Update app.py

app.py

@@ -3,95 +3,220 @@ import requests
 import urllib.parse
 import re
 import xmltodict
-from itertools import product, combinations
+from itertools import product
 import nltk
 from nltk.corpus import stopwords
 from nltk.stem import SnowballStemmer
-from collections import Counter
-import concurrent.futures
-import ssl
-import os
-
-# Configurar SSL para la descarga de NLTK
-try:
-    _create_unverified_https_context = ssl._create_unverified_context
-except AttributeError:
-    pass
-else:
-    ssl._create_default_https_context = _create_unverified_https_context
-
-# Función para descargar recursos de NLTK
-def download_nltk_resources():
-    resources = ['punkt', 'stopwords', 'words']
-    for resource in resources:
-        try:
-            nltk.data.find(f'tokenizers/{resource}')
-        except LookupError:
-            print(f"Descargando {resource}...")
-            nltk.download(resource, quiet=True)
-
-# Descargar recursos de NLTK
-download_nltk_resources()
 
-# Configurar el directorio de datos de NLTK
-nltk_data_dir = '/tmp/nltk_data'
-os.makedirs(nltk_data_dir, exist_ok=True)
-nltk.data.path.append(nltk_data_dir)
+# Descargar recursos necesarios de NLTK
+nltk.download('stopwords', quiet=True)
+nltk.download('punkt', quiet=True)
 
 # Inicializar el stemmer y la lista de stopwords
 stemmer = SnowballStemmer("spanish")
 stop_words = set(stopwords.words('spanish'))
-english_words = set(nltk.corpus.words.words())
 
-# El resto del código permanece igual...
+# Función para obtener sugerencias de DuckDuckGo
+def fetch_duckduckgo_suggestions(query, lang_code="es"):
+    encoded_query = urllib.parse.quote(query)
+    url = f"https://duckduckgo.com/ac/?q={encoded_query}&kl={lang_code}"
+    response = requests.get(url)
+    if response.status_code == 200:
+        try:
+            data = response.json()
+            return [item['phrase'] for item in data]
+        except ValueError:
+            print("Error decodificando JSON de DuckDuckGo")
+            return []
+    else:
+        return []
 
-# Función principal
-def main(keyword):
-    expanded_keywords = expand_keyword(keyword)
-    all_suggestions = []
+# Función para obtener sugerencias de Google
+def fetch_google_suggestions(query, lang_code="es"):
+    encoded_query = urllib.parse.quote(query)
+    url = f"http://suggestqueries.google.com/complete/search?client=firefox&hl={lang_code}&q={encoded_query}"
+    response = requests.get(url)
+    if response.status_code == 200:
+        try:
+            return response.json()[1]
+        except ValueError:
+            print("Error decodificando JSON de Google")
+            return []
+    else:
+        return []
+
+# Función para obtener sugerencias de YouTube
+def fetch_youtube_suggestions(query, lang_code="es"):
+    encoded_query = urllib.parse.quote(query)
+    url = f"http://suggestqueries.google.com/complete/search?client=youtube&hl={lang_code}&q={encoded_query}"
+    response = requests.get(url)
+    
+    if response.status_code == 200:
+        try:
+            match = re.search(r'window\.google\.ac\.h\(\["[^"]*",\[(.*?)\],', response.text)
+            if match:
+                suggestions_data = match.group(1)
+                suggestions = re.findall(r'\["([^"]+)"', suggestions_data)
+                return suggestions
+            else:
+                print("No se encontraron sugerencias en el formato esperado.")
+                return []
+        except Exception as e:
+            print(f"Error procesando la respuesta de YouTube: {e}")
+            return []
+    else:
+        return []
+
+# Función para obtener sugerencias de Bing
+def fetch_bing_suggestions(query, market="es-ES"):
+    url = "https://api.bing.com/qsml.aspx"
+    params = {
+        "Market": market,
+        "query": query
+    }
+    headers = {
+        "User-agent": "Mozilla/5.0"
+    }
+    response = requests.get(url, params=params, headers=headers)
     
-    with concurrent.futures.ThreadPoolExecutor(max_workers=10) as executor:
-        future_to_keyword = {executor.submit(fetch_all_suggestions, kw): kw for kw in expanded_keywords}
-        for future in concurrent.futures.as_completed(future_to_keyword):
-            kw = future_to_keyword[future]
-            try:
-                suggestions = future.result()
-                all_suggestions.extend(suggestions)
-            except Exception as exc:
-                print(f'{kw} generated an exception: {exc}')
+    if response.status_code == 200:
+        try:
+            obj = xmltodict.parse(response.content)
+            suggestList = []
+            if 'SearchSuggestion' in obj and obj['SearchSuggestion']['Section']:
+                suggestions = obj['SearchSuggestion']['Section']['Item']
+                if isinstance(suggestions, list):
+                    for s in suggestions:
+                        suggestList.append(s['Text'])
+                elif isinstance(suggestions, dict):
+                    suggestList.append(suggestions['Text'])
+            return suggestList
+        except Exception as e:
+            print(f"Error procesando la respuesta de Bing: {e}")
+            return []
+    else:
+        return []
+
+# Función para obtener sugerencias de Amazon
+def fetch_amazon_suggestions(query, market_id="A1F83G8C2ARO7P", alias="aps"):
+    url = "https://completion.amazon.com/api/2017/suggestions"
+    params = {
+        "mid": market_id,
+        "alias": alias,
+        "prefix": query
+    }
+    response = requests.get(url, params=params)
     
-    # Contar las sugerencias y sus plataformas
-    suggestion_counter = Counter(sugg for sugg, _ in all_suggestions)
-    platform_counter = Counter(platform for _, platform in all_suggestions)
+    if response.status_code == 200:
+        try:
+            data = response.json()
+            return [item['value'] for item in data.get('suggestions', [])]
+        except ValueError:
+            print("Error decodificando JSON de Amazon")
+            return []
+    else:
+        return []
+
+# Función para expandir la palabra clave
+def expand_keyword(keyword):
+    # Tokenizar la palabra clave
+    tokens = nltk.word_tokenize(keyword.lower())
     
-    # Ordenar las sugerencias por frecuencia
-    sorted_suggestions = sorted(suggestion_counter.items(), key=lambda x: x[1], reverse=True)
+    # Eliminar stopwords y aplicar stemming
+    tokens = [stemmer.stem(token) for token in tokens if token not in stop_words]
     
-    # Crear el HTML de salida
-    html_output = "<h3>Todas las sugerencias ordenadas por relevancia:</h3><ul>"
-    for suggestion, count in sorted_suggestions:
-        platforms = [platform for sugg, platform in all_suggestions if sugg == suggestion]
-        platforms_str = ", ".join(set(platforms))
-        relevance_score = count * len(set(platforms))
-        html_output += f"<li><strong>{suggestion}</strong> (Frecuencia: {count}, Plataformas: {platforms_str}, Puntuación de relevancia: {relevance_score})</li>"
-    html_output += "</ul>"
+    # Generar variaciones
+    variations = []
+    for i in range(1, len(tokens) + 1):
+        variations.extend(list(product(tokens, repeat=i)))
+    
+    # Convertir tuplas a strings
+    expanded_keywords = [" ".join(variation) for variation in variations]
+    
+    # Añadir variaciones con prefijos y sufijos
+    prefixes = ['como', 'que', 'donde', 'cuando', 'por que', 'cual']
+    suffixes = ['gratis', 'online', 'pdf', 'precios', 'opiniones']
     
-    # Añadir estadísticas por plataforma
-    html_output += "<h3>Estadísticas por plataforma:</h3><ul>"
-    for platform, count in platform_counter.most_common():
-        html_output += f"<li>{platform}: {count} sugerencias</li>"
-    html_output += "</ul>"
+    for prefix in prefixes:
+        expanded_keywords.append(f"{prefix} {keyword}")
     
-    # Añadir botón para copiar todas las sugerencias
-    all_suggestions_text = ", ".join(sugg for sugg, _ in sorted_suggestions)
-    html_output += f"""
+    for suffix in suffixes:
+        expanded_keywords.append(f"{keyword} {suffix}")
+    
+    # Añadir plurales (simplificado)
+    expanded_keywords.extend([f"{kw}s" for kw in expanded_keywords])
+    
+    # Eliminar duplicados y la palabra clave original vacía
+    expanded_keywords = list(set(expanded_keywords) - {''})
+    
+    return expanded_keywords
+
+# Función principal
+def main(keyword):
+    expanded_keywords = expand_keyword(keyword)
+    all_suggestions = {}
+    platform_suggestions = {
+        'Google': set(),
+        'DuckDuckGo': set(),
+        'YouTube': set(),
+        'Bing': set(),
+        'Amazon': set()
+    }
+
+    # Obtener sugerencias de todas las plataformas
+    for exp_keyword in expanded_keywords:
+        for platform, fetch_func in [
+            ('Google', fetch_google_suggestions),
+            ('DuckDuckGo', fetch_duckduckgo_suggestions),
+            ('YouTube', fetch_youtube_suggestions),
+            ('Bing', fetch_bing_suggestions),
+            ('Amazon', fetch_amazon_suggestions)
+        ]:
+            suggestions = fetch_func(exp_keyword)
+            platform_suggestions[platform].update(suggestions)
+            for suggestion in suggestions:
+                if suggestion in all_suggestions:
+                    all_suggestions[suggestion] += 1
+                else:
+                    all_suggestions[suggestion] = 1
+
+    # Ordenar y filtrar las sugerencias más frecuentes combinadas
+    sorted_suggestions = sorted(all_suggestions.items(), key=lambda item: item[1], reverse=True)
+    combined_top_suggestions = [sug for sug, freq in sorted_suggestions if freq >= 2][:50]  # Aumentado a top 50
+    suggestions_str = ", ".join(combined_top_suggestions)
+
+    # Crear la lista de todas las palabras clave con su número de repeticiones
+    all_suggestions_str = "<ul>"
+    for suggestion, freq in sorted_suggestions:
+        all_suggestions_str += f"<li>{suggestion} - {freq} repeticiones</li>"
+    all_suggestions_str += "</ul>"
+
+    # Crear el HTML de salida
+    html_output = f"""
     <div>
-        <h3>Copiar todas las sugerencias:</h3>
+        <b>Top 50 Sugerencias combinadas:</b> <span id='suggestions_text'>{suggestions_str}</span>
         <button class="lg secondary svelte-cmf5ev" style="font-size: small; padding: 2px; color: #808080ba; border: none; margin-left: 5px;" 
-        onclick='navigator.clipboard.writeText(`{all_suggestions_text}`).then(() => alert("Todas las sugerencias copiadas al portapapeles"))'>&nbsp;Copiar todas las sugerencias&nbsp;</button>
+        onclick='navigator.clipboard.writeText(document.getElementById("suggestions_text").innerText).then(() => alert("Texto copiado al portapapeles"))'>&nbsp;✂&nbsp;</button>
     </div>
     """
-    
+
+    # Agregar las top sugerencias de cada plataforma
+    for platform, suggestions in platform_suggestions.items():
+        html_output += f"""
+        <h4>Top 20 Sugerencias de {platform}:</h4>
+        <ul>
+        """
+        for suggestion in list(suggestions)[:20]:
+            freq = all_suggestions[suggestion]
+            html_output += f"<li>{suggestion} ({freq})</li>"
+        html_output += "</ul>"
+
+    # Agregar la lista completa de todas las palabras clave
+    html_output += """
+    <h4>Lista completa de palabras clave con su número de repeticiones:</h4>
+    """
+    html_output += all_suggestions_str
+
     return html_output
 
 # Interfaz de Gradio
@@ -100,8 +225,8 @@ iface = gr.Interface(
     inputs="text",
     outputs="html",
     title="<div style='margin:0 auto;text-align:center'><div style='margin:0 auto;text-align:center'><img style='width:100px;display: inline-table;margin-bottom:-10px' src='https://artxeweb.com/media/files/search.jpg'><p>Sugerencias Combinadas de Google, DuckDuckGo, YouTube, Bing y Amazon</p></div>",
-    description="<p style='margin-bottom:10px;text-align:center;background: #ffffff; padding: 8px; border-radius: 8px; border-width: 1px; border: solid 1px #e5e7eb;'>Ingrese una palabra clave para obtener sugerencias de búsqueda relacionadas de Google, DuckDuckGo, YouTube, Bing y Amazon. Se mostrarán todas las sugerencias ordenadas por relevancia.</p>",
+    description="<p style='margin-bottom:10px;text-align:center;background: #ffffff; padding: 8px; border-radius: 8px; border-width: 1px; border: solid 1px #e5e7eb;'>Ingrese una palabra clave para obtener sugerencias de búsqueda relacionadas de Google, DuckDuckGo, YouTube, Bing y Amazon. Se mostrarán las 50 primeras sugerencias combinadas y también las 20 principales de cada plataforma por separado.</p>",
     article="<div style='margin-top:10px'><p style='text-align: center !important; background: #ffffff; padding: 5px 30px; border-radius: 8px; border-width: 1px; border: solid 1px #e5e7eb; width: fit-content; margin: auto;'>Desarrollada por <a style='text-decoration: none !important; color: #e12a31 !important;' href='https://artxeweb.com'>© Artxe Web</a></p></div>"
 )
 
-iface.launch()
+iface.launch()