 +204,22 @@ def analyze_composite(uploaded_file, *texts):
 
     # 1) File upload
     if uploaded_file is not None:
-        # uploaded_file may be a file-like with .read(), or just a path string
+    # ... read `raw` exactly as you have it now ...
+
+    name = (uploaded_file.name if hasattr(uploaded_file, "name") else uploaded_file).lower()
+    if name.endswith((".png",".jpg",".jpeg",".tiff",".bmp",".gif")):
+        img = Image.open(io.BytesIO(raw))
+        arr = np.array(img.convert("RGB"))
+        texts = ocr_reader.readtext(arr, detail=0)
+        content = "\n".join(texts)
+    else:
         try:
-            raw = uploaded_file.read()
-        except Exception:
-            # fall back to treating uploaded_file as a filesystem path
-            with open(uploaded_file, "rb") as f:
-                raw = f.read()
-
-        # get the filename (or just use the string if no .name attr)
-        name = (
-            uploaded_file.name.lower()
-            if hasattr(uploaded_file, "name")
-            else uploaded_file.lower()
-        )
+            content = raw.decode("utf-8")
+        except:
+            content = raw.decode("latin-1")
 
-        # now branch on extension
-        if name.endswith((".png", ".jpg", ".jpeg", ".tiff", ".bmp", ".gif")):
-            img = Image.open(io.BytesIO(raw))
-            content = pytesseract.image_to_string(img)
-        else:
-            try:
-                content = raw.decode("utf-8")
-            except UnicodeDecodeError:
-                content = raw.decode("latin-1")
-
-        r = analyze_message(content)
+    r = analyze_message(content)
+    outputs.append(…)
         outputs.append(
             "── Uploaded File ──\n"
             f"Emotion Profile : {r['emotion_profile']}\n"