### Pyparsing Import and Namespace Setup

Source: https://github.com/keepertrading90/rpk-nexus-plan-maestro/blob/main/_SISTEMA/runtime_python/Lib/site-packages/pyparsing/ai/best_practices.md

Demonstrates the recommended way to import pyparsing and set up namespaces for common modules like `pyparsing.common` and `pyparsing.unicode`. This ensures consistent access to library features.

```python
import pyparsing as pp

# For pyparsing.common
ppc = pp.common

# For pyparsing.unicode
ppu = pp.unicode
```

--------------------------------

### Defining Keyword Constants with Using Each

Source: https://github.com/keepertrading90/rpk-nexus-plan-maestro/blob/main/_SISTEMA/runtime_python/Lib/site-packages/pyparsing/ai/best_practices.md

Shows a concise way to define multiple keyword constants using `using_each` with a list of keywords, avoiding repetitive individual assignments.

```python
import pyparsing as pp

# Define multiple keywords efficiently
IF, THEN, ELSE = pp.Keyword.using_each(["if", "then", "else"])

# Example usage
parser = IF + THEN + ELSE
result = parser.parse_string("if then else")
print(result)
```

--------------------------------

### Ensuring Full Input Consumption

Source: https://github.com/keepertrading90/rpk-nexus-plan-maestro/blob/main/_SISTEMA/runtime_python/Lib/site-packages/pyparsing/ai/best_practices.md

Explains how to use the `parse_all=True` argument with `parse_string` to ensure that the entire input string is consumed by the grammar.

```python
import pyparsing as pp

parser = pp.Word(pp.alphanums)

# This will raise an exception if there's unparsed input
result = parser.parse_string("hello world", parse_all=True)
```

--------------------------------

### Parsing Delimited Lists with Optional Handling

Source: https://github.com/keepertrading90/rpk-nexus-plan-maestro/blob/main/_SISTEMA/runtime_python/Lib/site-packages/pyparsing/ai/best_practices.md

Demonstrates the use of `pp.DelimitedList` for parsing comma-separated sequences and `pp.Optional` to handle cases where the list might be empty.

```python
import pyparsing as pp

# Parse a delimited list, optionally empty
items = pp.Optional(pp.DelimitedList(pp.Word(pp.alphanums)))('')

result_empty = items.parse_string("")
print(result_empty)

result_data = items.parse_string("apple,banana,cherry")
print(result_data)
```

--------------------------------

### Handling Comments with Ignore

Source: https://github.com/keepertrading90/rpk-nexus-plan-maestro/blob/main/_SISTEMA/runtime_python/Lib/site-packages/pyparsing/ai/best_practices.md

Demonstrates how to handle comments within the grammar using the `ignore()` method, recommending built-in comment parsers like `pp.cpp_style_comment` and `pp.python_style_comment`.

```python
import pyparsing as pp

# Define grammar elements
content = pp.Word(pp.alphanums)

# Ignore C++ style comments
parser = content.copy().ignore(pp.cpp_style_comment)

result = parser.parse_string("some_data // this is a comment")
print(result[0]) # Output: some_data
```

--------------------------------

### Using Operator Forms for Grammar Elements

Source: https://github.com/keepertrading90/rpk-nexus-plan-maestro/blob/main/_SISTEMA/runtime_python/Lib/site-packages/pyparsing/ai/best_practices.md

Demonstrates the preference for using operator forms like `+`, `|`, `^`, `~` over explicit pyparsing classes (e.g., `And`, `MatchFirst`, `Or`, `Not`) for improved readability in grammar definitions.

```python
import pyparsing as pp

# Example using operator forms
parser = pp.Word(pp.alphas) + pp.Optional(pp.Word(pp.nums)) | pp.Keyword("special")
```

--------------------------------

### Using Each for Order-Independent Elements

Source: https://github.com/keepertrading90/rpk-nexus-plan-maestro/blob/main/_SISTEMA/runtime_python/Lib/site-packages/pyparsing/ai/best_practices.md

Explains the use of `pp.Each()` to define elements that can appear in any order within the grammar. The '&' operator is presented as a more readable alternative.

```python
import pyparsing as pp

# Using Each operator
parser_each = pp.Each([
    pp.Keyword("first"),
    pp.Keyword("second")
])

# Using '&' operator (more readable)
parser_amp = pp.Keyword("first") & pp.Keyword("second")

result = parser_amp.parse_string("second first")
print(result)
```

--------------------------------

### Organizing Sub-expressions with Group

Source: https://github.com/keepertrading90/rpk-nexus-plan-maestro/blob/main/_SISTEMA/runtime_python/Lib/site-packages/pyparsing/ai/best_practices.md

Highlights the use of `pp.Group` for organizing related sub-expressions and preserving results names, especially when used within `OneOrMore` or `ZeroOrMore`.

```python
import pyparsing as pp

# Grouping related elements
coordinates = pp.Group(
    pp.Word(pp.nums)("x") + pp.Word(pp.nums)("y")
)

result = coordinates.parse_string("10 20")
print(result.x) # Output: 10
print(result.y) # Output: 20
```

--------------------------------

### GET /api/customers

Source: https://context7.com/keepertrading90/rpk-nexus-plan-maestro/llms.txt

Endpoint to list all customers along with their total stock value.

```APIDOC
## GET /api/customers

### Description
Endpoints for exploring inventory by customer, obtaining a list of items with their values, and calculating historical averages in specific date ranges.

### Method
GET

### Endpoint
/api/customers

### Parameters
#### Query Parameters
None

#### Request Body
None

### Request Example
```bash
curl -X GET "http://localhost:8000/api/customers"
```

### Response
#### Success Response (200)
- **customers** (array of objects) - A list of customers and their associated total stock value.
  - **Cliente** (string) - The name of the customer.
  - **Valor_Total** (float) - The total stock value for the customer.

#### Response Example
```json
{
    "customers": [
        {"Cliente": "FORD ESPAÑA", "Valor_Total": 2500000.50},
        {"Cliente": "SEAT S.A.", "Valor_Total": 1850000.25}
    ]
}
```
```

--------------------------------

### FastAPI Server Configuration (Python)

Source: https://github.com/keepertrading90/rpk-nexus-plan-maestro/blob/main/docs/architecture/RPK_NEXUS_ARCHITECTURE.md

This snippet illustrates the backend server setup using FastAPI, highlighting its use for routing API data and static files. It also mentions intelligent routing and security practices like 'Zero-Trust'.

```python
# backend/server_nexus.py
# Utiliza el framework FastAPI por su alto rendimiento y tipado estático (Pydantic).
# - Enrutamiento Inteligente: Gestiona tanto la API de datos como el servicio de archivos estáticos.
# - Redirecciones Relativas: Implementa lógica para forzar barras al final.
# - Seguridad: Sigue el patrón "Zero-Trust" validando cada entrada de datos.
```

--------------------------------

### Handling Whitespace and Line Endings

Source: https://github.com/keepertrading90/rpk-nexus-plan-maestro/blob/main/_SISTEMA/runtime_python/Lib/site-packages/pyparsing/ai/best_practices.md

Provides guidance on configuring pyparsing to handle whitespace and significant newlines. It shows how to set default whitespace characters and explicitly define line endings for line-oriented grammars.

```python
import pyparsing as pp

# Set default whitespace to only spaces and tabs
pp.ParserElement.set_default_whitespace_chars(" \t")

# Define a suppressor for line endings
NL = pp.LineEnd().suppress()
```

--------------------------------

### Enabling Packrat Parsing for Recursive Grammars

Source: https://github.com/keepertrading90/rpk-nexus-plan-maestro/blob/main/_SISTEMA/runtime_python/Lib/site-packages/pyparsing/ai/best_practices.md

Shows how to enable packrat parsing immediately after importing pyparsing to optimize performance for grammars with recursive elements, such as those using `Forward()` or `infix_notation()`.

```python
import pyparsing as pp

# Enable packrat parsing for performance
pp.ParserElement.enable_packrat()
```

--------------------------------

### Applying Parse Actions for Data Conversion

Source: https://github.com/keepertrading90/rpk-nexus-plan-maestro/blob/main/_SISTEMA/runtime_python/Lib/site-packages/pyparsing/ai/best_practices.md

Illustrates how to use parse actions to convert parsed string data into Python data types. It covers common types, unquoting strings, replacing keywords with Python constants, and using `aslist`/`asdict` for native containers.

```python
import pyparsing as pp

# Parse integer using pyparsing.common
integer_parser = pp.common.integer
result_int = integer_parser.parse_string("123")
print(type(result_int[0])) # Output: <class 'int'>

# Unquote and parse double-quoted string
dbl_quoted = pp.dbl_quoted_string().set_parse_action(pp.remove_quotes)
result_str = dbl_quoted.parse_string('"hello world"')
print(result_str[0]) # Output: hello world

# Replace keyword 'true' with Python True
boolean_true = pp.Keyword("true").set_parse_action(pp.replace_with(True))
result_bool = boolean_true.parse_string("true")
print(result_bool[0]) # Output: True

# Parse into native Python list
list_parser = pp.Group(pp.Word(pp.nums) + pp.Word(pp.alphanums), aslist=True)
result_list = list_parser.parse_string("10 abc")
print(result_list) # Output: [['10', 'abc']]

# Parse into native Python dict
dict_parser = pp.Dict(pp.Word(pp.alphas)("key") + pp.Word(pp.nums)("value"), asdict=True)
result_dict = dict_parser.parse_string("name 42")
print(result_dict) # Output: [{'key': 'name', 'value': '42'}]
```

--------------------------------

### API: Listar Escenarios

Source: https://context7.com/keepertrading90/rpk-nexus-plan-maestro/llms.txt

Realiza una solicitud GET a la ruta /api/scenarios para obtener una lista de todos los escenarios guardados en el sistema. No requiere parámetros de entrada.

```bash
curl -X GET "http://localhost:8000/api/scenarios"
```

--------------------------------

### API: Ejecutar Simulación

Source: https://context7.com/keepertrading90/rpk-nexus-plan-maestro/llms.txt

Ejecuta una simulación para un escenario específico utilizando una solicitud GET a /api/simulate/{id}. Permite especificar días laborales y horas de turno como parámetros de consulta.

```bash
curl -X GET "http://localhost:8000/api/simulate/1?dias_laborales=240&horas_turno=24"
```

--------------------------------

### API: Obtener Historial de Cambios

Source: https://context7.com/keepertrading90/rpk-nexus-plan-maestro/llms.txt

Recupera el historial de cambios de un escenario particular a través de una solicitud GET a /api/scenarios/{id}/history.

```bash
curl -X GET "http://localhost:8000/api/scenarios/1/history"
```

--------------------------------

### Using Keyword for Reserved Words

Source: https://github.com/keepertrading90/rpk-nexus-plan-maestro/blob/main/_SISTEMA/runtime_python/Lib/site-packages/pyparsing/ai/best_practices.md

Recommends using `pp.Keyword` over `pp.Literal` for matching reserved words to prevent partial matches. `pp.CaselessKeyword` is available for case-insensitive matching.

```python
import pyparsing as pp

# Using Keyword to match reserved words precisely
for_keyword = pp.Keyword("for")

# Using CaselessKeyword for case-insensitive matching
if_keyword = pp.CaselessKeyword("if")
```

--------------------------------

### Choosing Appropriate Parsing Methods

Source: https://github.com/keepertrading90/rpk-nexus-plan-maestro/blob/main/_SISTEMA/runtime_python/Lib/site-packages/pyparsing/ai/best_practices.md

Compares and contrasts the different parsing methods available in pyparsing: `parse_string` for full parsing, `search_string` for finding matches anywhere, `scan_string` for yielding all matches with positions, and `transform_string` for batch transformations.

```python
import pyparsing as pp

parser = pp.Word("abc")
text = "abc xyz abc"

# parse_string: Parses from the start, requires parse_all=True for full consumption
result_parse = parser.parse_string(text, parse_all=True)
print(f"parse_string: {result_parse}")

# search_string: Searches anywhere in the text
result_search = parser.search_string(text)
print(f"search_string: {list(result_search)}")

# scan_string: Yields all matches with positions
result_scan = parser.scan_string(text)
print(f"scan_string: {list(result_scan)}")

# transform_string: Batch transforms or conversions
def upper_transform(s):
    return s[0].upper()

transformed_text = parser.transform_string(text, upper_transform)
print(f"transform_string: {transformed_text}")
```

--------------------------------

### Defining Recursive Grammar Elements with Forward

Source: https://github.com/keepertrading90/rpk-nexus-plan-maestro/blob/main/_SISTEMA/runtime_python/Lib/site-packages/pyparsing/ai/best_practices.md

Illustrates the use of `pp.Forward()` to define placeholders for recursive grammar elements, which are later assigned using the `<<=` operator. Assigning meaningful names with `set_name()` is crucial for error reporting.

```python
import pyparsing as pp

# Define a forward declaration for a recursive element
recursive_element = pp.Forward().set_name("recursive_element")

# Later, assign the actual grammar to the forward
recursive_element <<= pp.Word(pp.alphanums) | pp.Group(recursive_element + recursive_element)
```

--------------------------------

### Creating Masked Arrays with Old and New NumPy Versions

Source: https://github.com/keepertrading90/rpk-nexus-plan-maestro/blob/main/_SISTEMA/runtime_python/Lib/site-packages/numpy/ma/README.rst

Demonstrates the creation of masked arrays using both the older numpy.core.ma and the newer maskedarray packages. It shows how the representation and comparison of these arrays differ, particularly in how masked values are displayed and how masks are handled.

```python
>>> import numpy.core.ma as old_ma
>>> import maskedarray as new_ma
>>> x = old_ma.array([1,2,3,4,5], mask=[0,0,1,0,0])
>>> x
array(data = 
 [     1      2 999999      4      5],
      mask = 
 [False False True False False],
      fill_value=999999)
>>> y = new_ma.array([1,2,3,4,5], mask=[0,0,1,0,0])
>>> y
array(data = [1 2 -- 4 5],
      mask = [False False True False False],
      fill_value=999999)
```

--------------------------------

### Comparing Masked Arrays from Old and New NumPy Versions

Source: https://github.com/keepertrading90/rpk-nexus-plan-maestro/blob/main/_SISTEMA/runtime_python/Lib/site-packages/numpy/ma/README.rst

Illustrates the comparison between masked arrays created with the old numpy.core.ma and the new maskedarray packages. It highlights how the equality comparison results in a masked array and how the masks themselves are extracted and compared.

```python
>>> x==y
array(data = 
 [True True True True True],
      mask = 
 [False False True False False],
      fill_value=?)
>>> old_ma.getmask(x) == new_ma.getmask(x)
array([True, True, True, True, True])
>>> old_ma.getmask(y) == new_ma.getmask(y)
array([True, True, False, True, True])
>>> old_ma.getmask(y)
False
```

--------------------------------

### GET /api/summary

Source: https://context7.com/keepertrading90/rpk-nexus-plan-maestro/llms.txt

Unified endpoint that returns KPIs, temporal evolution, and rankings according to the requesting module. It automatically detects if the request comes from the stock or times module based on the referer header.

```APIDOC
## GET /api/summary

### Description
Unified endpoint that returns KPIs, temporal evolution, and rankings according to the requesting module. It automatically detects if the request comes from the stock or times module based on the referer header.

### Method
GET

### Endpoint
/api/summary

### Parameters
#### Query Parameters
- **fecha_inicio** (string) - Optional - The start date for the data range (YYYY-MM-DD).
- **fecha_fin** (string) - Optional - The end date for the data range (YYYY-MM-DD).

#### Request Body
None

### Request Example (Load Times)
```bash
curl -X GET "http://localhost:8000/api/summary?fecha_inicio=2026-01-01&fecha_fin=2026-02-18"
```

### Response (Load Times)
#### Success Response (200)
- **kpis** (object) - Key performance indicators for load times.
  - **total_carga** (float) - Total load.
  - **media_carga** (float) - Average load.
  - **num_centros** (integer) - Number of work centers.
  - **num_dias** (integer) - Number of days in the period.
- **evolucion_total** (object) - Temporal evolution of total load.
  - **fechas** (array of strings) - Dates in the period.
  - **cargas** (array of floats) - Load values corresponding to the dates.
- **evolucion_centros** (object) - Temporal evolution of load per work center.
  - **[centro_id]** (object) - Data for a specific work center.
    - **fechas** (array of strings) - Dates.
    - **cargas** (array of floats) - Load values.
- **rankings** (array of objects) - Rankings of work centers.
  - **Centro** (string) - Work center identifier.
  - **Carga_Total** (float) - Total load for the center.
  - **Media_Diaria** (float) - Daily average load for the center.
- **ultima_fecha** (string) - The last date for which data is available.

#### Response Example (Load Times)
```json
{
    "kpis": {
        "total_carga": 45620.5,
        "media_carga": 1520.68,
        "num_centros": 45,
        "num_dias": 30
    },
    "evolucion_total": {
        "fechas": ["2026-01-01", "2026-01-02", "..."],
        "cargas": [1450.2, 1523.8, "..."]
    },
    "evolucion_centros": {
        "782": {"fechas": ["..."], "cargas": ["..."]},
        "640": {"fechas": ["..."], "cargas": ["..."]}
    },
    "rankings": [
        {"Centro": "782", "Carga_Total": 4520.5, "Media_Diaria": 150.68},
        {"Centro": "640", "Carga_Total": 3890.2, "Media_Diaria": 129.67}
    ],
    "ultima_fecha": "2026-02-18"
}
```

### Response (Stock)
#### Success Response (200)
- **kpis** (object) - Key performance indicators for stock.
- **evolucion_total** (object) - Temporal evolution of total stock.
- **top_customers** (array of objects) - Top customers by stock value.
- **top_items** (array of objects) - Top items by stock value.

#### Response Example (Stock)
```json
{
    "kpis": { ... },
    "evolucion_total": { ... },
    "top_customers": [ ... ],
    "top_items": [ ... ]
}
```
```

--------------------------------

### Get Work Center Article Breakdown by Month (REST API)

Source: https://context7.com/keepertrading90/rpk-nexus-plan-maestro/llms.txt

Provides a detailed breakdown of articles processed in a specific work center for a given month. The output includes article details, associated production orders, hours, days, and percentage contribution.

```bash
curl -X GET "http://localhost:8000/api/centro/782/articulos/mes/2026-02"
```

--------------------------------

### Get Pending Sales Order Articles (REST API)

Source: https://context7.com/keepertrading90/rpk-nexus-plan-maestro/llms.txt

Fetches a list of articles with pending sales orders for a specific date. The response includes article details, customer reference, quantity, and amount.

```bash
curl -X GET "http://localhost:8000/api/pedidos/articulos?fecha=2026-02-18"
```

--------------------------------

### Get Base Simulation (REST API)

Source: https://context7.com/keepertrading90/rpk-nexus-plan-maestro/llms.txt

Retrieves the base simulation results for production capacity without any modifications. It includes detailed article-level data and a summary of center-level metrics, based on specified working days and shift hours.

```bash
curl -X GET "http://localhost:8000/api/simulate/base?dias_laborales=238&horas_turno=16"
```

--------------------------------

### API: Previsualizar Simulación

Source: https://context7.com/keepertrading90/rpk-nexus-plan-maestro/llms.txt

Realiza una solicitud POST a /api/simulate/preview para obtener una vista previa de una simulación sin guardarla. Requiere un cuerpo JSON con detalles de configuración y posibles anulaciones.

```bash
curl -X POST "http://localhost:8000/api/simulate/preview" \
  -H "Content-Type: application/json" \
  -d '{
    "dias_laborales": 240,
    "horas_turno": 16,
    "center_configs": {"782": {"shifts": 24}},
    "overrides": [
        {"articulo": "123456", "centro": "782", "oee_override": 0.92}
    ]
}'
```

--------------------------------

### API: Actualizar Escenario Completo

Source: https://context7.com/keepertrading90/rpk-nexus-plan-maestro/llms.txt

Actualiza un escenario existente de forma completa mediante una solicitud PUT a /api/scenarios/{id}/full. El cuerpo de la solicitud debe contener todos los campos del escenario a actualizar.

```bash
curl -X PUT "http://localhost:8000/api/scenarios/1/full" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "Escenario Q3 2026 - Rev2",
    "description": "Revisión con nuevos datos de demanda",
    "dias_laborales": 242,
    "horas_turno_global": 16,
    "center_configs": {},
    "overrides": []
}'
```

--------------------------------

### Resumen de Tiempos de Carga API (Bash)

Source: https://context7.com/keepertrading90/rpk-nexus-plan-maestro/llms.txt

Endpoint unificado que retorna KPIs, evolución temporal y rankings para el módulo de tiempos de carga. Requiere filtros de fecha y detecta automáticamente el módulo solicitante por el header 'referer'.

```bash
# Resumen de Tiempos de Carga (con filtros de fecha)
curl -X GET "http://localhost:8000/api/summary?fecha_inicio=2026-01-01&fecha_fin=2026-02-18"

# Respuesta esperada (Tiempos):
{
    "kpis": {
        "total_carga": 45620.5,
        "media_carga": 1520.68,
        "num_centros": 45,
        "num_dias": 30
    },
    "evolucion_total": {
        "fechas": ["2026-01-01", "2026-01-02", "..."],
        "cargas": [1450.2, 1523.8, "..."]
    },
    "evolucion_centros": {
        "782": {"fechas": ["..."], "cargas": ["..."]},
        "640": {"fechas": ["..."], "cargas": ["..."]}
    },
    "rankings": [
        {"Centro": "782", "Carga_Total": 4520.5, "Media_Diaria": 150.68},
        {"Centro": "640", "Carga_Total": 3890.2, "Media_Diaria": 129.67}
    ],
    "ultima_fecha": "2026-02-18"
}

# Resumen de Stock (llamado desde módulo /mod/stock)
# Respuesta incluye: kpis, evolucion_total, top_customers, top_items
```

--------------------------------

### Modelos SQLAlchemy: Inicialización de Base de Datos

Source: https://context7.com/keepertrading90/rpk-nexus-plan-maestro/llms.txt

Inicializa las tablas necesarias para el sistema de escenarios de simulación utilizando la función init_sim_db() y crea una sesión de base de datos con SessionLocal().

```python
from backend.db.models_sim import init_sim_db, SessionLocal

# Inicializar tablas del simulador
init_sim_db()

# Crear sesión de base de datos
db = SessionLocal()
```

--------------------------------

### Suppressing Punctuation Tokens

Source: https://github.com/keepertrading90/rpk-nexus-plan-maestro/blob/main/_SISTEMA/runtime_python/Lib/site-packages/pyparsing/ai/best_practices.md

Shows a pattern for suppressing punctuation tokens to keep parse results clean, using `pp.Suppress.using_each()` for common punctuation characters.

```python
import pyparsing as pp

# Suppress common punctuation
LBRACK, RBRACK, LBRACE, RBRACE, COLON = pp.Suppress.using_each("[]{}:")

# Example usage
parser = LBRACK + pp.Word(pp.alphanums) + RBRACK
result = parser.parse_string("[data]")
print(result)
```

--------------------------------

### Frontend JavaScript: Inicialización y Comunicación con API REST

Source: https://context7.com/keepertrading90/rpk-nexus-plan-maestro/llms.txt

Este snippet muestra la configuración base de la API, el estado global de la aplicación, y funciones asíncronas para inicializar el sistema, renderizar el dashboard, mostrar desglose de centros, y actualizar gráficos de comparación. Utiliza `fetch` para interactuar con endpoints REST y `Promise.all` para peticiones concurrentes.

```javascript
// Configuración base de la API
const API_BASE = '/api';

// Estado global de la aplicación
const state = {
    currentTab: 'dashboard',
    fechas: [],
    centros: [],
    selectedCentros: [],
    charts: {},
    filters: { from: '', to: '' }
};

// Inicialización del sistema
async function initializeSystem() {
    const [fechasRes, centrosRes, statusRes] = await Promise.all([
        fetch(`${API_BASE}/fechas`),
        fetch(`${API_BASE}/centros`),
        fetch(`${API_BASE}/status`)
    ]);

    const fechasData = await fechasRes.json();
    const centrosData = await centrosRes.json();

    state.fechas = fechasData.fechas || [];
    state.centros = centrosData.centros || [];
}

// Renderizar dashboard principal
async function renderDashboard() {
    const params = new URLSearchParams({
        fecha_inicio: state.filters.from,
        fecha_fin: state.filters.to
    });

    const res = await fetch(`${API_BASE}/summary?${params}`);
    const data = await res.json();

    // Actualizar KPIs
    document.getElementById('kpi-total-carga').innerText = data.kpis.total_carga.toLocaleString();
    document.getElementById('kpi-media-carga').innerText = data.kpis.media_carga.toLocaleString();
    document.getElementById('kpi-centros').innerText = data.kpis.num_centros;
}

// Mostrar desglose (drilldown) de un centro
async function showDrilldown(centro) {
    const mes = state.filters.to.substring(0, 7);  // YYYY-MM
    const res = await fetch(`${API_BASE}/centro/${centro}/articulos/mes/${mes}`);
    const data = await res.json();

    // Renderizar tabla de artículos con horas, días y porcentaje de carga
    data.articulos.forEach(art => {
        console.log(`${art.articulo}: ${art.horas}h (${art.porcentaje}%)`);
    });
}

// Comparar evolución de múltiples centros
async function updateComparisonChart() {
    const params = new URLSearchParams({
        fecha_inicio: state.filters.from,
        fecha_fin: state.filters.to
    });

    const centrosIds = state.selectedCentros.join(',');
    const res = await fetch(`${API_BASE}/centro/${centrosIds}?${params}`);
    const data = await res.json();

    // data.fechas: Array de fechas
    // data.centros: { "782": { cargas: [...] }, "640": { cargas: [...] } }
}

```

--------------------------------

### Listar Clientes con Valor de Stock API (Bash)

Source: https://context7.com/keepertrading90/rpk-nexus-plan-maestro/llms.txt

Endpoint para explorar el inventario por cliente. Retorna un listado de todos los clientes junto con el valor total de stock asociado a cada uno de ellos.

```bash
# Listar todos los clientes con su valor de stock
curl -X GET "http://localhost:8000/api/customers"

# Respuesta:
{
    "customers": [
        {"Cliente": "FORD ESPAÑA", "Valor_Total": 2500000.50},
        {"Cliente": "SEAT S.A.", "Valor_Total": 1850000.25}
    ]
}
```

--------------------------------

### GET /api/customer/{cliente_nombre}/items

Source: https://context7.com/keepertrading90/rpk-nexus-plan-maestro/llms.txt

Endpoint to retrieve a specific customer's items, including their values within a specified date range.

```APIDOC
## GET /api/customer/{cliente_nombre}/items

### Description
Endpoint to retrieve a specific customer's items, including their values within a specified date range.

### Method
GET

### Endpoint
/api/customer/{cliente_nombre}/items

### Parameters
#### Path Parameters
- **cliente_nombre** (string) - Required - The name of the customer. URL-encoded if it contains spaces or special characters.

#### Query Parameters
- **fecha_inicio** (string) - Optional - The start date for filtering items (YYYY-MM-DD).
- **fecha_fin** (string) - Optional - The end date for filtering items (YYYY-MM-DD).

#### Request Body
None

### Request Example
```bash
curl -X GET "http://localhost:8000/api/customer/FORD%20ESPAÑA/items?fecha_inicio=2026-01-01&fecha_fin=2026-02-18"
```

### Response
#### Success Response (200)
- **items** (array of objects) - A list of items associated with the customer.
  - **Articulo** (string) - The item identifier.
  - **Cantidad** (integer) - The quantity of the item.
  - **Valor_Unitario** (float) - The unit value of the item.
  - **Valor_Total** (float) - The total value of the item (Cantidad * Valor_Unitario).
  - **Fecha** (string) - The date the item data pertains to.

#### Response Example
```json
{
    "items": [
        {"Articulo": "ITEM001", "Cantidad": 100, "Valor_Unitario": 25.00, "Valor_Total": 2500.00, "Fecha": "2026-01-15"},
        {"Articulo": "ITEM002", "Cantidad": 50, "Valor_Unitario": 37.00, "Valor_Total": 1850.00, "Fecha": "2026-01-15"}
    ]
}
```
```

--------------------------------

### GET /api/v1/status

Source: https://context7.com/keepertrading90/rpk-nexus-plan-maestro/llms.txt

Endpoint to verify the connectivity and general status of the NEXUS database. It returns information about the connection and record counts in the main tables.

```APIDOC
## GET /api/v1/status

### Description
Endpoint to verify the connectivity and general status of the NEXUS database. Returns information about the connection and record counts in the main tables.

### Method
GET

### Endpoint
/api/v1/status

### Parameters
#### Query Parameters
None

#### Request Body
None

### Request Example
```bash
curl -X GET "http://localhost:8000/api/v1/status"
```

### Response
#### Success Response (200)
- **status** (string) - The current status of the system (e.g., "online").
- **db_path** (string) - The absolute path to the SQLite database file.
- **records** (object) - An object containing the count of records in key tables.
  - **stock** (integer) - Number of records in the stock table.
  - **tiempos** (integer) - Number of records in the times/load table.
- **database** (string) - The name of the database file.

#### Response Example
```json
{
    "status": "online",
    "db_path": "/path/to/rpk_industrial.db",
    "records": {
        "stock": 15420,
        "tiempos": 8750
    },
    "database": "rpk_industrial.db"
}
```
```

--------------------------------

### Consultar Datos con Lenguaje Natural API (Bash)

Source: https://context7.com/keepertrading90/rpk-nexus-plan-maestro/llms.txt

Permite realizar consultas en lenguaje natural sobre los datos de producción. Utiliza un motor de traducción a SQL para interpretar preguntas sobre stock, saturación y carga de trabajo. Soporta diversas consultas predefinidas.

```bash
# Consultar datos mediante lenguaje natural
curl -X POST "http://localhost:8000/api/v1/chat" \
  -H "Content-Type: application/json" \
  -d '{"text": "¿Cuál es el stock total?"}'

# Respuesta esperada:
{
    "response": "He consultado la base de datos NEXUS.\n\nLos datos que he encontrado son:\n- **Total_Stock**: 2500000\n- **Valor_Total**: 15750000.50"
}

# Otras consultas soportadas:
curl -X POST "http://localhost:8000/api/v1/chat" \
  -H "Content-Type: application/json" \
  -d '{"text": "stock por cliente"}'

curl -X POST "http://localhost:8000/api/v1/chat" \
  -H "Content-Type: application/json" \
  -d '{"text": "saturacion critica"}'

curl -X POST "http://localhost:8000/api/v1/chat" \
  -H "Content-Type: application/json" \
  -d '{"text": "top 10 articulos"}'
```

--------------------------------

### Configure Matplotlib to Use Maskedarray

Source: https://github.com/keepertrading90/rpk-nexus-plan-maestro/blob/main/_SISTEMA/runtime_python/Lib/site-packages/numpy/ma/README.rst

Demonstrates how to configure Matplotlib to use the external maskedarray module instead of numpy.ma. This can be done by modifying the matplotlibrc file or using a command-line option.

```text
# In matplotlibrc file:
#maskedarray : False       # True to use external maskedarray module
                             # instead of numpy.ma; this is a temporary #
                             # setting for testing maskedarray.

# Command-line option:
# python simple_plot.py --maskedarray
```

--------------------------------

### Grouping Sub-expressions with Results Names

Source: https://github.com/keepertrading90/rpk-nexus-plan-maestro/blob/main/_SISTEMA/runtime_python/Lib/site-packages/pyparsing/ai/best_practices.md

Details the requirement to use `pp.Group` when adding a results name to an expression containing sub-expressions that also have results names. This preserves the results names of the nested elements.

```python
import pyparsing as pp

# Grouping is necessary when sub-expressions have results names
address = pp.Group(
    pp.Word(pp.alphanums)("street") +
    pp.Word(pp.alphanums)("city")
)("location")

result = address.parse_string("123 Main Anytown")
print(result.location[0].street) # Output: 123
print(result.location[0].city)  # Output: Main
```

--------------------------------

### Obtener Artículos de un Cliente Específico API (Bash)

Source: https://context7.com/keepertrading90/rpk-nexus-plan-maestro/llms.txt

Endpoint para explorar el inventario por cliente. Permite obtener el listado de artículos para un cliente específico dentro de un rango de fechas determinado.

```bash
# Obtener artículos de un cliente específico
curl -X GET "http://localhost:8000/api/customer/FORD%20ESPAÑA/items?fecha_inicio=2026-01-01&fecha_fin=2026-02-18"
```

--------------------------------

### Apply Date Filters

Source: https://github.com/keepertrading90/rpk-nexus-plan-maestro/blob/main/frontend/modules/stock/index.html

Applies date filters to the data by calling the loadData function with start and end dates obtained from input elements with IDs 'date-start' and 'date-end'.

```javascript
function applyFilters() {
  loadData(document.getElementById('date-start').value, document.getElementById('date-end').value);
}
```

--------------------------------

### Fetch and Update Ranking Section (JavaScript)

Source: https://github.com/keepertrading90/rpk-nexus-plan-maestro/blob/main/frontend/modules/stock/index.html

Fetches customer data from the API based on selected start and end dates and then updates the ranking table. This function is asynchronous.

```javascript
async function updateRankingSection() {
  const start = document.getElementById('date-ranking-start').value;
  const end = document.getElementById('date-ranking-end').value;
  const res = await fetch(`/api/customers?fecha_inicio=${start}&fecha_fin=${end}`);
  const data = await res.json();
  updateRankingTable(data.customers);
}
```

--------------------------------

### Modelos SQLAlchemy: Creación de Escenario

Source: https://context7.com/keepertrading90/rpk-nexus-plan-maestro/llms.txt

Crea un nuevo escenario de simulación programáticamente utilizando los modelos SQLAlchemy definidos. Permite especificar detalles como nombre, descripción, días laborales, horas de turno y configuraciones de centro.

```python
from backend.db.models_sim import Scenario, SessionLocal

db = SessionLocal()

# Crear un nuevo escenario programáticamente
nuevo_escenario = Scenario(
    name="Plan Expansión 2027",
    description="Evaluación de nueva línea de producción",
    dias_laborales=245,
    horas_turno_global=24,
    center_configs_json='{"782": {"shifts": 24, "personnel_ratio": 1.5}}'
)
db.add(nuevo_escenario)
db.commit()
```

--------------------------------

### GET /api/v1/hub_stats

Source: https://context7.com/keepertrading90/rpk-nexus-plan-maestro/llms.txt

Endpoint that provides the main KPIs for the central dashboard: total stock, average saturation of work centers, and theoretical coverage days calculated through data cross-analysis.

```APIDOC
## GET /api/v1/hub_stats

### Description
Endpoint that provides the main KPIs for the central dashboard: total stock, average saturation of work centers, and theoretical coverage days calculated through data cross-analysis.

### Method
GET

### Endpoint
/api/v1/hub_stats

### Parameters
#### Query Parameters
None

#### Request Body
None

### Request Example
```bash
curl -X GET "http://localhost:8000/api/v1/hub_stats"
```

### Response
#### Success Response (200)
- **stock** (object) - Information about the total stock.
  - **total** (integer) - The total value or quantity of stock.
  - **items** (integer) - The number of distinct stock items.
- **saturation** (float) - The average saturation percentage of work centers.
- **cobertura** (float) - The theoretical coverage in days.

#### Response Example
```json
{
    "stock": {
        "total": 2500000,
        "items": 1847
    },
    "saturation": 74.5,
    "cobertura": 12.4
}
```
```

--------------------------------

### Get TT Table Class by Tag

Source: https://github.com/keepertrading90/rpk-nexus-plan-maestro/blob/main/_SISTEMA/runtime_python/Lib/site-packages/fontTools/ttLib/tables/table_API_readme.txt

Demonstrates how to obtain the converter class for a given TT table tag using `ttLib.getTableClass`. This function is useful for instantiating the correct table converter class.

```python
from fontTools import ttLib

print(ttLib.getTableClass('glyf'))
```

--------------------------------

### Obtener Estadísticas Principales del Hub API (Bash)

Source: https://context7.com/keepertrading90/rpk-nexus-plan-maestro/llms.txt

Proporciona los KPIs principales para el dashboard central, incluyendo stock total, saturación media de centros de trabajo y días de cobertura teórica. Estos datos se calculan mediante análisis cruzado de la información disponible.

```bash
# Obtener estadísticas principales del Hub
curl -X GET "http://localhost:8000/api/v1/hub_stats"

# Respuesta esperada:
{
    "stock": {
        "total": 2500000,
        "items": 1847
    },
    "saturation": 74.5,
    "cobertura": 12.4
}
```

--------------------------------

### Project Structure Overview (Text)

Source: https://github.com/keepertrading90/rpk-nexus-plan-maestro/blob/main/docs/architecture/RPK_NEXUS_ARCHITECTURE.md

This snippet displays the physical layout of the RPK NEXUS project, following the RPK Agentic Standard. It outlines the main directories and their purposes, including agent rules, Python runtime, backend logic, frontend modules, scripts, and documentation.

```text
Plan Maestro RPK NEXUS/
├── .agent/                  # 🤖 Reglas y workflows del Agente AI
├── _SISTEMA/                # 🚀 Entorno Python Portable (runtime_python)
├── backend/                 # 🧠 Lógica de Negocio y Servidor
│   ├── core/                # Motores de simulación y analítica
│   ├── db/                  # Capa de datos (SQLite + Esquemas)
│   ├── api/                 # Endpoints especializados (Futuros)
│   └── server_nexus.py      # Servidor Central FastAPI (El Corazón)
├── frontend/                # 🎨 Interfaz de Usuario
│   ├── modules/             # Micro-Frontends (Dashboard Paneles)
│   │   ├── pedidos/         # Panel de Gestión de Pedidos de Venta
│   │   ├── simulador/       # Simulador de Producción / Escenarios
│   │   ├── stock/           # Dashboard de Existencias y Valoración
│   │   └── tiempos/         # Planificación de Cargas y Saturation
│   ├── ui/                  # Portal Hub Central (Portal de Usuario)
│   └── assets/              # Recursos globales (Logos, CSS base)
├── scripts/                 # ⚙️ Utilidades de Operaciones (Ops)
│   ├── qa_scanner.py        # Auditoría de sintaxis y patrones RPK
│   ├── ops_sync.py          # Sincronización Golden con GitHub
│   └── sync_nexus.py        # Motor ETL de ingesta diaria
├── docs/                    # 📄 Documentación y Logs
│   ├── architecture/        # [NUEVO] Este documento y diagramas
│   └── logs/                # Historial de cambios y actualizaciones
├── INICIAR_NEXUS.bat        # ⚡ Lanzador Maestro Único
└── README.md                # Guía rápida del proyecto
```

--------------------------------

### Masked Array Division - Default Behavior

Source: https://github.com/keepertrading90/rpk-nexus-plan-maestro/blob/main/_SISTEMA/runtime_python/Lib/site-packages/numpy/ma/README.rst

Illustrates the default behavior of masked array division in NumPy, where input arrays are filled, the operation is performed, and then the mask is set for the result. This can lead to divisions by zero.

```python
import numpy
import maskedarray as ma
from numpy import logical_or

x = ma.array([1,2,3,4],mask=[1,0,0,0], dtype=numpy.float_)
y = ma.array([-1,0,1,2], mask=[0,0,0,1], dtype=numpy.float_)

d1 = x.filled(0) # d1 = array([0., 2., 3., 4.])
d2 = y.filled(1) # array([-1.,  0.,  1.,  1.])
m = ma.mask_or(ma.getmask(x), ma.getmask(y)) # m = array([True,False,False,True])
dm = ma.divide.domain(d1,d2) # array([False,  True, False, False])
result = (d1/d2).view(ma.MaskedArray) # masked_array([-0. inf, 3., 4.])
result._mask = logical_or(m, dm)
```

--------------------------------

### Consultor Lenguaje Natural: Traducción a SQL

Source: https://context7.com/keepertrading90/rpk-nexus-plan-maestro/llms.txt

Traduce preguntas formuladas en lenguaje natural a consultas SQL utilizando la función traducir_a_sql(). Está diseñado para una futura integración con Google Gemini API.

```python
from backend.db.consultor import traducir_a_sql, ejecutar_consulta

# Traducir pregunta a SQL
sql = traducir_a_sql("¿Cuál es el stock total?")
print(sql)
# Salida: "SELECT SUM(cantidad) as Total_Stock, SUM(valor_total) as Valor_Total FROM stock_snapshot"
```

--------------------------------

### Verificar Estado del Sistema API (Bash)

Source: https://context7.com/keepertrading90/rpk-nexus-plan-maestro/llms.txt

Verifica la conectividad y el estado general de la base de datos NEXUS. Retorna información sobre la conexión, la ruta de la base de datos y el conteo de registros en tablas clave.

```bash
# Verificar estado del sistema
curl -X GET "http://localhost:8000/api/v1/status"

# Respuesta esperada:
{
    "status": "online",
    "db_path": "/path/to/rpk_industrial.db",
    "records": {
        "stock": 15420,
        "tiempos": 8750
    },
    "database": "rpk_industrial.db"
}
```

--------------------------------

### Assigning Results Names for Parsed Data Access

Source: https://github.com/keepertrading90/rpk-nexus-plan-maestro/blob/main/_SISTEMA/runtime_python/Lib/site-packages/pyparsing/ai/best_practices.md

Explains how to use results names for accessing parsed data fields, enabling attribute-style access on `ParseResults`. Results names should be valid Python identifiers and replace numeric indexing.

```python
import pyparsing as pp

# Assigning results names using call format
full_name = pp.Word(pp.alphas)("first_name") + pp.Word(pp.alphas)("last_name")

# Example usage
result = full_name.parse_string("John Doe")
print(result.first_name) # Output: John
print(result.last_name)  # Output: Doe
```

--------------------------------

### Initialize Application

Source: https://github.com/keepertrading90/rpk-nexus-plan-maestro/blob/main/frontend/modules/stock/index.html

Calls the initialization function for the application.

```javascript
init();
```

--------------------------------

### Masked Array Division - Without Filling Arrays

Source: https://github.com/keepertrading90/rpk-nexus-plan-maestro/blob/main/_SISTEMA/runtime_python/Lib/site-packages/numpy/ma/README.rst

Demonstrates masked array division without filling the arrays first. This method is faster but may result in floating-point exceptions like 'inf' if divisions by zero occur.

```python
import numpy
import maskedarray as ma
from numpy import logical_or

x = ma.array([1,2,3,4],mask=[1,0,0,0], dtype=numpy.float_)
y = ma.array([-1,0,1,2], mask=[0,0,0,1], dtype=numpy.float_)

d1 = x._data # d1 = array([1., 2., 3., 4.])
d2 = y._data # array([-1.,  0.,  1.,  2.])
dm = ma.divide.domain(d1,d2) # array([False,  True, False, False])
m = ma.mask_or(ma.getmask(x), ma.getmask(y)) # m = array([True,False,False,True])
result = (d1/d2).view(ma.MaskedArray) # masked_array([-1. inf, 3., 2.])
result._mask = logical_or(m, dm)
```

--------------------------------

### Consultor Lenguaje Natural: Ejecución de Consultas

Source: https://context7.com/keepertrading90/rpk-nexus-plan-maestro/llms.txt

Ejecuta una consulta SQL generada por traducir_a_sql() y devuelve los resultados y los nombres de las columnas. La función ejecutar_consulta() permite interactuar con la base de datos.

```python
from backend.db.consultor import ejecutar_consulta

# Ejecutar consulta y obtener resultados
resultados, columnas = ejecutar_consulta(sql)
print(columnas)     # ['Total_Stock', 'Valor_Total']
print(resultados)   # [(2500000.0, 15750000.50)]
```

--------------------------------

### Get TT Table Module by Tag

Source: https://github.com/keepertrading90/rpk-nexus-plan-maestro/blob/main/_SISTEMA/runtime_python/Lib/site-packages/fontTools/ttLib/tables/table_API_readme.txt

Shows how to retrieve the Python module corresponding to a specific TT table tag using `ttLib.getTableModule`. This function helps in accessing the converter module dynamically based on the table tag.

```python
from fontTools import ttLib

print(ttLib.getTableModule('glyf'))
```