Update elastic stack test

catch-all/05_infra_test/04_elastic_stack/app/Dockerfile

@@ -0,0 +1,15 @@
+# Usa la imagen base de Python
+FROM python:3.9-slim
+
+# Establece el directorio de trabajo
+WORKDIR /app
+
+# Copia el archivo requirements.txt e instala las dependencias
+COPY requirements.txt /app/
+RUN pip install --no-cache-dir -r requirements.txt
+
+# Copia el código fuente a /app
+COPY . /app/
+
+# Comando para ejecutar la aplicación
+CMD ["python", "main.py"]

catch-all/05_infra_test/04_elastic_stack/app/main.py

@@ -0,0 +1,114 @@
+import pandas as pd
+from elasticsearch import Elasticsearch
+from elasticsearch.helpers import bulk
+
+# Configura la conexión a Elasticsearch
+es = Elasticsearch("http://elasticsearch:9200")
+
+
+def create_index():
+    """
+    Crea un índice en Elasticsearch con el nombre 'movies' si no existe.
+    Define el mapeo del índice para los campos de los documentos.
+    """
+
+    # Define el mapeo del índice 'movies'
+    mappings = {
+        "properties": {
+            # Campo para el título de la película
+            "title": {"type": "text", "analyzer": "english"},
+            # Campo para la etnicidad
+            "ethnicity": {"type": "text", "analyzer": "standard"},
+            # Campo para el director
+            "director": {"type": "text", "analyzer": "standard"},
+            # Campo para el elenco
+            "cast": {"type": "text", "analyzer": "standard"},
+            # Campo para el género
+            "genre": {"type": "text", "analyzer": "standard"},
+            # Campo para el argumento de la película
+            "plot": {"type": "text", "analyzer": "english"},
+            # Campo para el año de lanzamiento
+            "year": {"type": "integer"},
+            # Campo para la página de Wikipedia
+            "wiki_page": {"type": "keyword"}
+        }
+    }
+
+    # Verifica si el índice 'movies' ya existe
+    if not es.indices.exists(index="movies"):
+
+        # Crea el índice 'movies' si no existe
+        es.indices.create(index="movies", mappings=mappings)
+        print("\n[+] Índice 'movies' creado.")
+
+    else:
+
+        print("\n[!] El índice 'movies' ya existe.")
+
+
+def load_data():
+    """
+    Carga datos desde un archivo CSV a Elasticsearch.
+    """
+
+    try:
+
+        # Lee el archivo CSV
+        df = pd.read_csv("/app/wiki_movie_plots_deduped.csv", quoting=1)
+
+        # Verifica el número de filas en el DataFrame
+        num_rows = len(df)
+        sample_size = min(5000, num_rows)
+
+        # Elimina filas con valores nulos y toma una muestra
+        df = df.dropna().sample(sample_size, random_state=42).reset_index(drop=True)
+
+    except Exception as e:
+
+        print(f"\n[!] Error al leer el archivo CSV: {e}")
+
+        return
+
+    # Prepara los datos para la carga en Elasticsearch
+    bulk_data = [
+        {
+            "_index": "movies",  # Nombre del índice en Elasticsearch
+            "_id": i,  # ID del documento en Elasticsearch
+            "_source": {
+                "title": row["Title"],  # Título de la película
+                "ethnicity": row["Origin/Ethnicity"],  # Etnicidad
+                "director": row["Director"],  # Director
+                "cast": row["Cast"],  # Elenco
+                "genre": row["Genre"],  # Género
+                "plot": row["Plot"],  # Argumento
+                "year": row["Release Year"],  # Año de lanzamiento
+                "wiki_page": row["Wiki Page"],  # Página de Wikipedia
+            }
+        }
+        for i, row in df.iterrows()  # Itera sobre cada fila del DataFrame
+    ]
+
+    try:
+
+        # Carga los datos en Elasticsearch en bloques
+        bulk(es, bulk_data)
+        print("\n[+] Datos cargados en Elasticsearch.")
+
+    except Exception as e:
+
+        print(f"\n[!] Error al cargar datos en Elasticsearch: {e}")
+
+
+def main():
+    """
+    Función principal que crea el índice y carga los datos.
+    """
+
+    create_index()  # Crea el índice en Elasticsearch
+    load_data()  # Carga los datos en Elasticsearch
+
+
+if __name__ == "__main__":
+    # Ejecuta la función principal si el script se ejecuta directamente
+
+    main()

catch-all/05_infra_test/04_elastic_stack/app/requirements.txt

@@ -0,0 +1,3 @@
+pandas==2.0.1
+numpy==1.24.2
+elasticsearch==8.8.0

catch-all/05_infra_test/04_elastic_stack/app/wiki_movie_plots_deduped.csv

@@ -0,0 +1,7 @@
+Title,Origin/Ethnicity,Director,Cast,Genre,Plot,Release Year,Wiki Page
+The Shawshank Redemption,American,Frank Darabont,"Tim Robbins, Morgan Freeman",Drama,"Two imprisoned men bond over a number of years, finding solace and eventual redemption through acts of common decency.",1994,https://en.wikipedia.org/wiki/The_Shawshank_Redemption
+The Godfather,American,Francis Ford Coppola,"Marlon Brando, Al Pacino",Crime,"The aging patriarch of an organized crime dynasty transfers control of his clandestine empire to his reluctant son.",1972,https://en.wikipedia.org/wiki/The_Godfather
+The Dark Knight,American,Christopher Nolan,"Christian Bale, Heath Ledger",Action,"When the menace known as the Joker emerges from his mysterious past, he wreaks havoc and chaos on the people of Gotham.",2008,https://en.wikipedia.org/wiki/The_Dark_Knight
+Pulp Fiction,American,Quentin Tarantino,"John Travolta, Uma Thurman",Crime,"The lives of two mob hitmen, a boxer, a gangster's wife, and a pair of diner bandits intertwine in four tales of violence and redemption.",1994,https://en.wikipedia.org/wiki/Pulp_Fiction
+The Lord of the Rings: The Return of the King,American,Peter Jackson,"Elijah Wood, Viggo Mortensen",Fantasy,"The final battle for Middle-earth begins. The forces of good and evil are drawn into a confrontation and the outcome will determine the fate of the world.",2003,https://en.wikipedia.org/wiki/The_Lord_of_the_Rings:_The_Return_of_the_King
+Inception,American,Christopher Nolan,"Leonardo DiCaprio, Joseph Gordon-Levitt",Sci-Fi,"A thief who enters the dreams of others to steal secrets from their subconscious is given the inverse task of planting an idea into the mind of a CEO.",2010,https://en.wikipedia.org/wiki/Inception