Update app.py

app.py

@@ -111,27 +111,11 @@ def split_text(text, max_tokens=500):
     return chunks
 
 def create_bibtex_entry(data):
-    author = data.get('Author', '').strip()
-    title = data.get('Title', '').strip()
-    journal = data.get('Journal', '').strip()
-    year = data.get('Year', '').strip()
-    volume = data.get('Volume', '').strip()
-    pages = data.get('Pages', '').strip()
-    doi = data.get('Doi', '').strip()
-
-    # Remove "doi: " prefix if present
-    doi = doi.replace('doi: ', '')
-
     bibtex = "@article{idnothing,\n"
-    if author: bibtex += f"  author = {{{author}}},\n"
-    if title: bibtex += f"  title = {{{title}}},\n"
-    if journal: bibtex += f"  journal = {{{journal}}},\n"
-    if year: bibtex += f"  year = {{{year}}},\n"
-    if volume: bibtex += f"  volume = {{{volume}}},\n"
-    if pages: bibtex += f"  pages = {{{pages}}},\n"
-    if doi: bibtex += f"  doi = {{{doi}}},\n"
-    bibtex += "}"
-
+    for key, value in data.items():
+        if key != 'None' and value.strip():
+            bibtex += f"  {key.lower()} = {{{value.strip()}}},\n"
+    bibtex = bibtex.rstrip(',\n') + "\n}"
     return bibtex
 
 
@@ -151,13 +135,13 @@ def transform_chunks(marianne_segmentation):
         result_entity = "[" + entity_group.capitalize() + "]"
         word = row['word']
         
-        if entity_group in ['Author', 'Title', 'Journal', 'Pages', 'Doi']:
+        if entity_group != 'None':
             if entity_group in bibtex_data:
                 bibtex_data[entity_group] += ' ' + word
             else:
                 bibtex_data[entity_group] = word
             current_entity = entity_group
-        elif entity_group == 'None':
+        else:
             if current_entity:
                 bibtex_data[current_entity] += ' ' + word
             else:
@@ -165,21 +149,11 @@ def transform_chunks(marianne_segmentation):
         
         html_output.append(f'<div class="manuscript"><div class="annotation">{result_entity}</div><div class="content">{word}</div></div>')
 
-    # Extract year from the 'None' field if present
-    none_content = bibtex_data.get('None', '')
-    year_match = re.search(r'\((\d{4})\)', none_content)
-    if year_match:
-        bibtex_data['Year'] = year_match.group(1)
-
-    # Extract volume from the 'None' field if present
-    volume_match = re.search(r',\s*(\d+),', none_content)
-    if volume_match:
-        bibtex_data['Volume'] = volume_match.group(1)
-
     bibtex_entry = create_bibtex_entry(bibtex_data)
     
     final_html = '\n'.join(html_output)
     return final_html, bibtex_entry
+    
 
 
 # Class to encapsulate the Falcon chatbot
@@ -203,7 +177,19 @@ class MistralChatBot:
             classified_list.append(df)
     
         classified_list = pd.concat(classified_list)
+        
+        # Debugging: Print the classified list
+        print("Classified List:")
+        print(classified_list)
+        
         html_output, bibtex_entry = transform_chunks(classified_list)
+        
+        # Debugging: Print the outputs
+        print("HTML Output:")
+        print(html_output)
+        print("BibTeX Entry:")
+        print(bibtex_entry)
+        
         generated_text = f'{css}<h2 style="text-align:center">Edited text</h2>\n<div class="generation">{html_output}</div>'
         return generated_text, bibtex_entry
 
@@ -224,7 +210,7 @@ demo = gr.Blocks()
 
 with gr.Blocks(theme='JohnSmith9982/small_and_pretty') as demo:
     gr.HTML("""<h1 style="text-align:center">Reversed Zotero</h1>""")
-    text_input = gr.Textbox(label="Your text", type="text", lines=1)
+    text_input = gr.Textbox(label="Your text", type="text", lines=5)
     text_button = gr.Button("Extract a structured bibtex")
     text_output = gr.HTML(label="Metadata")
     bibtex_output = gr.Textbox(label="BibTeX Entry", lines=10)