all tabs have speed search capability

app.py

@@ -13,7 +13,7 @@ import re
 from collections import defaultdict
 from huggingface_hub import hf_hub_download
 import json        
-from huggingface_hub import hf_hub_download
+
 
 
 pattern = r'"(.*?)"'
@@ -144,8 +144,8 @@ def visualize_isnad(taraf_num, yaxis):
     isnad_hadith = isnad_hadith.merge(narrator_bios[['Rawi ID', 'Famous Name']], left_on='Destination', right_on='Rawi ID', how='left').rename(columns={'Famous Name': 'Student'})
     isnad_hadith[['Source', 'Destination']] = isnad_hadith[['Source', 'Destination']].astype(str)
     # Fill missing values with 'فلان'
-    isnad_hadith['Teacher'].fillna('فلان', inplace=True)
-    isnad_hadith['Student'].fillna('فلان', inplace=True)
+    # isnad_hadith['Teacher'].fillna('فلان', inplace=True)
+    # isnad_hadith['Student'].fillna('فلان', inplace=True)
 
     end_nodes = df['End Transmitter ID'].tolist() 
     G = nx.from_pandas_edgelist(isnad_hadith, source = 'Source', target = 'Destination', create_using = nx.DiGraph())
@@ -221,17 +221,15 @@ def visualize_subTaraf(taraf_num, hadith_str, yaxis):
     sub_taraf = taraf[taraf['Index'].isin(hadiths)]
 
     taraf_hadith = sub_taraf['bookid_hadithid'].to_list()
-    taraf_hadith_split = [i.split('_') for i in taraf_hadith]
-
-    hadith_cleaned = isnad_info['Tarafs Cleaned'].apply(lambda x: taraf_num in x)
-    isnad_hadith = isnad_info[hadith_cleaned]
-    subtaraf_filter = isnad_info['Hadiths Cleaned'].apply(lambda x: any(i in x for i in taraf_hadith_split))
-    isnad_hadith = isnad_hadith[subtaraf_filter]
+    isnad_hadith = isnad_info.iloc[np.array([hadith_lookup[i] for i in taraf_hadith]).reshape(-1)][['Source', 'Destination']]
+    isnad_hadith[['Source', 'Destination']] = isnad_hadith[['Source', 'Destination']].astype(int)
     
-    isnad_hadith['Teacher'] = isnad_hadith['Source'].apply(lambda x: narrator_bios[narrator_bios['Rawi ID'].astype(int) == int(x)]['Famous Name'].to_list())
-    isnad_hadith['Student'] = isnad_hadith['Destination'].apply(lambda x: narrator_bios[narrator_bios['Rawi ID'].astype(int) == int(x)]['Famous Name'].to_list())
-    isnad_hadith['Teacher'] = isnad_hadith['Teacher'].apply(lambda x: x[0] if len(x)==1 else 'فلان')
-    isnad_hadith['Student'] = isnad_hadith['Student'].apply(lambda x: x[0] if len(x)==1 else 'فلان')
+    # Merge isnad_hadith with narrator_bios for Teacher and Student
+    isnad_hadith = isnad_hadith.merge(narrator_bios[['Rawi ID', 'Famous Name']], left_on='Source', right_on='Rawi ID', how='left').rename(columns={'Famous Name': 'Teacher'})
+    isnad_hadith = isnad_hadith.merge(narrator_bios[['Rawi ID', 'Famous Name']], left_on='Destination', right_on='Rawi ID', how='left').rename(columns={'Famous Name': 'Student'})
+    isnad_hadith[['Source', 'Destination']] = isnad_hadith[['Source', 'Destination']].astype(str)
+    # isnad_hadith['Teacher'].fillna('فلان', inplace=True)
+    # isnad_hadith['Student'].fillna('فلان', inplace=True)
 
     G = nx.from_pandas_edgelist(isnad_hadith, source = 'Source', target = 'Destination', create_using = nx.DiGraph())
     isnad_pos = nx.nx_agraph.graphviz_layout(G, prog='dot')
@@ -267,13 +265,16 @@ def taraf_booknum(taraf_num):
 def visualize_hadith_isnad(df, yaxis):
     df['bookid_hadithid'] = df['Book_ID'].astype(str) + '_' + df['Hadith Number'].astype(str)
     hadith = matn_info[matn_info['bookid_hadithid'].isin(df['bookid_hadithid'])]
-    taraf_hadith_split = [i.split('_') for i in hadith['bookid_hadithid'].to_list()]
-    hadith_cleaned = isnad_info['Hadiths Cleaned'].apply(lambda x: any(i in x for i in taraf_hadith_split))
-    isnad_hadith = isnad_info[hadith_cleaned]
-    isnad_hadith['Teacher'] = isnad_hadith['Source'].apply(lambda x: narrator_bios[narrator_bios['Rawi ID'].astype(int) == int(x)]['Famous Name'].to_list())
-    isnad_hadith['Student'] = isnad_hadith['Destination'].apply(lambda x: narrator_bios[narrator_bios['Rawi ID'].astype(int) == int(x)]['Famous Name'].to_list())
-    isnad_hadith['Teacher'] = isnad_hadith['Teacher'].apply(lambda x: x[0] if len(x)==1 else 'فلان')
-    isnad_hadith['Student'] = isnad_hadith['Student'].apply(lambda x: x[0] if len(x)==1 else 'فلان')
+    taraf_hadith = df['bookid_hadithid'].to_list()
+    isnad_hadith = isnad_info.iloc[np.array([hadith_lookup[i] for i in taraf_hadith]).reshape(-1)][['Source', 'Destination']]
+    isnad_hadith[['Source', 'Destination']] = isnad_hadith[['Source', 'Destination']].astype(int)
+    
+    # Merge isnad_hadith with narrator_bios for Teacher and Student
+    isnad_hadith = isnad_hadith.merge(narrator_bios[['Rawi ID', 'Famous Name']], left_on='Source', right_on='Rawi ID', how='left').rename(columns={'Famous Name': 'Teacher'})
+    isnad_hadith = isnad_hadith.merge(narrator_bios[['Rawi ID', 'Famous Name']], left_on='Destination', right_on='Rawi ID', how='left').rename(columns={'Famous Name': 'Student'})
+    isnad_hadith[['Source', 'Destination']] = isnad_hadith[['Source', 'Destination']].astype(str)
+    # isnad_hadith['Teacher'].fillna('فلان', inplace=True)
+    # isnad_hadith['Student'].fillna('فلان', inplace=True)
 
     G = nx.from_pandas_edgelist(isnad_hadith, source = 'Source', target = 'Destination', create_using = nx.DiGraph())
     isnad_pos = nx.nx_agraph.graphviz_layout(G, prog='dot')
@@ -302,64 +303,61 @@ def visualize_hadith_isnad(df, yaxis):
 
 
 def visualize_narrator_taraf(taraf_num, narrator, yaxis):
-    taraf = matn_info[matn_info['taraf_ID'] == taraf_num]
-    num_hadith = taraf.shape[0]
-    taraf['Index'] = np.arange(num_hadith)
+    taraf = matn_info[matn_info['taraf_ID'] == taraf_num].copy()
+    taraf['Index'] = np.arange(len(taraf))
     hadith_cleaned = isnad_info['Tarafs Cleaned'].apply(lambda x: taraf_num in x)
     isnad_hadith = isnad_info[hadith_cleaned]
-    isnad_hadith['Teacher'] = isnad_hadith['Source'].apply(lambda x: narrator_bios[narrator_bios['Rawi ID'].astype(int) == int(x)]['Famous Name'].to_list())
-    isnad_hadith['Student'] = isnad_hadith['Destination'].apply(lambda x: narrator_bios[narrator_bios['Rawi ID'].astype(int) == int(x)]['Famous Name'].to_list())
-    isnad_hadith['Teacher'] = isnad_hadith['Teacher'].apply(lambda x: x[0] if len(x)==1 else 'فلان')
-    isnad_hadith['Student'] = isnad_hadith['Student'].apply(lambda x: x[0] if len(x)==1 else 'فلان')
+    isnad_hadith[['Source', 'Destination']] = isnad_hadith[['Source', 'Destination']].astype(int)
+    # Merge isnad_hadith with narrator_bios for Teacher and Student
+    isnad_hadith = isnad_hadith.merge(narrator_bios[['Rawi ID', 'Famous Name']], left_on='Source', right_on='Rawi ID', how='left').rename(columns={'Famous Name': 'Teacher'})
+    isnad_hadith = isnad_hadith.merge(narrator_bios[['Rawi ID', 'Famous Name']], left_on='Destination', right_on='Rawi ID', how='left').rename(columns={'Famous Name': 'Student'})
+    isnad_hadith[['Source', 'Destination']] = isnad_hadith[['Source', 'Destination']].astype(str)
     taraf_hadith = taraf['bookid_hadithid'].to_list()
-    taraf_hadith_split = [i.split('_') for i in taraf_hadith]
 
+    # original graph of whole taraf
     G = nx.from_pandas_edgelist(isnad_hadith, source = 'Source', target = 'Destination', create_using = nx.DiGraph())
     if narrator not in G.nodes():
         raise gr.Error('Narrator not in Isnad of Taraf!')
-    # nodes = list(G.nodes)
-    # nodes = [node for node in nodes if node not in list(nx.ancestors(G, narrator)) + list(nx.descendants(G, narrator)) and node !=narrator]
-    # [G.remove_nodes_from(nodes)]
+
     matns_with_narrator = []
     end_node = {}
-    for i in range(len(taraf_hadith_split)):
-        isnad_in_hadith1 = isnad_hadith['Hadiths Cleaned'].apply(lambda x: taraf_hadith_split[i] in x )
-        isnad_hadith1 = isnad_hadith[isnad_in_hadith1]
-        G1 = nx.from_pandas_edgelist(isnad_hadith1, source = 'Source', target = 'Destination', create_using = nx.DiGraph())
-        node = [n for n, d in G1.out_degree() if d == 0]
-        if narrator in list(G1.nodes):
-            matns_with_narrator.append(taraf_hadith[i])
-            for n in node:
-                if n not in list(end_node.keys()):
-                    end_node[n] = []
-                end_node[n].append(str(i))
-
-    matns_with_narrator_split = [i.split('_') for i in matns_with_narrator]
-    hadith_cleaned = isnad_hadith['Hadiths Cleaned'].apply(lambda x: any(i in x for i in matns_with_narrator_split))
+
+    # Process each hadith in taraf_hadith_split
+    for idx, split_hadith in enumerate(taraf_hadith):
+        isnad_hadith1 = isnad_info.iloc[hadith_lookup[taraf_hadith[i]]][['Source', 'Destination']]
+        G1 = nx.from_pandas_edgelist(isnad_hadith1, source='Source', target='Destination', create_using=nx.DiGraph())
+        if narrator in G1.nodes:
+            matns_with_narrator.append(taraf_hadith[idx])
+            for node in (n for n, d in G1.out_degree() if d == 0):
+                end_node.setdefault(node, []).append(str(idx))
+                
+    # Update the graph
+    hadith_cleaned = isnad_info.iloc[np.array([hadith_lookup[i] for i in matns_with_narrator]).reshape(-1)][['Source', 'Destination']]
     isnad_hadith = isnad_hadith[hadith_cleaned]
     G = nx.from_pandas_edgelist(isnad_hadith, source = 'Source', target = 'Destination', create_using = nx.DiGraph())
     isnad_pos = nx.nx_agraph.graphviz_layout(G, prog='dot')
 
     narrator_matn_info = taraf[taraf['bookid_hadithid'].isin(matns_with_narrator)]
-    num_hadith = narrator_matn_info.shape[0]
-    narrator_matn_info['Subset Index'] = np.arange(num_hadith)
+    narrator_matn_info['Subset Index'] = np.arange(len(narrator_matn_info))
+
+    # Visualization with pyvis
     x_stretch = 4
     y_stretch = 4
     net = Network(directed =True, select_menu=True, cdn_resources='remote')
 
     for node, pos in isnad_pos.items():
-        node_info,student_narrations,student_gen, student_rank, node_name = get_node_info(node)
-        if node == '99999':
-            net.add_node(node, font = {'size':50, 'color': 'black'}, color = '#000000', label = f'{node_name} \n ID: {node} - Gen {student_gen}', x= pos[0]*x_stretch, y= -1*pos[1]*y_stretch, size= 70)   
-        elif node in list(end_node.keys()):
-            net.add_node(node, font = {'size':30, 'color': 'red'}, color = value_to_hex(student_narrations), label = f'{node_name} \n {student_rank} \n ID: {node} - Gen {student_gen} \n Hadiths {", ".join(end_node[node])}', x= pos[0]*x_stretch, y= -1*pos[1]*y_stretch, size= 50)
-        else:
-            net.add_node(node, font = {'size':30, 'color': 'red'}, color = value_to_hex(student_narrations), label = f'{node_name} \n {student_rank} \n ID: {node} - Gen {student_gen}', x= pos[0]*x_stretch, y= -1*pos[1]*y_stretch, size= 50)
+        node_info, student_narrations, student_gen, student_rank, node_name = get_node_info(node)
+        label = f'{node_name} \n ID: {node} - Gen {student_gen}'
+        size = 70 if node == '99999' else 50
+        font_color = 'black' if node == '99999' else 'red'
+        hadiths = f" \n Hadiths {', '.join(end_node[node])}" if node in end_node else ''
+        net.add_node(node, font={'size': 30, 'color': font_color}, color=value_to_hex(student_narrations),
+                     label=f"{label} {hadiths}", x=pos[0] * x_stretch, y=-pos[1] * y_stretch, size=size)
+
     for edge in G.edges:
         row = isnad_hadith[(isnad_hadith['Source'] == edge[0]) & (isnad_hadith['Destination'] == edge[1])].iloc[0]
-        source = row['Source']
-        target = row['Destination']
-        net.add_edge(source, target, color = value_to_hex(int(row[f'{yaxis} Count'])), label = f"{row[f'{yaxis} Count']}")
+        net.add_edge(edge[0], edge[1], color=value_to_hex(int(row[f'{yaxis} Count'])), label=f"{row[f'{yaxis} Count']}")
+        
     net.toggle_physics(False)    
     html = net.generate_html()
     html = html.replace("'", "\"")