Update src/main.py

src/main.py

@@ -25,104 +25,61 @@ def spell_out_word(word):
     """단어를 개별 알파벳으로 분리하는 함수"""
     return ' '.join(list(word.lower()))
 
-def translate_korean_to_english(text):
+def translate_quoted_word(word):
+    """따옴표 안의 단어를 개별적으로 번역"""
     try:
-        # 작은따옴표로 묶인 단어들 찾기
-        quoted_words = find_quoted_words(text)
-        translated_quoted_words = []
-        
-        # 각 따옴표 단어를 먼저 번역
         url = "https://translate.googleapis.com/translate_a/single"
-        for word in quoted_words:
-            params = {
-                "client": "gtx",
-                "sl": "ko",
-                "tl": "en",
-                "dt": "t",
-                "q": word
-            }
-            response = requests.get(url, params=params)
-            if response.status_code == 200:
-                translated_word = response.json()[0][0][0].upper()
-                translated_quoted_words.append(translated_word)
-            else:
-                translated_quoted_words.append(word)
-
-        # 원본 텍스트에서 따옴표 부분을 임시 마커로 대체
-        temp_text = text
-        for i, word in enumerate(quoted_words):
-            temp_text = temp_text.replace(f"'{word}'", f"QUOTED_WORD_{i}")
-
-        # 전체 문장 번역
         params = {
             "client": "gtx",
             "sl": "ko",
             "tl": "en",
             "dt": "t",
-            "q": temp_text.strip()
+            "q": word
         }
         response = requests.get(url, params=params)
-        
         if response.status_code == 200:
-            translated_text = ' '.join(item[0] for item in response.json()[0] if item[0])
-            
-            # 번역된 텍스트에서 마커를 번역된 따옴표 단어로 대체
-            for i, translated_word in enumerate(translated_quoted_words):
-                translated_text = translated_text.replace(f"QUOTED_WORD_{i}", f"'{translated_word}'")
-            
-            return translated_text
-        else:
-            raise Exception(f"Translation API returned status code: {response.status_code}")
+            translated = response.json()[0][0][0].upper()
+            return translated
+        return word
     except Exception as e:
-        print(f"Translation error: {e}")
-        return text
+        print(f"Word translation error: {e}")
+        return word
 
-def generate_complete_video(gloss_list, dataset, list_2000_tokens):
+def translate_korean_to_english(text):
+    """전체 텍스트 번역 함수"""
     try:
-        frames = []
-        is_spelling = False
-        
-        for gloss in gloss_list:
-            if gloss == 'FINGERSPELL-START':
-                is_spelling = True
-                continue
-            elif gloss == 'FINGERSPELL-END':
-                is_spelling = False
-                continue
-                
-            for frame in dg.generate_video([gloss], dataset, list_2000_tokens):
-                frame_data = frame.split(b'\r\n\r\n')[1]
-                nparr = np.frombuffer(frame_data, np.uint8)
-                img = cv2.imdecode(nparr, cv2.IMREAD_COLOR)
-                frames.append(img)
+        # 1. 따옴표로 묶인 부분을 찾아서 따로 번역
+        pattern = r"'([^']*)'|([^']+)"
+        parts = re.findall(pattern, text)
+        translated_parts = []
         
-        if not frames:
-            raise Exception("No frames generated")
-
-        height, width = frames[0].shape[:2]
-        fourcc = cv2.VideoWriter_fourcc(*'mp4v')
-        
-        with tempfile.NamedTemporaryFile(suffix='.mp4', delete=False) as temp_file:
-            temp_path = temp_file.name
-            
-        out = cv2.VideoWriter(temp_path, fourcc, 25, (width, height))
-        
-        for frame in frames:
-            out.write(frame)
-        out.release()
-        
-        with open(temp_path, 'rb') as f:
-            video_bytes = f.read()
+        for quoted, unquoted in parts:
+            if quoted:  # 따옴표로 묶인 부분
+                translated_word = translate_quoted_word(quoted)
+                translated_parts.append(f"'{translated_word}'")
+            elif unquoted:  # 일반 텍스트
+                # 일반 텍스트 번역
+                url = "https://translate.googleapis.com/translate_a/single"
+                params = {
+                    "client": "gtx",
+                    "sl": "ko",
+                    "tl": "en",
+                    "dt": "t",
+                    "q": unquoted.strip()
+                }
+                response = requests.get(url, params=params)
+                if response.status_code == 200:
+                    translated = ' '.join(item[0] for item in response.json()[0] if item[0])
+                    translated_parts.append(translated)
+                else:
+                    translated_parts.append(unquoted)
         
-        os.remove(temp_path)
-        return video_bytes
+        # 번역된 부분들을 합치기
+        result = ''.join(translated_parts).strip()
+        return result
     except Exception as e:
-        print(f"Error generating video: {str(e)}")
-        raise
-
-@app.route('/')
-def index():
-    return render_template('index.html', title=app.config['TITLE'])
+        print(f"Translation error: {e}")
+        return text
 
 @app.route('/translate/', methods=['POST'])
 def result():
@@ -132,14 +89,17 @@ def result():
             return render_template('error.html', error="Please enter text to translate")
         
         try:
+            # 번역 수행
             english_text = translate_korean_to_english(input_text)
             if not english_text:
                 raise Exception("Translation failed")
-                
-            # 작은따옴표로 묶인 단어들 찾기
-            quoted_words = find_quoted_words(english_text)
             
-            eng_to_asl_translator = NlpSpacyBaseTranslator(sentence=english_text.replace("'", ""))
+            # 따옴표로 묶인 단어 추출 (번역된 영어 텍스트에서)
+            quoted_words = re.findall(r"'([^']*)'", english_text)
+            
+            # 번역된 텍스트에서 따옴표 제거하고 ASL 변환
+            clean_english = re.sub(r"'([^']*)'", r"\1", english_text)
+            eng_to_asl_translator = NlpSpacyBaseTranslator(sentence=clean_english)
             generated_gloss = eng_to_asl_translator.translate_to_gloss()
             
             # 단어 처리
@@ -148,7 +108,7 @@ def result():
             
             for word in words:
                 word_upper = word.upper()
-                if any(quoted_word.upper() == word_upper for quoted_word in quoted_words):
+                if any(quoted.upper() == word_upper for quoted in quoted_words):
                     # 고유명사인 경우 철자를 하나씩 분리
                     spelled_word = spell_out_word(word)
                     processed_gloss.extend(['FINGERSPELL-START'] + spelled_word.split() + ['FINGERSPELL-END'])
@@ -190,6 +150,55 @@ def result():
         except Exception as e:
             return render_template('error.html', error=f"Translation error: {str(e)}")
 
+def generate_complete_video(gloss_list, dataset, list_2000_tokens):
+    try:
+        frames = []
+        is_spelling = False
+        
+        for gloss in gloss_list:
+            if gloss == 'FINGERSPELL-START':
+                is_spelling = True
+                continue
+            elif gloss == 'FINGERSPELL-END':
+                is_spelling = False
+                continue
+                
+            for frame in dg.generate_video([gloss], dataset, list_2000_tokens):
+                frame_data = frame.split(b'\r\n\r\n')[1]
+                nparr = np.frombuffer(frame_data, np.uint8)
+                img = cv2.imdecode(nparr, cv2.IMREAD_COLOR)
+                frames.append(img)
+        
+        if not frames:
+            raise Exception("No frames generated")
+
+        height, width = frames[0].shape[:2]
+        fourcc = cv2.VideoWriter_fourcc(*'mp4v')
+        
+        with tempfile.NamedTemporaryFile(suffix='.mp4', delete=False) as temp_file:
+            temp_path = temp_file.name
+            
+        out = cv2.VideoWriter(temp_path, fourcc, 25, (width, height))
+        
+        for frame in frames:
+            out.write(frame)
+        out.release()
+        
+        with open(temp_path, 'rb') as f:
+            video_bytes = f.read()
+        
+        os.remove(temp_path)
+        return video_bytes
+    except Exception as e:
+        print(f"Error generating video: {str(e)}")
+        raise
+
+@app.route('/')
+def index():
+    return render_template('index.html', title=app.config['TITLE'])
+
+
+
 @app.route('/video_feed')
 def video_feed():
     sentence = request.args.get('gloss_sentence_to_display', '')