Update modules/text_analysis/discourse_analysis.py

modules/text_analysis/discourse_analysis.py

@@ -3,52 +3,50 @@ import spacy
 import networkx as nx
 import matplotlib.pyplot as plt
 from collections import defaultdict
-from .semantic_analysis import visualize_semantic_relations, create_semantic_graph, POS_COLORS, POS_TRANSLATIONS
+from .semantic_analysis import (
+    create_concept_graph,
+    visualize_concept_graph,
+    identify_and_contextualize_entities,
+    POS_COLORS,
+    POS_TRANSLATIONS,
+    ENTITY_LABELS
+)
 
-##################################################################################################################
 def compare_semantic_analysis(text1, text2, nlp, lang):
     doc1 = nlp(text1)
     doc2 = nlp(text2)
-    
-    G1, pos_counts1 = create_semantic_graph(doc1, lang)
-    G2, pos_counts2 = create_semantic_graph(doc2, lang)
-    
-    # Create two separate figures with a smaller size
-    fig1, ax1 = plt.subplots(figsize=(18, 13))
-    fig2, ax2 = plt.subplots(figsize=(18, 13))
-    
-    # Draw the first graph
-    pos1 = nx.spring_layout(G1, k=0.7, iterations=50)
-    nx.draw(G1, pos1, ax=ax1, node_color=[POS_COLORS.get(G1.nodes[node]['pos'], '#CCCCCC') for node in G1.nodes()],
-            with_labels=True, node_size=4000, font_size=10, font_weight='bold',
-            arrows=True, arrowsize=20, width=2, edge_color='gray')
-    nx.draw_networkx_edge_labels(G1, pos1, edge_labels=nx.get_edge_attributes(G1, 'label'), font_size=8, ax=ax1)
-    
-    # Draw the second graph
-    pos2 = nx.spring_layout(G2, k=0.7, iterations=50)
-    nx.draw(G2, pos2, ax=ax2, node_color=[POS_COLORS.get(G2.nodes[node]['pos'], '#CCCCCC') for node in G2.nodes()],
-            with_labels=True, node_size=4000, font_size=10, font_weight='bold',
-            arrows=True, arrowsize=20, width=2, edge_color='gray')
-    nx.draw_networkx_edge_labels(G2, pos2, edge_labels=nx.get_edge_attributes(G2, 'label'), font_size=8, ax=ax2)
-    
-    ax1.set_title("Documento 1: Relaciones Semánticas Relevantes", fontsize=14, fontweight='bold')
-    ax2.set_title("Documento 2: Relaciones Semánticas Relevantes", fontsize=14, fontweight='bold')
-    
-    ax1.axis('off')
-    ax2.axis('off')
-    
-    # Add legends
-    legend_elements = [plt.Rectangle((0,0),1,1,fc=POS_COLORS.get(pos, '#CCCCCC'), edgecolor='none', 
-                       label=f"{POS_TRANSLATIONS[lang].get(pos, pos)}")
-                       for pos in ['NOUN', 'VERB']]
-    ax1.legend(handles=legend_elements, loc='upper left', bbox_to_anchor=(0, 1), fontsize=8)
-    ax2.legend(handles=legend_elements, loc='upper left', bbox_to_anchor=(0, 1), fontsize=8)
-    
-    plt.tight_layout()
-    
-    return fig1, fig2
 
-##################################################################################################################
+    # Identificar entidades y conceptos clave para ambos documentos
+    entities1, key_concepts1 = identify_and_contextualize_entities(doc1, lang)
+    entities2, key_concepts2 = identify_and_contextualize_entities(doc2, lang)
+
+    # Crear grafos de conceptos para ambos documentos
+    concepts1 = [concept for concept, _ in key_concepts1]
+    concepts2 = [concept for concept, _ in key_concepts2]
+    G1 = create_concept_graph(text1, concepts1)
+    G2 = create_concept_graph(text2, concepts2)
+
+    # Visualizar los grafos de conceptos
+    fig1 = visualize_concept_graph(G1, lang)
+    fig2 = visualize_concept_graph(G2, lang)
+
+    # Añadir títulos específicos para cada documento
+    fig1.suptitle("Documento 1: Relaciones Conceptuales", fontsize=16, fontweight='bold')
+    fig2.suptitle("Documento 2: Relaciones Conceptuales", fontsize=16, fontweight='bold')
+
+    return fig1, fig2, entities1, entities2, key_concepts1, key_concepts2
+
 def perform_discourse_analysis(text1, text2, nlp, lang):
-    graph1, graph2 = compare_semantic_analysis(text1, text2, nlp, lang)
-    return graph1, graph2
+    graph1, graph2, entities1, entities2, key_concepts1, key_concepts2 = compare_semantic_analysis(text1, text2, nlp, lang)
+    
+    # Aquí puedes añadir más análisis de discurso si lo necesitas
+    # Por ejemplo, podrías comparar las entidades y conceptos clave entre los dos textos
+
+    return {
+        'graph1': graph1,
+        'graph2': graph2,
+        'entities1': entities1,
+        'entities2': entities2,
+        'key_concepts1': key_concepts1,
+        'key_concepts2': key_concepts2
+    }