Upload app.py

app.py

@@ -1,48 +1,82 @@
-from fastapi import FastAPI, HTTPException
-from pydantic import BaseModel
-from transformers import pipeline
-import torch
-import os
-
-# Version: 2026-01-15 - Updated model
-app = FastAPI(title="Ibani Translator API")
-
-# Model configuration
-MODEL_ID = "williampepple1/ibani-translator"
-
-print(f"Loading model {MODEL_ID}...")
-try:
-    # Use pipeline for easy inference
-    # device=-1 forces CPU usage which is what HF Spaces free tier provides
-    # force_download=True ensures we get the latest model version
-    translator = pipeline(
-        "translation", 
-        model=MODEL_ID, 
-        device=-1,
-        model_kwargs={"force_download": True}
-    )
-    print("Model loaded successfully!")
-except Exception as e:
-    print(f"Error loading model: {e}")
-    translator = None
-
-class TranslationRequest(BaseModel):
-    text: str
-
-@app.get("/")
-def read_root():
-    return {"status": "healthy", "model": MODEL_ID}
-
-@app.post("/translate")
-async def translate(request: TranslationRequest):
-    if translator is None:
-        raise HTTPException(status_code=503, detail="Model not loaded")
-    
-    try:
-        result = translator(request.text)
-        return {
-            "translated_text": result[0]['translation_text'],
-            "original_text": request.text
-        }
-    except Exception as e:
-        raise HTTPException(status_code=500, detail=str(e))
+from fastapi import FastAPI, HTTPException
+from pydantic import BaseModel
+from transformers import pipeline
+import torch
+import os
+import re
+
+
+def fix_ibani_spacing(text: str) -> str:
+    """Remove unwanted spaces around special Ibani characters (á, ḅ)."""
+    if not text:
+        return text
+    
+    # All Ibani letters (including special characters) that can appear in words
+    ibani_letters = (
+        r'a-zA-Z'           # Basic ASCII letters
+        r'ạẹịọụ'            # Vowels with dot below (lowercase)
+        r'ẠẸỊỌỤ'            # Vowels with dot below (uppercase)
+        r'áéíóú'            # Vowels with acute (lowercase)
+        r'ÁÉÍÓÚ'            # Vowels with acute (uppercase)
+        r'àèìòù'            # Vowels with grave (lowercase)
+        r'ÀÈÌÒÙ'            # Vowels with grave (uppercase)
+        r'ḅ'                # B with dot below (lowercase)
+        r'Ḅ'                # B with dot below (uppercase)
+        r'ńṅ'               # N with diacritics (lowercase)
+        r'ŃṄ'               # N with diacritics (uppercase)
+    )
+    
+    # Characters that cause spacing issues
+    special_chars = ['á', 'Á', 'ḅ', 'Ḅ']
+    
+    for char in special_chars:
+        # Remove space before the character (when preceded by any Ibani letter)
+        text = re.sub(r'([' + ibani_letters + r'])\s+' + re.escape(char), r'\1' + char, text)
+        # Remove space after the character (when followed by any Ibani letter)
+        text = re.sub(re.escape(char) + r'\s+([' + ibani_letters + r'])', char + r'\1', text)
+    
+    return text
+
+# Version: 2026-01-15 - Updated model
+app = FastAPI(title="Ibani Translator API")
+
+# Model configuration
+MODEL_ID = "williampepple1/ibani-translator"
+
+print(f"Loading model {MODEL_ID}...")
+try:
+    # Use pipeline for easy inference
+    # device=-1 forces CPU usage which is what HF Spaces free tier provides
+    # force_download=True ensures we get the latest model version
+    translator = pipeline(
+        "translation", 
+        model=MODEL_ID, 
+        device=-1,
+        model_kwargs={"force_download": True}
+    )
+    print("Model loaded successfully!")
+except Exception as e:
+    print(f"Error loading model: {e}")
+    translator = None
+
+class TranslationRequest(BaseModel):
+    text: str
+
+@app.get("/")
+def read_root():
+    return {"status": "healthy", "model": MODEL_ID}
+
+@app.post("/translate")
+async def translate(request: TranslationRequest):
+    if translator is None:
+        raise HTTPException(status_code=503, detail="Model not loaded")
+    
+    try:
+        result = translator(request.text)
+        translated_text = fix_ibani_spacing(result[0]['translation_text'])
+        return {
+            "translated_text": translated_text,
+            "original_text": request.text
+        }
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=str(e))