Spaces:
Sleeping
Sleeping
| # modules/studentact/current_situation_interface.py | |
| import streamlit as st | |
| import logging | |
| from ..utils.widget_utils import generate_unique_key | |
| import matplotlib.pyplot as plt | |
| import numpy as np | |
| from ..database.current_situation_mongo_db import store_current_situation_result | |
| # Importaciones locales | |
| from translations import get_translations | |
| # Importamos la función de recomendaciones personalizadas si existe | |
| try: | |
| from .claude_recommendations import display_personalized_recommendations | |
| except ImportError: | |
| # Si no existe el módulo, definimos una función placeholder | |
| def display_personalized_recommendations(text, metrics, text_type, lang_code, t): | |
| # Obtener el mensaje de advertencia traducido si está disponible | |
| warning = t.get('module_not_available', "Módulo de recomendaciones personalizadas no disponible. Por favor, contacte al administrador.") | |
| st.warning(warning) | |
| from .current_situation_analysis import ( | |
| analyze_text_dimensions, | |
| analyze_clarity, | |
| analyze_vocabulary_diversity, | |
| analyze_cohesion, | |
| analyze_structure, | |
| get_dependency_depths, | |
| normalize_score, | |
| generate_sentence_graphs, | |
| generate_word_connections, | |
| generate_connection_paths, | |
| create_vocabulary_network, | |
| create_syntax_complexity_graph, | |
| create_cohesion_heatmap | |
| ) | |
| # Configuración del estilo de matplotlib para el gráfico de radar | |
| plt.rcParams['font.family'] = 'sans-serif' | |
| plt.rcParams['axes.grid'] = True | |
| plt.rcParams['axes.spines.top'] = False | |
| plt.rcParams['axes.spines.right'] = False | |
| logger = logging.getLogger(__name__) | |
| # Definición de tipos de texto con umbrales | |
| TEXT_TYPES = { | |
| 'academic_article': { | |
| # Los nombres se obtendrán de las traducciones | |
| 'thresholds': { | |
| 'vocabulary': {'min': 0.70, 'target': 0.85}, | |
| 'structure': {'min': 0.75, 'target': 0.90}, | |
| 'cohesion': {'min': 0.65, 'target': 0.80}, | |
| 'clarity': {'min': 0.70, 'target': 0.85} | |
| } | |
| }, | |
| 'student_essay': { | |
| 'thresholds': { | |
| 'vocabulary': {'min': 0.60, 'target': 0.75}, | |
| 'structure': {'min': 0.65, 'target': 0.80}, | |
| 'cohesion': {'min': 0.55, 'target': 0.70}, | |
| 'clarity': {'min': 0.60, 'target': 0.75} | |
| } | |
| }, | |
| 'general_communication': { | |
| 'thresholds': { | |
| 'vocabulary': {'min': 0.50, 'target': 0.65}, | |
| 'structure': {'min': 0.55, 'target': 0.70}, | |
| 'cohesion': {'min': 0.45, 'target': 0.60}, | |
| 'clarity': {'min': 0.50, 'target': 0.65} | |
| } | |
| } | |
| } | |
| #################################################### | |
| #################################################### | |
| def display_current_situation_interface(lang_code, nlp_models, t): | |
| """ | |
| Interfaz simplificada con gráfico de radar para visualizar métricas. | |
| """ | |
| # Agregar logs para depuración | |
| logger.info(f"Idioma: {lang_code}") | |
| logger.info(f"Claves en t: {list(t.keys())}") | |
| # Inicializar estados si no existen | |
| if 'text_input' not in st.session_state: | |
| st.session_state.text_input = "" | |
| if 'text_area' not in st.session_state: | |
| st.session_state.text_area = "" | |
| if 'show_results' not in st.session_state: | |
| st.session_state.show_results = False | |
| if 'current_doc' not in st.session_state: | |
| st.session_state.current_doc = None | |
| if 'current_metrics' not in st.session_state: | |
| st.session_state.current_metrics = None | |
| if 'current_recommendations' not in st.session_state: | |
| st.session_state.current_recommendations = None | |
| try: | |
| # Container principal con dos columnas | |
| with st.container(): | |
| input_col, results_col = st.columns([1,2]) | |
| ############################################################################################### | |
| # CSS personalizado para que el formulario ocupe todo el alto disponible | |
| st.markdown(""" | |
| <style> | |
| /* Hacer que la columna tenga una altura definida */ | |
| [data-testid="column"] { | |
| min-height: 900px; | |
| height: 100vh; /* 100% del alto visible de la ventana */ | |
| } | |
| /* Hacer que el formulario ocupe el espacio disponible en la columna */ | |
| .stForm { | |
| height: calc(100% - 40px); /* Ajuste por márgenes y paddings */ | |
| display: flex; | |
| flex-direction: column; | |
| } | |
| /* Hacer que el área de texto se expanda dentro del formulario */ | |
| .stForm .stTextArea { | |
| flex: 1; | |
| display: flex; | |
| flex-direction: column; | |
| } | |
| /* El textarea en sí debe expandirse */ | |
| .stForm .stTextArea textarea { | |
| flex: 1; | |
| min-height: 750px !important; | |
| } | |
| </style> | |
| """, unsafe_allow_html=True) | |
| ############################################################################################### | |
| with input_col: | |
| with st.form(key=f"text_input_form_{lang_code}"): | |
| text_input = st.text_area( | |
| t.get('input_prompt', "Escribe o pega tu texto aquí:"), | |
| height=800, | |
| key=f"text_area_{lang_code}", | |
| value=st.session_state.text_input, | |
| help=t.get('help', "Este texto será analizado para darte recomendaciones personalizadas") | |
| ) | |
| submit_button = st.form_submit_button( | |
| t.get('analyze_button', "Analizar mi escritura"), | |
| type="primary", | |
| use_container_width=True | |
| ) | |
| if submit_button: | |
| if text_input.strip(): | |
| st.session_state.text_input = text_input | |
| ####################################################################### | |
| # Código para análisis... | |
| try: | |
| with st.spinner(t.get('processing', "Analizando...")): # Usando t.get directamente | |
| doc = nlp_models[lang_code](text_input) | |
| metrics = analyze_text_dimensions(doc) | |
| storage_success = store_current_situation_result( | |
| username=st.session_state.username, | |
| text=text_input, | |
| metrics=metrics, | |
| feedback=None | |
| ) | |
| if not storage_success: | |
| logger.warning("No se pudo guardar el análisis en la base de datos") | |
| st.session_state.current_doc = doc | |
| st.session_state.current_metrics = metrics | |
| st.session_state.show_results = True | |
| except Exception as e: | |
| logger.error(f"Error en análisis: {str(e)}") | |
| st.error(t.get('analysis_error', "Error al analizar el texto")) # Usando t.get directamente | |
| # Mostrar resultados en la columna derecha | |
| with results_col: | |
| if st.session_state.show_results and st.session_state.current_metrics is not None: | |
| # Primero los radio buttons para tipo de texto - usando t.get directamente | |
| st.markdown(f"### {t.get('text_type_header', 'Tipo de texto')}") | |
| # Preparar opciones de tipos de texto con nombres traducidos | |
| text_type_options = {} | |
| for text_type_key in TEXT_TYPES.keys(): | |
| # Fallback a nombres genéricos si no hay traducción | |
| default_names = { | |
| 'academic_article': 'Academic Article' if lang_code == 'en' else 'Article Académique' if lang_code == 'fr' else 'Artigo Acadêmico' if lang_code == 'pt' else 'Artículo Académico', | |
| 'student_essay': 'Student Essay' if lang_code == 'en' else 'Devoir Universitaire' if lang_code == 'fr' else 'Trabalho Universitário' if lang_code == 'pt' else 'Trabajo Universitario', | |
| 'general_communication': 'General Communication' if lang_code == 'en' else 'Communication Générale' if lang_code == 'fr' else 'Comunicação Geral' if lang_code == 'pt' else 'Comunicación General' | |
| } | |
| text_type_options[text_type_key] = default_names.get(text_type_key, text_type_key) | |
| text_type = st.radio( | |
| label=t.get('text_type_header', "Tipo de texto"), # Usando t.get directamente | |
| options=list(TEXT_TYPES.keys()), | |
| format_func=lambda x: text_type_options.get(x, x), | |
| horizontal=True, | |
| key="text_type_radio", | |
| label_visibility="collapsed", | |
| help=t.get('text_type_help', "Selecciona el tipo de texto para ajustar los criterios de evaluación") # Usando t.get directamente | |
| ) | |
| st.session_state.current_text_type = text_type | |
| # Crear subtabs con nombres traducidos | |
| diagnosis_tab = "Diagnosis" if lang_code == 'en' else "Diagnostic" if lang_code == 'fr' else "Diagnóstico" if lang_code == 'pt' else "Diagnóstico" | |
| recommendations_tab = "Recommendations" if lang_code == 'en' else "Recommandations" if lang_code == 'fr' else "Recomendações" if lang_code == 'pt' else "Recomendaciones" | |
| subtab1, subtab2 = st.tabs([diagnosis_tab, recommendations_tab]) | |
| # Mostrar resultados en el primer subtab | |
| with subtab1: | |
| display_diagnosis( | |
| metrics=st.session_state.current_metrics, | |
| text_type=text_type, | |
| lang_code=lang_code, | |
| t=t # Pasar t directamente, no current_situation_t | |
| ) | |
| # Mostrar recomendaciones en el segundo subtab | |
| with subtab2: | |
| # Llamar directamente a la función de recomendaciones personalizadas | |
| display_personalized_recommendations( | |
| text=text_input, | |
| metrics=st.session_state.current_metrics, | |
| text_type=text_type, | |
| lang_code=lang_code, | |
| t=t | |
| ) | |
| except Exception as e: | |
| logger.error(f"Error en interfaz principal: {str(e)}") | |
| st.error(t.get('error_interface', "Ocurrió un error al cargar la interfaz")) # Usando t.get directamente | |
| ################################################################# | |
| ################################################################# | |
| def display_diagnosis(metrics, text_type=None, lang_code='es', t=None): | |
| """ | |
| Muestra los resultados del análisis: métricas verticalmente y gráfico radar. | |
| """ | |
| try: | |
| # Asegurar que tenemos traducciones | |
| if t is None: | |
| t = {} | |
| # Traducciones para títulos y etiquetas | |
| dimension_labels = { | |
| 'es': { | |
| 'title': "Tipo de texto", | |
| 'vocabulary': "Vocabulario", | |
| 'structure': "Estructura", | |
| 'cohesion': "Cohesión", | |
| 'clarity': "Claridad", | |
| 'improvement': "⚠️ Por mejorar", | |
| 'acceptable': "📈 Aceptable", | |
| 'optimal': "✅ Óptimo", | |
| 'target': "Meta: {:.2f}" | |
| }, | |
| 'en': { | |
| 'title': "Text Type", | |
| 'vocabulary': "Vocabulary", | |
| 'structure': "Structure", | |
| 'cohesion': "Cohesion", | |
| 'clarity': "Clarity", | |
| 'improvement': "⚠️ Needs improvement", | |
| 'acceptable': "📈 Acceptable", | |
| 'optimal': "✅ Optimal", | |
| 'target': "Target: {:.2f}" | |
| }, | |
| 'fr': { | |
| 'title': "Type de texte", | |
| 'vocabulary': "Vocabulaire", | |
| 'structure': "Structure", | |
| 'cohesion': "Cohésion", | |
| 'clarity': "Clarté", | |
| 'improvement': "⚠️ À améliorer", | |
| 'acceptable': "📈 Acceptable", | |
| 'optimal': "✅ Optimal", | |
| 'target': "Objectif: {:.2f}" | |
| }, | |
| 'pt': { | |
| 'title': "Tipo de texto", | |
| 'vocabulary': "Vocabulário", | |
| 'structure': "Estrutura", | |
| 'cohesion': "Coesão", | |
| 'clarity': "Clareza", | |
| 'improvement': "⚠️ Precisa melhorar", | |
| 'acceptable': "📈 Aceitável", | |
| 'optimal': "✅ Ótimo", | |
| 'target': "Meta: {:.2f}" | |
| } | |
| } | |
| # Obtener traducciones para el idioma actual, con fallback a español | |
| labels = dimension_labels.get(lang_code, dimension_labels['es']) | |
| # Usar valor por defecto si no se especifica tipo | |
| text_type = text_type or 'student_essay' | |
| # Obtener umbrales según el tipo de texto | |
| thresholds = TEXT_TYPES[text_type]['thresholds'] | |
| # Crear dos columnas para las métricas y el gráfico | |
| metrics_col, graph_col = st.columns([1, 1.5]) | |
| # Columna de métricas | |
| with metrics_col: | |
| metrics_config = [ | |
| { | |
| 'label': labels['vocabulary'], | |
| 'key': 'vocabulary', | |
| 'value': metrics['vocabulary']['normalized_score'], | |
| 'help': t.get('vocabulary_help', "Riqueza y variedad del vocabulario"), | |
| 'thresholds': thresholds['vocabulary'] | |
| }, | |
| { | |
| 'label': labels['structure'], | |
| 'key': 'structure', | |
| 'value': metrics['structure']['normalized_score'], | |
| 'help': t.get('structure_help', "Organización y complejidad de oraciones"), | |
| 'thresholds': thresholds['structure'] | |
| }, | |
| { | |
| 'label': labels['cohesion'], | |
| 'key': 'cohesion', | |
| 'value': metrics['cohesion']['normalized_score'], | |
| 'help': t.get('cohesion_help', "Conexión y fluidez entre ideas"), | |
| 'thresholds': thresholds['cohesion'] | |
| }, | |
| { | |
| 'label': labels['clarity'], | |
| 'key': 'clarity', | |
| 'value': metrics['clarity']['normalized_score'], | |
| 'help': t.get('clarity_help', "Facilidad de comprensión del texto"), | |
| 'thresholds': thresholds['clarity'] | |
| } | |
| ] | |
| # Mostrar métricas con textos traducidos | |
| for metric in metrics_config: | |
| value = metric['value'] | |
| if value < metric['thresholds']['min']: | |
| status = labels['improvement'] | |
| color = "inverse" | |
| elif value < metric['thresholds']['target']: | |
| status = labels['acceptable'] | |
| color = "off" | |
| else: | |
| status = labels['optimal'] | |
| color = "normal" | |
| target_text = labels['target'].format(metric['thresholds']['target']) | |
| st.metric( | |
| metric['label'], | |
| f"{value:.2f}", | |
| f"{status} ({target_text})", | |
| delta_color=color, | |
| help=metric['help'] | |
| ) | |
| st.markdown("<div style='margin-bottom: 0.5rem;'></div>", unsafe_allow_html=True) | |
| # Gráfico radar en la columna derecha | |
| with graph_col: | |
| display_radar_chart(metrics_config, thresholds, lang_code) # Pasar el parámetro lang_code | |
| except Exception as e: | |
| logger.error(f"Error mostrando resultados: {str(e)}") | |
| st.error(t.get('error_results', "Error al mostrar los resultados")) | |
| ################################################################## | |
| ################################################################## | |
| def display_radar_chart(metrics_config, thresholds, lang_code='es'): | |
| """ | |
| Muestra el gráfico radar con los resultados. | |
| """ | |
| try: | |
| # Traducción de las etiquetas de leyenda según el idioma | |
| legend_translations = { | |
| 'es': {'min': 'Mínimo', 'target': 'Meta', 'user': 'Tu escritura'}, | |
| 'en': {'min': 'Minimum', 'target': 'Target', 'user': 'Your writing'}, | |
| 'fr': {'min': 'Minimum', 'target': 'Objectif', 'user': 'Votre écriture'}, | |
| 'pt': {'min': 'Mínimo', 'target': 'Meta', 'user': 'Sua escrita'} | |
| } | |
| # Usar español por defecto si el idioma no está soportado | |
| translations = legend_translations.get(lang_code, legend_translations['es']) | |
| # Preparar datos para el gráfico | |
| categories = [m['label'] for m in metrics_config] | |
| values_user = [m['value'] for m in metrics_config] | |
| min_values = [m['thresholds']['min'] for m in metrics_config] | |
| target_values = [m['thresholds']['target'] for m in metrics_config] | |
| # Crear y configurar gráfico | |
| fig = plt.figure(figsize=(8, 8)) | |
| ax = fig.add_subplot(111, projection='polar') | |
| # Configurar radar | |
| angles = [n / float(len(categories)) * 2 * np.pi for n in range(len(categories))] | |
| angles += angles[:1] | |
| values_user += values_user[:1] | |
| min_values += min_values[:1] | |
| target_values += target_values[:1] | |
| # Configurar ejes | |
| ax.set_xticks(angles[:-1]) | |
| ax.set_xticklabels(categories, fontsize=10) | |
| circle_ticks = np.arange(0, 1.1, 0.2) | |
| ax.set_yticks(circle_ticks) | |
| ax.set_yticklabels([f'{tick:.1f}' for tick in circle_ticks], fontsize=8) | |
| ax.set_ylim(0, 1) | |
| # Dibujar áreas de umbrales con etiquetas traducidas | |
| ax.plot(angles, min_values, '#e74c3c', linestyle='--', linewidth=1, label=translations['min'], alpha=0.5) | |
| ax.plot(angles, target_values, '#2ecc71', linestyle='--', linewidth=1, label=translations['target'], alpha=0.5) | |
| ax.fill_between(angles, target_values, [1]*len(angles), color='#2ecc71', alpha=0.1) | |
| ax.fill_between(angles, [0]*len(angles), min_values, color='#e74c3c', alpha=0.1) | |
| # Dibujar valores del usuario con etiqueta traducida | |
| ax.plot(angles, values_user, '#3498db', linewidth=2, label=translations['user']) | |
| ax.fill(angles, values_user, '#3498db', alpha=0.2) | |
| # Ajustar leyenda | |
| ax.legend( | |
| loc='upper right', | |
| bbox_to_anchor=(1.3, 1.1), | |
| fontsize=10, | |
| frameon=True, | |
| facecolor='white', | |
| edgecolor='none', | |
| shadow=True | |
| ) | |
| plt.tight_layout() | |
| st.pyplot(fig) | |
| plt.close() | |
| except Exception as e: | |
| logger.error(f"Error mostrando gráfico radar: {str(e)}") | |
| st.error("Error al mostrar el gráfico") |