bug fix extractor, results

app.py

@@ -106,7 +106,7 @@ class LongAudioProcessor:
 class TaggedPhraseExtractor:
     def __init__(self, text=''):
         self.text = text
-        self.patterns = {}
+        self.patterns = patterns 
 
     def set_text(self, text):
         """Set the text to search within."""
@@ -119,7 +119,6 @@ class TaggedPhraseExtractor:
     def extract_phrases(self):
         """Extract phrases for all colors and patterns added, including the three longest phrases."""
         matches = {}
-        three_matches = {}
         for color, pattern in self.patterns.items():
             found_phrases = re.findall(pattern, self.text)
             sorted_phrases = sorted(found_phrases, key=len, reverse=True)
@@ -244,27 +243,32 @@ def process_input(image=None, file=None, audio=None, text="", translateto = "Eng
 
     audio_output = process_text_to_audio(processed_text, translateto, translateto)
     extractor = TaggedPhraseExtractor(final_text)
-    extractor.add_pattern("red", patterns["red"])
-    extractor.add_pattern("blue", patterns["blue"])
-    extractor.add_pattern("green", patterns["green"])
     matches = extractor.extract_phrases()
 
     top_phrases = []
     for color, phrases in matches.items():
         top_phrases.extend(phrases)
 
+    while len(top_phrases) < 3:
+        top_phrases.append("")
+
     audio_outputs = []
     translations = []
     for phrase in top_phrases:
-        translated_phrase = translate_text(phrase, translatefrom=translatefrom, translateto=translateto)
-        translations.append(translated_phrase)
-        target_audio = process_text_to_audio(phrase, translatefrom=translateto, translateto=translateto)
-        native_audio = process_text_to_audio(translated_phrase, translatefrom=translatefrom, translateto=translatefrom)
-        audio_outputs.append((target_audio, native_audio))
+        if phrase:
+            translated_phrase = translate_text(phrase, translatefrom=translatefrom, translateto=translateto)
+            translations.append(translated_phrase)
+            target_audio = process_text_to_audio(phrase, translatefrom=translateto, translateto=translateto)
+            native_audio = process_text_to_audio(translated_phrase, translatefrom=translatefrom, translateto=translatefrom)
+            audio_outputs.append((target_audio, native_audio))
+        else:
+            translations.append("")
+            audio_outputs.append(("", ""))
 
     return final_text, audio_output, top_phrases, translations, audio_outputs
 
 
+
 # Define the inputs and outputs for the Gradio Interface
 inputs = [
     
@@ -292,28 +296,20 @@ outputs = [
     gr.Audio(label="Audio Output (Native Language) 3"),
     gr.Audio(label="Audio Output (Target Language) 3")
 ]
+
 def update_outputs(inputlanguage, target_language, audio, image, text, file):
-    final_text, audio_output, top_phrases, translations, audio_outputs = process_input(
+    processed_text, audio_output, top_phrases, translations, audio_outputs = process_input(
         image=image, file=file, audio=audio, text=text, 
         translateto=target_language, translatefrom=inputlanguage 
     )
-    processed_text_output = final_text
-    audio_output_native_phrases = [native for _, native in audio_outputs]
-    audio_output_target_phrases = [target for target, _ in audio_outputs]
-
-    longest_phrases_outputs = top_phrases[:3]
-    translated_phrases_outputs = translations[:3]
-    audio_outputs_native = audio_output_native_phrases[:3]
-    audio_outputs_target = audio_output_target_phrases[:3]
-
-    return (
-        processed_text_output,
-        audio_output,  
-        *audio_outputs_native, 
-        *audio_outputs_target, 
-        *longest_phrases_outputs, 
-        *translated_phrases_outputs
-    )
+    output_tuple = (processed_text, audio_output)
+    for i in range(len(top_phrases)):
+        output_tuple += (top_phrases[i], translations[i], audio_outputs[i][0], audio_outputs[i][1])
+
+    while len(output_tuple) < 14: 
+        output_tuple += ("", "", "", "")
+
+    return output_tuple
 
 def interface_func(inputlanguage, target_language, audio, image, text, file):
     return update_outputs(inputlanguage, target_language, audio, image, text, file)