Add metrics historical backfill scripts, fix up all new metrics calc queries and add combined script to run

This commit is contained in:
2025-03-31 22:15:41 -04:00
parent 796a2e5d1f
commit 5035dda733
26 changed files with 2592 additions and 958 deletions

View File

@@ -0,0 +1,608 @@
// run-all-updates.js
const path = require('path');
const fs = require('fs');
const { Pool } = require('pg'); // Assuming you use 'pg'
// --- Configuration ---
// Toggle these constants to enable/disable specific steps for testing
const RUN_DAILY_SNAPSHOTS = true;
const RUN_PRODUCT_METRICS = true;
const RUN_PERIODIC_METRICS = true;
// Maximum execution time for the entire sequence (e.g., 90 minutes)
const MAX_EXECUTION_TIME_TOTAL = 90 * 60 * 1000;
// Maximum execution time per individual SQL step (e.g., 30 minutes)
const MAX_EXECUTION_TIME_PER_STEP = 30 * 60 * 1000;
// Query cancellation timeout
const CANCEL_QUERY_AFTER_SECONDS = 5;
// --- End Configuration ---
// Change working directory to script directory
process.chdir(path.dirname(__filename));
// Log script path for debugging
console.log('Script running from:', __dirname);
// Try to load environment variables from multiple locations
const envPaths = [
path.resolve(__dirname, '../..', '.env'), // Two levels up (inventory/.env)
path.resolve(__dirname, '..', '.env'), // One level up (inventory-server/.env)
path.resolve(__dirname, '.env'), // Same directory
'/var/www/html/inventory/.env' // Server absolute path
];
let envLoaded = false;
for (const envPath of envPaths) {
if (fs.existsSync(envPath)) {
console.log(`Loading environment from: ${envPath}`);
require('dotenv').config({ path: envPath });
envLoaded = true;
break;
}
}
if (!envLoaded) {
console.warn('WARNING: Could not find .env file in any of the expected locations.');
console.warn('Checked paths:', envPaths);
}
// --- Database Setup ---
// Make sure we have the required DB credentials
if (!process.env.DB_HOST && !process.env.DATABASE_URL) {
console.error('WARNING: Neither DB_HOST nor DATABASE_URL environment variables found');
}
// Only validate individual parameters if not using connection string
if (!process.env.DATABASE_URL) {
if (!process.env.DB_USER) console.error('WARNING: DB_USER environment variable is missing');
if (!process.env.DB_NAME) console.error('WARNING: DB_NAME environment variable is missing');
// Password must be a string for PostgreSQL SCRAM authentication
if (!process.env.DB_PASSWORD || typeof process.env.DB_PASSWORD !== 'string') {
console.error('WARNING: DB_PASSWORD environment variable is missing or not a string');
}
}
// Configure database connection to match individual scripts
let dbConfig;
// Check if a DATABASE_URL exists (common in production environments)
if (process.env.DATABASE_URL && typeof process.env.DATABASE_URL === 'string') {
console.log('Using DATABASE_URL for connection');
dbConfig = {
connectionString: process.env.DATABASE_URL,
ssl: process.env.DB_SSL === 'true' ? { rejectUnauthorized: false } : false,
// Add performance optimizations
max: 10, // connection pool max size
idleTimeoutMillis: 30000,
connectionTimeoutMillis: 60000,
// Set timeouts for long-running queries
statement_timeout: 1800000, // 30 minutes
query_timeout: 1800000 // 30 minutes
};
} else {
// Use individual connection parameters
dbConfig = {
host: process.env.DB_HOST,
user: process.env.DB_USER,
password: process.env.DB_PASSWORD,
database: process.env.DB_NAME,
port: process.env.DB_PORT || 5432,
ssl: process.env.DB_SSL === 'true',
// Add performance optimizations
max: 10, // connection pool max size
idleTimeoutMillis: 30000,
connectionTimeoutMillis: 60000,
// Set timeouts for long-running queries
statement_timeout: 1800000, // 30 minutes
query_timeout: 1800000 // 30 minutes
};
}
// Try to load from utils DB module as a last resort
try {
if (!process.env.DB_HOST && !process.env.DATABASE_URL) {
console.log('Attempting to load DB config from individual script modules...');
const dbModule = require('./metrics-new/utils/db');
if (dbModule && dbModule.dbConfig) {
console.log('Found DB config in individual script module');
dbConfig = {
...dbModule.dbConfig,
// Add performance optimizations if not present
max: dbModule.dbConfig.max || 10,
idleTimeoutMillis: dbModule.dbConfig.idleTimeoutMillis || 30000,
connectionTimeoutMillis: dbModule.dbConfig.connectionTimeoutMillis || 60000,
statement_timeout: 1800000,
query_timeout: 1800000
};
}
}
} catch (err) {
console.warn('Could not load DB config from individual script modules:', err.message);
}
// Debug log connection info (without password)
console.log('DB Connection Info:', {
connectionString: dbConfig.connectionString ? 'PROVIDED' : undefined,
host: dbConfig.host,
user: dbConfig.user,
database: dbConfig.database,
port: dbConfig.port,
ssl: dbConfig.ssl ? 'ENABLED' : 'DISABLED',
password: (dbConfig.password || dbConfig.connectionString) ? '****' : 'MISSING' // Only show if credentials exist
});
const pool = new Pool(dbConfig);
const getConnection = () => {
return pool.connect();
};
const closePool = () => {
console.log("Closing database connection pool.");
return pool.end();
};
// --- Progress Utilities ---
// Using functions directly instead of globals
const progressUtils = require('./metrics-new/utils/progress'); // Assuming utils/progress.js exports these
// --- State & Cancellation ---
let isCancelled = false;
let currentStep = ''; // Track which step is running for cancellation message
let overallStartTime = null;
let mainTimeoutHandle = null;
let stepTimeoutHandle = null;
async function cancelCalculation(reason = 'cancelled by user') {
if (isCancelled) return; // Prevent multiple cancellations
isCancelled = true;
console.log(`Calculation ${reason}. Attempting to cancel active step: ${currentStep}`);
// Clear timeouts
if (mainTimeoutHandle) clearTimeout(mainTimeoutHandle);
if (stepTimeoutHandle) clearTimeout(stepTimeoutHandle);
// Attempt to cancel the long-running query in Postgres
let conn = null;
try {
console.log(`Attempting to cancel queries running longer than ${CANCEL_QUERY_AFTER_SECONDS} seconds...`);
conn = await getConnection();
const result = await conn.query(`
SELECT pg_cancel_backend(pid)
FROM pg_stat_activity
WHERE query_start < now() - interval '${CANCEL_QUERY_AFTER_SECONDS} seconds'
AND application_name = 'node-metrics-calculator' -- Match specific app name
AND state = 'active' -- Only cancel active queries
AND query NOT LIKE '%pg_cancel_backend%'
AND pid <> pg_backend_pid(); -- Don't cancel self
`);
console.log(`Sent ${result.rowCount} cancellation signal(s).`);
conn.release();
} catch (err) {
console.error('Error during database query cancellation:', err.message);
if (conn) {
try { conn.release(); } catch (e) { console.error("Error releasing cancellation connection", e); }
}
// Proceed with script termination attempt even if DB cancel fails
} finally {
// Update progress to show cancellation
progressUtils.outputProgress({
status: 'cancelled',
operation: `Calculation ${reason} during step: ${currentStep}`,
current: 0, // Reset progress indicators
total: 100,
elapsed: overallStartTime ? progressUtils.formatElapsedTime(overallStartTime) : 'N/A',
remaining: null,
rate: 0,
percentage: '0', // Or keep last known percentage?
timing: {
start_time: overallStartTime ? new Date(overallStartTime).toISOString() : 'N/A',
end_time: new Date().toISOString(),
elapsed_seconds: overallStartTime ? Math.round((Date.now() - overallStartTime) / 1000) : 0
}
});
}
// Note: We don't force exit here anymore. We let the main function's error
// handling catch the cancellation error thrown by executeSqlStep or the timeout.
return {
success: true, // Indicates cancellation was initiated
message: `Calculation ${reason}`
};
}
// Handle SIGINT (Ctrl+C) and SIGTERM (kill) signals
process.on('SIGINT', () => {
console.log('\nReceived SIGINT (Ctrl+C).');
cancelCalculation('cancelled by user (SIGINT)');
// Give cancellation a moment to propagate before force-exiting if needed
setTimeout(() => process.exit(1), 2000);
});
process.on('SIGTERM', () => {
console.log('Received SIGTERM.');
cancelCalculation('cancelled by system (SIGTERM)');
// Give cancellation a moment to propagate before force-exiting if needed
setTimeout(() => process.exit(1), 2000);
});
// Add error handlers for uncaught exceptions/rejections
process.on('uncaughtException', (error) => {
console.error('Uncaught Exception:', error);
// Attempt graceful shutdown/logging if possible, then exit
cancelCalculation('failed due to uncaught exception').finally(() => {
closePool().finally(() => process.exit(1));
});
});
process.on('unhandledRejection', (reason, promise) => {
console.error('Unhandled Rejection at:', promise, 'reason:', reason);
// Attempt graceful shutdown/logging if possible, then exit
cancelCalculation('failed due to unhandled rejection').finally(() => {
closePool().finally(() => process.exit(1));
});
});
// --- Core Logic ---
/**
* Executes a single SQL calculation step.
* @param {object} config - Configuration for the step.
* @param {string} config.name - User-friendly name of the step.
* @param {string} config.sqlFile - Path to the SQL file.
* @param {string} config.historyType - Type identifier for calculate_history.
* @param {string} config.statusModule - Module name for calculate_status.
* @param {object} progress - Progress utility functions.
* @returns {Promise<{success: boolean, message: string, duration: number}>}
*/
async function executeSqlStep(config, progress) {
if (isCancelled) throw new Error(`Calculation skipped step ${config.name} due to prior cancellation.`);
currentStep = config.name; // Update global state
console.log(`\n--- Starting Step: ${config.name} ---`);
const stepStartTime = Date.now();
let connection = null;
let calculateHistoryId = null;
// Set timeout for this specific step
if (stepTimeoutHandle) clearTimeout(stepTimeoutHandle); // Clear previous step's timeout
stepTimeoutHandle = setTimeout(() => {
// Don't exit directly, throw an error to be caught by the main loop
const timeoutError = new Error(`Step "${config.name}" timed out after ${MAX_EXECUTION_TIME_PER_STEP / 1000} seconds.`);
cancelCalculation(`timed out during step: ${config.name}`); // Initiate cancellation process
// The error will likely be thrown before cancelCalculation fully completes,
// but cancelCalculation attempts to stop the query.
// The main catch block will handle cleanup.
}, MAX_EXECUTION_TIME_PER_STEP);
try {
// 1. Read SQL File
const sqlFilePath = path.resolve(__dirname, config.sqlFile);
if (!fs.existsSync(sqlFilePath)) {
throw new Error(`SQL file not found: ${sqlFilePath}`);
}
const sqlQuery = fs.readFileSync(sqlFilePath, 'utf8');
console.log(`Read SQL file: ${config.sqlFile}`);
// Check for potential parameter references that might cause issues
const parameterMatches = sqlQuery.match(/\$\d+(?!\:\:)/g);
if (parameterMatches && parameterMatches.length > 0) {
console.warn(`WARNING: Found ${parameterMatches.length} untyped parameters in SQL: ${parameterMatches.slice(0, 5).join(', ')}${parameterMatches.length > 5 ? '...' : ''}`);
console.warn('These might cause "could not determine data type of parameter" errors.');
}
// 2. Get Database Connection
connection = await getConnection();
console.log("Database connection acquired.");
// 3. Clean up Previous Runs & Create History Record (within a transaction)
await connection.query('BEGIN');
// Ensure calculate_status table exists
await connection.query(`
CREATE TABLE IF NOT EXISTS calculate_status (
module_name TEXT PRIMARY KEY,
last_calculation_timestamp TIMESTAMP WITH TIME ZONE NOT NULL DEFAULT CURRENT_TIMESTAMP
);
`);
// Ensure calculate_history table exists (basic structure)
await connection.query(`
CREATE TABLE IF NOT EXISTS calculate_history (
id SERIAL PRIMARY KEY,
start_time TIMESTAMP WITH TIME ZONE DEFAULT CURRENT_TIMESTAMP,
end_time TIMESTAMP WITH TIME ZONE,
duration_seconds INTEGER,
status TEXT, -- 'running', 'completed', 'failed', 'cancelled'
error_message TEXT,
additional_info JSONB
);
`);
// Mark previous runs of this type as cancelled
await connection.query(`
UPDATE calculate_history
SET
status = 'cancelled',
end_time = NOW(),
duration_seconds = EXTRACT(EPOCH FROM (NOW() - start_time))::INTEGER,
error_message = 'Previous calculation was not completed properly or was superseded.'
WHERE status = 'running' AND additional_info->>'type' = $1::text;
`, [config.historyType]);
// Create history record for this run
const historyResult = await connection.query(`
INSERT INTO calculate_history (status, additional_info)
VALUES ('running', jsonb_build_object('type', $1::text, 'sql_file', $2::text))
RETURNING id;
`, [config.historyType, config.sqlFile]);
calculateHistoryId = historyResult.rows[0].id;
await connection.query('COMMIT');
console.log(`Created history record ID: ${calculateHistoryId}`);
// 4. Initial Progress Update
progress.outputProgress({
status: 'running',
operation: `Starting: ${config.name}`,
current: 0, total: 100,
elapsed: progress.formatElapsedTime(stepStartTime),
remaining: 'Calculating...', rate: 0, percentage: '0',
timing: { start_time: new Date(stepStartTime).toISOString() }
});
// 5. Execute the Main SQL Query
progress.outputProgress({
status: 'running',
operation: `Executing SQL: ${config.name}`,
current: 25, total: 100,
elapsed: progress.formatElapsedTime(stepStartTime),
remaining: 'Executing...', rate: 0, percentage: '25',
timing: { start_time: new Date(stepStartTime).toISOString() }
});
console.log(`Executing SQL for ${config.name}...`);
try {
// Try executing exactly as individual scripts do
console.log('Executing SQL with simple query method...');
await connection.query(sqlQuery);
} catch (sqlError) {
if (sqlError.message.includes('could not determine data type of parameter')) {
console.log('Simple query failed with parameter type error, trying alternative method...');
try {
// Execute with explicit text mode to avoid parameter confusion
await connection.query({
text: sqlQuery,
rowMode: 'text'
});
} catch (altError) {
console.error('Alternative execution method also failed:', altError.message);
throw altError; // Re-throw the alternative error
}
} else {
console.error('SQL Execution Error:', sqlError.message);
if (sqlError.position) {
// If the error has a position, try to show the relevant part of the SQL query
const position = parseInt(sqlError.position, 10);
const startPos = Math.max(0, position - 100);
const endPos = Math.min(sqlQuery.length, position + 100);
console.error('SQL Error Context:');
console.error('...' + sqlQuery.substring(startPos, position) + ' [ERROR HERE] ' + sqlQuery.substring(position, endPos) + '...');
}
throw sqlError; // Re-throw to be caught by the main try/catch
}
}
// Check for cancellation immediately after query finishes
if (isCancelled) throw new Error(`Calculation cancelled during SQL execution for ${config.name}`);
console.log(`SQL execution finished for ${config.name}.`);
// 6. Update Status & History (within a transaction)
await connection.query('BEGIN');
await connection.query(`
INSERT INTO calculate_status (module_name, last_calculation_timestamp)
VALUES ($1::text, NOW())
ON CONFLICT (module_name) DO UPDATE
SET last_calculation_timestamp = EXCLUDED.last_calculation_timestamp;
`, [config.statusModule]);
const stepDuration = Math.round((Date.now() - stepStartTime) / 1000);
await connection.query(`
UPDATE calculate_history
SET
end_time = NOW(),
duration_seconds = $1::integer,
status = 'completed'
WHERE id = $2::integer;
`, [stepDuration, calculateHistoryId]);
await connection.query('COMMIT');
// 7. Final Progress Update for Step
progress.outputProgress({
status: 'complete',
operation: `Completed: ${config.name}`,
current: 100, total: 100,
elapsed: progress.formatElapsedTime(stepStartTime),
remaining: '0s', rate: 0, percentage: '100',
timing: {
start_time: new Date(stepStartTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: stepDuration
}
});
console.log(`--- Finished Step: ${config.name} (Duration: ${progress.formatElapsedTime(stepStartTime)}) ---`);
return {
success: true,
message: `${config.name} completed successfully`,
duration: stepDuration
};
} catch (error) {
clearTimeout(stepTimeoutHandle); // Clear timeout on error
const errorEndTime = Date.now();
const errorDuration = Math.round((errorEndTime - stepStartTime) / 1000);
const finalStatus = isCancelled ? 'cancelled' : 'failed';
const errorMessage = error.message || 'Unknown error';
console.error(`--- ERROR in Step: ${config.name} ---`);
console.error(error); // Log the full error
console.error(`------------------------------------`);
// Update history with error/cancellation status
if (connection && calculateHistoryId) {
try {
// Use a separate transaction for error logging
await connection.query('ROLLBACK'); // Rollback any partial transaction from try block
await connection.query('BEGIN');
await connection.query(`
UPDATE calculate_history
SET
end_time = NOW(),
duration_seconds = $1::integer,
status = $2::text,
error_message = $3::text
WHERE id = $4::integer;
`, [errorDuration, finalStatus, errorMessage.substring(0, 1000), calculateHistoryId]); // Limit error message size
await connection.query('COMMIT');
console.log(`Updated history record ID ${calculateHistoryId} with status: ${finalStatus}`);
} catch (historyError) {
console.error("FATAL: Failed to update history record on error:", historyError);
// Cannot rollback here if already rolled back or commit failed
}
} else {
console.warn("Could not update history record on error (no connection or history ID).");
}
// Update progress file with error/cancellation
progress.outputProgress({
status: finalStatus,
operation: `Error in ${config.name}: ${errorMessage.split('\n')[0]}`, // Show first line of error
current: 50, total: 100, // Indicate partial completion
elapsed: progress.formatElapsedTime(stepStartTime),
remaining: null, rate: 0, percentage: '50',
timing: {
start_time: new Date(stepStartTime).toISOString(),
end_time: new Date(errorEndTime).toISOString(),
elapsed_seconds: errorDuration
}
});
// Rethrow the error to be caught by the main runCalculations function
throw error; // Add context if needed: new Error(`Step ${config.name} failed: ${errorMessage}`)
} finally {
clearTimeout(stepTimeoutHandle); // Ensure timeout is cleared
currentStep = ''; // Reset current step
if (connection) {
try {
await connection.release();
console.log("Database connection released.");
} catch (releaseError) {
console.error("Error releasing database connection:", releaseError);
}
}
}
}
/**
* Main function to run all calculation steps sequentially.
*/
async function runAllCalculations() {
overallStartTime = Date.now();
isCancelled = false; // Reset cancellation flag at start
// Overall timeout for the entire script
mainTimeoutHandle = setTimeout(() => {
console.error(`--- OVERALL TIMEOUT REACHED (${MAX_EXECUTION_TIME_TOTAL / 1000}s) ---`);
cancelCalculation(`overall timeout reached`);
// The process should exit via the unhandled rejection/exception handlers
// or the SIGTERM/SIGINT handlers after cancellation attempt.
}, MAX_EXECUTION_TIME_TOTAL);
const steps = [
{
run: RUN_DAILY_SNAPSHOTS,
name: 'Daily Snapshots Update',
sqlFile: 'metrics-new/update_daily_snapshots.sql',
historyType: 'daily_snapshots',
statusModule: 'daily_snapshots'
},
{
run: RUN_PRODUCT_METRICS,
name: 'Product Metrics Update',
sqlFile: 'metrics-new/update_product_metrics.sql', // ASSUMING the initial population is now part of a regular update
historyType: 'product_metrics',
statusModule: 'product_metrics'
},
{
run: RUN_PERIODIC_METRICS,
name: 'Periodic Metrics Update',
sqlFile: 'metrics-new/update_periodic_metrics.sql',
historyType: 'periodic_metrics',
statusModule: 'periodic_metrics'
}
];
let overallSuccess = true;
try {
for (const step of steps) {
if (step.run) {
if (isCancelled) {
console.log(`Skipping step "${step.name}" due to cancellation.`);
overallSuccess = false; // Mark as not fully successful if steps are skipped due to cancel
continue; // Skip to next step
}
// Pass the progress utilities to the step executor
await executeSqlStep(step, progressUtils);
} else {
console.log(`Skipping step "${step.name}" (disabled by configuration).`);
}
}
// If we finished naturally (no errors thrown out)
clearTimeout(mainTimeoutHandle); // Clear the main timeout
if (isCancelled) {
console.log("\n--- Calculation finished with cancellation ---");
overallSuccess = false;
} else {
console.log("\n--- All enabled calculations finished successfully ---");
progressUtils.clearProgress(); // Clear progress only on full success
}
} catch (error) {
clearTimeout(mainTimeoutHandle); // Clear the main timeout
console.error("\n--- SCRIPT EXECUTION FAILED ---");
// Error details were already logged by executeSqlStep or global handlers
overallSuccess = false;
// Don't re-log the error here unless adding context
// console.error("Overall failure reason:", error.message);
} finally {
await closePool();
console.log(`Total execution time: ${progressUtils.formatElapsedTime(overallStartTime)}`);
process.exit(overallSuccess ? 0 : 1);
}
}
// --- Script Execution ---
if (require.main === module) {
runAllCalculations();
} else {
// Export functions if needed as a module (e.g., for testing or API)
module.exports = {
runAllCalculations,
cancelCalculation,
// Expose individual steps if useful, wrapping them slightly
runDailySnapshots: () => executeSqlStep({ name: 'Daily Snapshots Update', sqlFile: 'update_daily_snapshots.sql', historyType: 'daily_snapshots', statusModule: 'daily_snapshots' }, progressUtils),
runProductMetrics: () => executeSqlStep({ name: 'Product Metrics Update', sqlFile: 'update_product_metrics.sql', historyType: 'product_metrics', statusModule: 'product_metrics' }, progressUtils),
runPeriodicMetrics: () => executeSqlStep({ name: 'Periodic Metrics Update', sqlFile: 'update_periodic_metrics.sql', historyType: 'periodic_metrics', statusModule: 'periodic_metrics' }, progressUtils),
getProgress: progressUtils.getProgress
};
}

View File

@@ -1,558 +0,0 @@
const path = require('path');
// Change working directory to script directory
process.chdir(path.dirname(__filename));
require('dotenv').config({ path: path.resolve(__dirname, '..', '.env') });
// Configuration flags for controlling which metrics to calculate
// Set to 1 to skip the corresponding calculation, 0 to run it
const SKIP_PRODUCT_METRICS = 0;
const SKIP_TIME_AGGREGATES = 0;
const SKIP_FINANCIAL_METRICS = 0;
const SKIP_VENDOR_METRICS = 0;
const SKIP_CATEGORY_METRICS = 0;
const SKIP_BRAND_METRICS = 0;
const SKIP_SALES_FORECASTS = 0;
// Add error handler for uncaught exceptions
process.on('uncaughtException', (error) => {
console.error('Uncaught Exception:', error);
process.exit(1);
});
// Add error handler for unhandled promise rejections
process.on('unhandledRejection', (reason, promise) => {
console.error('Unhandled Rejection at:', promise, 'reason:', reason);
process.exit(1);
});
const progress = require('./metrics/utils/progress');
console.log('Progress module loaded:', {
modulePath: require.resolve('./metrics/utils/progress'),
exports: Object.keys(progress),
currentDir: process.cwd(),
scriptDir: __dirname
});
// Store progress functions in global scope to ensure availability
global.formatElapsedTime = progress.formatElapsedTime;
global.estimateRemaining = progress.estimateRemaining;
global.calculateRate = progress.calculateRate;
global.outputProgress = progress.outputProgress;
global.clearProgress = progress.clearProgress;
global.getProgress = progress.getProgress;
global.logError = progress.logError;
// List of temporary tables used in the calculation process
const TEMP_TABLES = [
'temp_revenue_ranks',
'temp_sales_metrics',
'temp_purchase_metrics',
'temp_product_metrics',
'temp_vendor_metrics',
'temp_category_metrics',
'temp_brand_metrics',
'temp_forecast_dates',
'temp_daily_sales',
'temp_product_stats',
'temp_category_sales',
'temp_category_stats',
'temp_beginning_inventory',
'temp_monthly_inventory'
];
// Add cleanup function for temporary tables
async function cleanupTemporaryTables(connection) {
try {
// Drop each temporary table if it exists
for (const table of TEMP_TABLES) {
await connection.query(`DROP TABLE IF EXISTS ${table}`);
}
} catch (err) {
console.error('Error cleaning up temporary tables:', err);
}
}
const { getConnection, closePool } = require('./metrics/utils/db');
const calculateProductMetrics = require('./metrics/product-metrics');
const calculateTimeAggregates = require('./metrics/time-aggregates');
const calculateFinancialMetrics = require('./metrics/financial-metrics');
const calculateVendorMetrics = require('./metrics/vendor-metrics');
const calculateCategoryMetrics = require('./metrics/category-metrics');
const calculateBrandMetrics = require('./metrics/brand-metrics');
const calculateSalesForecasts = require('./metrics/sales-forecasts');
// Add cancel handler
let isCancelled = false;
function cancelCalculation() {
isCancelled = true;
console.log('Calculation has been cancelled by user');
// Force-terminate any query that's been running for more than 5 seconds
try {
const connection = getConnection();
connection.then(async (conn) => {
try {
// Identify and terminate long-running queries from our application
await conn.query(`
SELECT pg_cancel_backend(pid)
FROM pg_stat_activity
WHERE query_start < now() - interval '5 seconds'
AND application_name LIKE '%node%'
AND query NOT LIKE '%pg_cancel_backend%'
`);
// Clean up any temporary tables
await cleanupTemporaryTables(conn);
// Release connection
conn.release();
} catch (err) {
console.error('Error during force cancellation:', err);
conn.release();
}
}).catch(err => {
console.error('Could not get connection for cancellation:', err);
});
} catch (err) {
console.error('Failed to terminate running queries:', err);
}
return {
success: true,
message: 'Calculation has been cancelled'
};
}
// Handle SIGTERM signal for cancellation
process.on('SIGTERM', cancelCalculation);
// Update the main calculation function to use the new modular structure
async function calculateMetrics() {
let connection;
const startTime = Date.now();
let processedProducts = 0;
let processedOrders = 0;
let processedPurchaseOrders = 0;
let totalProducts = 0;
let totalOrders = 0;
let totalPurchaseOrders = 0;
let calculateHistoryId;
// Set a maximum execution time (30 minutes)
const MAX_EXECUTION_TIME = 30 * 60 * 1000;
const timeout = setTimeout(() => {
console.error(`Calculation timed out after ${MAX_EXECUTION_TIME/1000} seconds, forcing termination`);
// Call cancel and force exit
cancelCalculation();
process.exit(1);
}, MAX_EXECUTION_TIME);
try {
// Clean up any previously running calculations
connection = await getConnection();
await connection.query(`
UPDATE calculate_history
SET
status = 'cancelled',
end_time = NOW(),
duration_seconds = EXTRACT(EPOCH FROM (NOW() - start_time))::INTEGER,
error_message = 'Previous calculation was not completed properly'
WHERE status = 'running'
`);
// Get counts from all relevant tables
const [productCountResult, orderCountResult, poCountResult] = await Promise.all([
connection.query('SELECT COUNT(*) as total FROM products'),
connection.query('SELECT COUNT(*) as total FROM orders'),
connection.query('SELECT COUNT(*) as total FROM purchase_orders')
]);
totalProducts = parseInt(productCountResult.rows[0].total);
totalOrders = parseInt(orderCountResult.rows[0].total);
totalPurchaseOrders = parseInt(poCountResult.rows[0].total);
// Create history record for this calculation
const historyResult = await connection.query(`
INSERT INTO calculate_history (
start_time,
status,
total_products,
total_orders,
total_purchase_orders,
additional_info
) VALUES (
NOW(),
'running',
$1,
$2,
$3,
jsonb_build_object(
'skip_product_metrics', ($4::int > 0),
'skip_time_aggregates', ($5::int > 0),
'skip_financial_metrics', ($6::int > 0),
'skip_vendor_metrics', ($7::int > 0),
'skip_category_metrics', ($8::int > 0),
'skip_brand_metrics', ($9::int > 0),
'skip_sales_forecasts', ($10::int > 0)
)
) RETURNING id
`, [
totalProducts,
totalOrders,
totalPurchaseOrders,
SKIP_PRODUCT_METRICS,
SKIP_TIME_AGGREGATES,
SKIP_FINANCIAL_METRICS,
SKIP_VENDOR_METRICS,
SKIP_CATEGORY_METRICS,
SKIP_BRAND_METRICS,
SKIP_SALES_FORECASTS
]);
calculateHistoryId = historyResult.rows[0].id;
// Add debug logging for the progress functions
console.log('Debug - Progress functions:', {
formatElapsedTime: typeof global.formatElapsedTime,
estimateRemaining: typeof global.estimateRemaining,
calculateRate: typeof global.calculateRate,
startTime: startTime
});
try {
const elapsed = global.formatElapsedTime(startTime);
console.log('Debug - formatElapsedTime test successful:', elapsed);
} catch (err) {
console.error('Debug - Error testing formatElapsedTime:', err);
throw err;
}
// Release the connection before getting a new one
connection.release();
isCancelled = false;
connection = await getConnection();
try {
global.outputProgress({
status: 'running',
operation: 'Starting metrics calculation',
current: 0,
total: 100,
elapsed: '0s',
remaining: 'Calculating...',
rate: 0,
percentage: '0',
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
// Update progress periodically
const updateProgress = async (products = null, orders = null, purchaseOrders = null) => {
// Ensure all values are valid numbers or default to previous value
if (products !== null) processedProducts = Number(products) || processedProducts || 0;
if (orders !== null) processedOrders = Number(orders) || processedOrders || 0;
if (purchaseOrders !== null) processedPurchaseOrders = Number(purchaseOrders) || processedPurchaseOrders || 0;
// Ensure we never send NaN to the database
const safeProducts = Number(processedProducts) || 0;
const safeOrders = Number(processedOrders) || 0;
const safePurchaseOrders = Number(processedPurchaseOrders) || 0;
await connection.query(`
UPDATE calculate_history
SET
processed_products = $1,
processed_orders = $2,
processed_purchase_orders = $3
WHERE id = $4
`, [safeProducts, safeOrders, safePurchaseOrders, calculateHistoryId]);
};
// Helper function to ensure valid progress numbers
const ensureValidProgress = (current, total) => ({
current: Number(current) || 0,
total: Number(total) || 1, // Default to 1 to avoid division by zero
percentage: (((Number(current) || 0) / (Number(total) || 1)) * 100).toFixed(1)
});
// Initial progress
const initialProgress = ensureValidProgress(0, totalProducts);
global.outputProgress({
status: 'running',
operation: 'Starting metrics calculation',
current: initialProgress.current,
total: initialProgress.total,
elapsed: '0s',
remaining: 'Calculating...',
rate: 0,
percentage: initialProgress.percentage,
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
if (!SKIP_PRODUCT_METRICS) {
const result = await calculateProductMetrics(startTime, totalProducts);
await updateProgress(result.processedProducts, result.processedOrders, result.processedPurchaseOrders);
if (!result.success) {
throw new Error('Product metrics calculation failed');
}
} else {
console.log('Skipping product metrics calculation...');
processedProducts = Math.floor(totalProducts * 0.6);
await updateProgress(processedProducts);
global.outputProgress({
status: 'running',
operation: 'Skipping product metrics calculation',
current: processedProducts,
total: totalProducts,
elapsed: global.formatElapsedTime(startTime),
remaining: global.estimateRemaining(startTime, processedProducts, totalProducts),
rate: global.calculateRate(startTime, processedProducts),
percentage: '60',
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
}
// Calculate time-based aggregates
if (!SKIP_TIME_AGGREGATES) {
const result = await calculateTimeAggregates(startTime, totalProducts, processedProducts);
await updateProgress(result.processedProducts, result.processedOrders, result.processedPurchaseOrders);
if (!result.success) {
throw new Error('Time aggregates calculation failed');
}
} else {
console.log('Skipping time aggregates calculation');
}
// Calculate financial metrics
if (!SKIP_FINANCIAL_METRICS) {
const result = await calculateFinancialMetrics(startTime, totalProducts, processedProducts);
await updateProgress(result.processedProducts, result.processedOrders, result.processedPurchaseOrders);
if (!result.success) {
throw new Error('Financial metrics calculation failed');
}
} else {
console.log('Skipping financial metrics calculation');
}
// Calculate vendor metrics
if (!SKIP_VENDOR_METRICS) {
const result = await calculateVendorMetrics(startTime, totalProducts, processedProducts);
await updateProgress(result.processedProducts, result.processedOrders, result.processedPurchaseOrders);
if (!result.success) {
throw new Error('Vendor metrics calculation failed');
}
} else {
console.log('Skipping vendor metrics calculation');
}
// Calculate category metrics
if (!SKIP_CATEGORY_METRICS) {
const result = await calculateCategoryMetrics(startTime, totalProducts, processedProducts);
await updateProgress(result.processedProducts, result.processedOrders, result.processedPurchaseOrders);
if (!result.success) {
throw new Error('Category metrics calculation failed');
}
} else {
console.log('Skipping category metrics calculation');
}
// Calculate brand metrics
if (!SKIP_BRAND_METRICS) {
const result = await calculateBrandMetrics(startTime, totalProducts, processedProducts);
await updateProgress(result.processedProducts, result.processedOrders, result.processedPurchaseOrders);
if (!result.success) {
throw new Error('Brand metrics calculation failed');
}
} else {
console.log('Skipping brand metrics calculation');
}
// Calculate sales forecasts
if (!SKIP_SALES_FORECASTS) {
const result = await calculateSalesForecasts(startTime, totalProducts, processedProducts);
await updateProgress(result.processedProducts, result.processedOrders, result.processedPurchaseOrders);
if (!result.success) {
throw new Error('Sales forecasts calculation failed');
}
} else {
console.log('Skipping sales forecasts calculation');
}
// Final progress update with guaranteed valid numbers
const finalProgress = ensureValidProgress(totalProducts, totalProducts);
// Final success message
outputProgress({
status: 'complete',
operation: 'Metrics calculation complete',
current: finalProgress.current,
total: finalProgress.total,
elapsed: global.formatElapsedTime(startTime),
remaining: '0s',
rate: global.calculateRate(startTime, finalProgress.current),
percentage: '100',
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
// Ensure all values are valid numbers before final update
const finalStats = {
processedProducts: Number(processedProducts) || 0,
processedOrders: Number(processedOrders) || 0,
processedPurchaseOrders: Number(processedPurchaseOrders) || 0
};
// Update history with completion
await connection.query(`
UPDATE calculate_history
SET
end_time = NOW(),
duration_seconds = $1,
processed_products = $2,
processed_orders = $3,
processed_purchase_orders = $4,
status = 'completed'
WHERE id = $5
`, [Math.round((Date.now() - startTime) / 1000),
finalStats.processedProducts,
finalStats.processedOrders,
finalStats.processedPurchaseOrders,
calculateHistoryId]);
// Clear progress file on successful completion
global.clearProgress();
return {
success: true,
message: 'Calculation completed successfully',
duration: Math.round((Date.now() - startTime) / 1000)
};
} catch (error) {
const endTime = Date.now();
const totalElapsedSeconds = Math.round((endTime - startTime) / 1000);
// Update history with error
await connection.query(`
UPDATE calculate_history
SET
end_time = NOW(),
duration_seconds = $1,
processed_products = $2,
processed_orders = $3,
processed_purchase_orders = $4,
status = $5,
error_message = $6
WHERE id = $7
`, [
totalElapsedSeconds,
processedProducts || 0, // Ensure we have a valid number
processedOrders || 0, // Ensure we have a valid number
processedPurchaseOrders || 0, // Ensure we have a valid number
isCancelled ? 'cancelled' : 'failed',
error.message,
calculateHistoryId
]);
if (isCancelled) {
global.outputProgress({
status: 'cancelled',
operation: 'Calculation cancelled',
current: processedProducts,
total: totalProducts || 0,
elapsed: global.formatElapsedTime(startTime),
remaining: null,
rate: global.calculateRate(startTime, processedProducts),
percentage: ((processedProducts / (totalProducts || 1)) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
} else {
global.outputProgress({
status: 'error',
operation: 'Error: ' + error.message,
current: processedProducts,
total: totalProducts || 0,
elapsed: global.formatElapsedTime(startTime),
remaining: null,
rate: global.calculateRate(startTime, processedProducts),
percentage: ((processedProducts / (totalProducts || 1)) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
}
throw error;
} finally {
// Clear the timeout to prevent forced termination
clearTimeout(timeout);
// Always clean up and release connection
if (connection) {
try {
await cleanupTemporaryTables(connection);
connection.release();
} catch (err) {
console.error('Error in final cleanup:', err);
}
}
}
} catch (error) {
console.error('Error in metrics calculation', error);
try {
if (connection) {
await connection.query(`
UPDATE calculate_history
SET
status = 'failed',
end_time = NOW(),
duration_seconds = EXTRACT(EPOCH FROM (NOW() - start_time))::INTEGER,
error_message = $1
WHERE id = $2
`, [error.message.substring(0, 500), calculateHistoryId]);
}
} catch (updateError) {
console.error('Error updating calculation history:', updateError);
}
throw error;
}
}
// Export as a module with all necessary functions
module.exports = {
calculateMetrics,
cancelCalculation,
getProgress: global.getProgress
};
// Run directly if called from command line
if (require.main === module) {
calculateMetrics().catch(error => {
if (!error.message.includes('Operation cancelled')) {
console.error('Error:', error);
}
process.exit(1);
});
}

View File

@@ -0,0 +1,426 @@
const path = require('path');
const fs = require('fs');
const progress = require('./utils/progress'); // Assuming progress utils are here
const { getConnection, closePool } = require('./utils/db'); // Assuming db utils are here
const os = require('os'); // For detecting number of CPU cores
// --- Configuration ---
const BATCH_SIZE_DAYS = 1; // Process 1 day per database function call
const SQL_FUNCTION_FILE = path.resolve(__dirname, 'backfill_historical_snapshots.sql'); // Correct path
const LOG_PROGRESS_INTERVAL_MS = 5000; // Update console progress roughly every 5 seconds
const HISTORY_TYPE = 'backfill_snapshots'; // Identifier for history table
const MAX_WORKERS = Math.max(1, Math.floor(os.cpus().length / 2)); // Use half of available CPU cores
const USE_PARALLEL = false; // Set to true to enable parallel processing
const PG_STATEMENT_TIMEOUT_MS = 1800000; // 30 minutes max per query
// --- Cancellation Handling ---
let isCancelled = false;
let runningQueryPromise = null; // To potentially track the active query
function requestCancellation() {
if (!isCancelled) {
isCancelled = true;
console.warn('\nCancellation requested. Finishing current batch then stopping...');
// Note: We are NOT forcefully cancelling the backend query anymore.
}
}
process.on('SIGINT', requestCancellation); // Handle Ctrl+C
process.on('SIGTERM', requestCancellation); // Handle termination signals
// --- Main Backfill Function ---
async function backfillSnapshots(cmdStartDate, cmdEndDate, cmdStartBatch = 1) {
let connection;
const overallStartTime = Date.now();
let calculateHistoryId = null;
let processedDaysTotal = 0; // Track total days processed across all batches executed in this run
let currentBatchNum = cmdStartBatch > 0 ? cmdStartBatch : 1;
let totalBatches = 0; // Initialize totalBatches
let totalDays = 0; // Initialize totalDays
console.log(`Starting snapshot backfill process...`);
console.log(`SQL Function definition file: ${SQL_FUNCTION_FILE}`);
if (!fs.existsSync(SQL_FUNCTION_FILE)) {
console.error(`FATAL: SQL file not found at ${SQL_FUNCTION_FILE}`);
process.exit(1); // Exit early if file doesn't exist
}
try {
// Set up a connection with higher memory limits
connection = await getConnection({
// Add performance-related settings
application_name: 'backfill_snapshots',
statement_timeout: PG_STATEMENT_TIMEOUT_MS, // 30 min timeout per statement
// These parameters may need to be configured in your database:
// work_mem: '1GB',
// maintenance_work_mem: '2GB',
// temp_buffers: '1GB',
});
console.log('Database connection acquired.');
// --- Ensure Function Exists ---
console.log('Ensuring database function is up-to-date...');
try {
const sqlFunctionDef = fs.readFileSync(SQL_FUNCTION_FILE, 'utf8');
if (!sqlFunctionDef.includes('CREATE OR REPLACE FUNCTION backfill_daily_snapshots_range_final')) {
throw new Error(`SQL file ${SQL_FUNCTION_FILE} does not seem to contain the function definition.`);
}
await connection.query(sqlFunctionDef); // Execute the whole file
console.log('Database function `backfill_daily_snapshots_range_final` created/updated.');
// Add performance query hints to the database
await connection.query(`
-- Analyze tables for better query planning
ANALYZE public.products;
ANALYZE public.imported_daily_inventory;
ANALYZE public.imported_product_stat_history;
ANALYZE public.daily_product_snapshots;
ANALYZE public.imported_product_current_prices;
`).catch(err => {
// Non-fatal if analyze fails
console.warn('Failed to analyze tables (non-fatal):', err.message);
});
} catch (err) {
console.error(`Error processing SQL function file ${SQL_FUNCTION_FILE}:`, err);
throw new Error(`Failed to create or replace DB function: ${err.message}`);
}
// --- Prepare History Record ---
console.log('Preparing calculation history record...');
// Ensure history table exists (optional, could be done elsewhere)
await connection.query(`
CREATE TABLE IF NOT EXISTS public.calculate_history (
id SERIAL PRIMARY KEY,
start_time TIMESTAMPTZ NOT NULL DEFAULT NOW(),
end_time TIMESTAMPTZ,
duration_seconds INTEGER,
status VARCHAR(20) NOT NULL, -- e.g., 'running', 'completed', 'failed', 'cancelled'
error_message TEXT,
additional_info JSONB -- Store type, file, batch info etc.
);
`);
// Mark previous runs of this type as potentially failed if they were left 'running'
await connection.query(`
UPDATE public.calculate_history
SET status = 'failed', error_message = 'Interrupted by new run.'
WHERE status = 'running' AND additional_info->>'type' = $1;
`, [HISTORY_TYPE]);
// Create new history record
const historyResult = await connection.query(`
INSERT INTO public.calculate_history (start_time, status, additional_info)
VALUES (NOW(), 'running', jsonb_build_object('type', $1::text, 'sql_file', $2::text, 'start_batch', $3::integer))
RETURNING id;
`, [HISTORY_TYPE, path.basename(SQL_FUNCTION_FILE), cmdStartBatch]);
calculateHistoryId = historyResult.rows[0].id;
console.log(`Calculation history record created with ID: ${calculateHistoryId}`);
// --- Determine Date Range ---
console.log('Determining date range...');
let effectiveStartDate, effectiveEndDate;
// Use command-line dates if provided, otherwise query DB
if (cmdStartDate) {
effectiveStartDate = cmdStartDate;
} else {
const minDateResult = await connection.query(`
SELECT LEAST(
COALESCE((SELECT MIN(date) FROM public.imported_daily_inventory WHERE date > '1970-01-01'), CURRENT_DATE),
COALESCE((SELECT MIN(date) FROM public.imported_product_stat_history WHERE date > '1970-01-01'), CURRENT_DATE)
)::date as min_date;
`);
effectiveStartDate = minDateResult.rows[0]?.min_date || new Date().toISOString().split('T')[0]; // Fallback
console.log(`Auto-detected start date: ${effectiveStartDate}`);
}
if (cmdEndDate) {
effectiveEndDate = cmdEndDate;
} else {
const maxDateResult = await connection.query(`
SELECT GREATEST(
COALESCE((SELECT MAX(date) FROM public.imported_daily_inventory WHERE date < CURRENT_DATE), '1970-01-01'::date),
COALESCE((SELECT MAX(date) FROM public.imported_product_stat_history WHERE date < CURRENT_DATE), '1970-01-01'::date)
)::date as max_date;
`);
// Ensure end date is not today or in the future
effectiveEndDate = maxDateResult.rows[0]?.max_date || new Date(Date.now() - 86400000).toISOString().split('T')[0]; // Default yesterday
if (new Date(effectiveEndDate) >= new Date(new Date().toISOString().split('T')[0])) {
effectiveEndDate = new Date(Date.now() - 86400000).toISOString().split('T')[0]; // Set to yesterday if >= today
}
console.log(`Auto-detected end date: ${effectiveEndDate}`);
}
// Validate dates
const dStart = new Date(effectiveStartDate);
const dEnd = new Date(effectiveEndDate);
if (isNaN(dStart.getTime()) || isNaN(dEnd.getTime()) || dStart > dEnd) {
throw new Error(`Invalid date range: Start "${effectiveStartDate}", End "${effectiveEndDate}"`);
}
// --- Batch Processing ---
totalDays = Math.ceil((dEnd - dStart) / (1000 * 60 * 60 * 24)) + 1; // Inclusive
totalBatches = Math.ceil(totalDays / BATCH_SIZE_DAYS);
console.log(`Target Date Range: ${effectiveStartDate} to ${effectiveEndDate} (${totalDays} days)`);
console.log(`Total Batches: ${totalBatches} (Batch Size: ${BATCH_SIZE_DAYS} days)`);
console.log(`Starting from Batch: ${currentBatchNum}`);
// Initial progress update
progress.outputProgress({
status: 'running',
operation: 'Starting Batch Processing',
currentBatch: currentBatchNum,
totalBatches: totalBatches,
totalDays: totalDays,
elapsed: '0s',
remaining: 'Calculating...',
rate: 0,
historyId: calculateHistoryId // Include history ID in the object
});
while (currentBatchNum <= totalBatches && !isCancelled) {
const batchOffset = (currentBatchNum - 1) * BATCH_SIZE_DAYS;
const batchStartDate = new Date(dStart);
batchStartDate.setDate(dStart.getDate() + batchOffset);
const batchEndDate = new Date(batchStartDate);
batchEndDate.setDate(batchStartDate.getDate() + BATCH_SIZE_DAYS - 1);
// Clamp batch end date to the overall effective end date
if (batchEndDate > dEnd) {
batchEndDate.setTime(dEnd.getTime());
}
const batchStartDateStr = batchStartDate.toISOString().split('T')[0];
const batchEndDateStr = batchEndDate.toISOString().split('T')[0];
const batchStartTime = Date.now();
console.log(`\n--- Processing Batch ${currentBatchNum} / ${totalBatches} ---`);
console.log(` Dates: ${batchStartDateStr} to ${batchEndDateStr}`);
// Execute the function for the batch
try {
progress.outputProgress({
status: 'running',
operation: `Executing DB function for batch ${currentBatchNum}...`,
currentBatch: currentBatchNum,
totalBatches: totalBatches,
totalDays: totalDays,
elapsed: progress.formatElapsedTime(overallStartTime),
remaining: 'Executing...',
rate: 0,
historyId: calculateHistoryId
});
// Performance improvement: Add batch processing hint
await connection.query('SET LOCAL enable_parallel_append = on; SET LOCAL enable_parallel_hash = on; SET LOCAL max_parallel_workers_per_gather = 4;');
// Store promise in case we need to try and cancel (though not implemented forcefully)
runningQueryPromise = connection.query(
`SELECT backfill_daily_snapshots_range_final($1::date, $2::date);`,
[batchStartDateStr, batchEndDateStr]
);
await runningQueryPromise; // Wait for the function call to complete
runningQueryPromise = null; // Clear the promise
const batchDurationMs = Date.now() - batchStartTime;
const daysInThisBatch = Math.ceil((batchEndDate - batchStartDate) / (1000 * 60 * 60 * 24)) + 1;
processedDaysTotal += daysInThisBatch;
console.log(` Batch ${currentBatchNum} completed in ${progress.formatElapsedTime(batchStartTime)}.`);
// --- Update Progress & History ---
const overallElapsedSec = Math.round((Date.now() - overallStartTime) / 1000);
progress.outputProgress({
status: 'running',
operation: `Completed batch ${currentBatchNum}`,
currentBatch: currentBatchNum,
totalBatches: totalBatches,
totalDays: totalDays,
processedDays: processedDaysTotal,
elapsed: progress.formatElapsedTime(overallStartTime),
remaining: progress.estimateRemaining(overallStartTime, processedDaysTotal, totalDays),
rate: progress.calculateRate(overallStartTime, processedDaysTotal),
batchDuration: progress.formatElapsedTime(batchStartTime),
historyId: calculateHistoryId
});
// Save checkpoint in history
await connection.query(`
UPDATE public.calculate_history
SET additional_info = jsonb_set(additional_info, '{last_completed_batch}', $1::jsonb)
|| jsonb_build_object('last_processed_date', $2::text)
WHERE id = $3::integer;
`, [JSON.stringify(currentBatchNum), batchEndDateStr, calculateHistoryId]);
} catch (batchError) {
console.error(`\n--- ERROR in Batch ${currentBatchNum} (${batchStartDateStr} to ${batchEndDateStr}) ---`);
console.error(' Database Error:', batchError.message);
console.error(' DB Error Code:', batchError.code);
// Log detailed error to history and re-throw to stop the process
await connection.query(`
UPDATE public.calculate_history
SET status = 'failed',
end_time = NOW(),
duration_seconds = $1::integer,
error_message = $2::text,
additional_info = additional_info || jsonb_build_object('failed_batch', $3::integer, 'failed_date_range', $4::text)
WHERE id = $5::integer;
`, [
Math.round((Date.now() - overallStartTime) / 1000),
`Batch ${currentBatchNum} failed: ${batchError.message} (Code: ${batchError.code || 'N/A'})`,
currentBatchNum,
`${batchStartDateStr} to ${batchEndDateStr}`,
calculateHistoryId
]);
throw batchError; // Stop execution
}
currentBatchNum++;
// Optional delay between batches
// await new Promise(resolve => setTimeout(resolve, 500));
} // End while loop
// --- Final Outcome ---
const finalStatus = isCancelled ? 'cancelled' : 'completed';
const finalMessage = isCancelled ? `Calculation stopped after completing batch ${currentBatchNum - 1}.` : 'Historical snapshots backfill completed successfully.';
const finalDurationSec = Math.round((Date.now() - overallStartTime) / 1000);
console.log(`\n--- Backfill ${finalStatus.toUpperCase()} ---`);
console.log(finalMessage);
console.log(`Total duration: ${progress.formatElapsedTime(overallStartTime)}`);
// Update history record
await connection.query(`
UPDATE public.calculate_history SET status = $1::calculation_status, end_time = NOW(), duration_seconds = $2::integer, error_message = $3
WHERE id = $4::integer;
`, [finalStatus, finalDurationSec, (isCancelled ? 'User cancelled' : null), calculateHistoryId]);
if (!isCancelled) {
progress.clearProgress(); // Clear progress state only on successful completion
} else {
progress.outputProgress({ // Final cancelled status update
status: 'cancelled',
operation: finalMessage,
currentBatch: currentBatchNum - 1,
totalBatches: totalBatches,
totalDays: totalDays,
processedDays: processedDaysTotal,
elapsed: progress.formatElapsedTime(overallStartTime),
remaining: 'Cancelled',
rate: 0,
historyId: calculateHistoryId
});
}
return { success: true, status: finalStatus, message: finalMessage, duration: finalDurationSec };
} catch (error) {
console.error('\n--- Backfill encountered an unrecoverable error ---');
console.error(error.message);
const finalDurationSec = Math.round((Date.now() - overallStartTime) / 1000);
// Update history if possible
if (connection && calculateHistoryId) {
try {
await connection.query(`
UPDATE public.calculate_history
SET status = $1::calculation_status, end_time = NOW(), duration_seconds = $2::integer, error_message = $3::text
WHERE id = $4::integer;
`, [
isCancelled ? 'cancelled' : 'failed',
finalDurationSec,
error.message,
calculateHistoryId
]);
} catch (histError) {
console.error("Failed to update history record with error state:", histError);
}
} else {
console.error("Could not update history record (no ID or connection).");
}
// FIX: Use initialized value or a default if loop never started
const batchNumForError = currentBatchNum > cmdStartBatch ? currentBatchNum - 1 : cmdStartBatch - 1;
// Update progress.outputProgress call to match actual function signature
try {
// Create progress data object
const progressData = {
status: 'failed',
operation: 'Backfill failed',
message: error.message,
currentBatch: batchNumForError,
totalBatches: totalBatches,
totalDays: totalDays,
processedDays: processedDaysTotal,
elapsed: progress.formatElapsedTime(overallStartTime),
remaining: 'Failed',
rate: 0,
// Include history ID in progress data if needed
historyId: calculateHistoryId
};
// Call with single object parameter (not separate historyId)
progress.outputProgress(progressData);
} catch (progressError) {
console.error('Failed to report progress:', progressError);
}
return { success: false, status: 'failed', error: error.message, duration: finalDurationSec };
} finally {
if (connection) {
console.log('Releasing database connection.');
connection.release();
}
// Close pool only if this script is meant to be standalone
// If part of a larger app, the app should manage pool closure
// console.log('Closing database pool.');
// await closePool();
}
}
// --- Script Execution ---
// Parse command-line arguments
const args = process.argv.slice(2);
let cmdStartDateArg, cmdEndDateArg, cmdStartBatchArg = 1; // Default start batch is 1
for (let i = 0; i < args.length; i++) {
if (args[i] === '--start-date' && args[i+1]) cmdStartDateArg = args[++i];
else if (args[i] === '--end-date' && args[i+1]) cmdEndDateArg = args[++i];
else if (args[i] === '--start-batch' && args[i+1]) cmdStartBatchArg = parseInt(args[++i], 10);
}
if (isNaN(cmdStartBatchArg) || cmdStartBatchArg < 1) {
console.warn(`Invalid --start-batch value. Defaulting to 1.`);
cmdStartBatchArg = 1;
}
// Run the backfill process
backfillSnapshots(cmdStartDateArg, cmdEndDateArg, cmdStartBatchArg)
.then(result => {
if (result.success) {
console.log(`\n${result.message} (Duration: ${result.duration}s)`);
process.exitCode = 0; // Success
} else {
console.error(`\n❌ Backfill failed: ${result.error || 'Unknown error'} (Duration: ${result.duration}s)`);
process.exitCode = 1; // Failure
}
})
.catch(err => {
console.error('\n❌ Unexpected error during backfill execution:', err);
process.exitCode = 1; // Failure
})
.finally(async () => {
// Ensure pool is closed if run standalone
console.log('Backfill script finished. Closing pool.');
await closePool(); // Make sure closePool exists and works in your db utils
process.exit(process.exitCode); // Exit with appropriate code
});

View File

@@ -0,0 +1,161 @@
-- Description: Backfills the daily_product_snapshots table using imported historical unit data
-- (daily inventory/stats) and historical price data (current prices table).
-- - Uses imported daily sales/receipt UNIT counts for accuracy.
-- - ESTIMATES historical stock levels using a forward calculation.
-- - APPROXIMATES historical REVENUE using looked-up historical base prices.
-- - APPROXIMATES historical COGS, PROFIT, and STOCK VALUE using CURRENT product costs/prices.
-- Run ONCE after importing historical data and before initial product_metrics population.
-- Dependencies: Core import tables (products), imported history tables (imported_daily_inventory,
-- imported_product_stat_history, imported_product_current_prices),
-- daily_product_snapshots table must exist.
-- Frequency: Run ONCE.
CREATE OR REPLACE FUNCTION backfill_daily_snapshots_range_final(
_start_date DATE,
_end_date DATE
)
RETURNS VOID AS $$
DECLARE
_current_processing_date DATE := _start_date;
_batch_start_time TIMESTAMPTZ;
_row_count INTEGER;
BEGIN
RAISE NOTICE 'Starting FINAL historical snapshot backfill from % to %.', _start_date, _end_date;
RAISE NOTICE 'Using historical units and historical prices (for revenue approximation).';
RAISE NOTICE 'WARNING: Historical COGS, Profit, and Stock Value use CURRENT product costs/prices.';
-- Ensure end date is not in the future
IF _end_date >= CURRENT_DATE THEN
_end_date := CURRENT_DATE - INTERVAL '1 day';
RAISE NOTICE 'Adjusted end date to % to avoid conflict with hourly script.', _end_date;
END IF;
-- Performance: Create temporary table with product info to avoid repeated lookups
CREATE TEMP TABLE IF NOT EXISTS temp_product_info AS
SELECT
pid,
sku,
COALESCE(landing_cost_price, cost_price, 0.00) as effective_cost_price,
COALESCE(price, 0.00) as current_price,
COALESCE(regular_price, 0.00) as current_regular_price
FROM public.products;
-- Performance: Create index on temporary table
CREATE INDEX IF NOT EXISTS temp_product_info_pid_idx ON temp_product_info(pid);
ANALYZE temp_product_info;
RAISE NOTICE 'Created temporary product info table with % products', (SELECT COUNT(*) FROM temp_product_info);
WHILE _current_processing_date <= _end_date LOOP
_batch_start_time := clock_timestamp();
RAISE NOTICE 'Processing date: %', _current_processing_date;
-- Get Daily Transaction Unit Info from imported history
WITH DailyHistoryUnits AS (
SELECT
pids.pid,
-- Prioritize daily_inventory, fallback to product_stat_history for sold qty
COALESCE(di.amountsold, ps.qty_sold, 0)::integer as units_sold_today,
COALESCE(di.qtyreceived, 0)::integer as units_received_today
FROM
(SELECT DISTINCT pid FROM temp_product_info) pids -- Ensure all products are considered
LEFT JOIN public.imported_daily_inventory di
ON pids.pid = di.pid AND di.date = _current_processing_date
LEFT JOIN public.imported_product_stat_history ps
ON pids.pid = ps.pid AND ps.date = _current_processing_date
-- Removed WHERE clause to ensure snapshots are created even for days with 0 activity,
-- allowing stock carry-over. The main query will handle products properly.
),
HistoricalPrice AS (
-- Find the base price (qty_buy=1) active on the processing date
SELECT DISTINCT ON (pid)
pid,
price_each
FROM public.imported_product_current_prices
WHERE
qty_buy = 1
-- Use TIMESTAMPTZ comparison logic:
AND date_active <= (_current_processing_date + interval '1 day' - interval '1 second') -- Active sometime on or before end of processing day
AND (date_deactive IS NULL OR date_deactive > _current_processing_date) -- Not deactivated before start of processing day
-- Assuming 'active' flag isn't needed if dates are correct; add 'AND active != 0' if necessary
ORDER BY
pid, date_active DESC -- Get the most recently activated price
),
PreviousStock AS (
-- Get the estimated stock from the PREVIOUS day snapshot
SELECT pid, eod_stock_quantity
FROM public.daily_product_snapshots
WHERE snapshot_date = _current_processing_date - INTERVAL '1 day'
)
-- Insert into the daily snapshots table
INSERT INTO public.daily_product_snapshots (
snapshot_date, pid, sku,
eod_stock_quantity, eod_stock_cost, eod_stock_retail, eod_stock_gross, stockout_flag,
units_sold, units_returned,
gross_revenue, discounts, returns_revenue,
net_revenue, cogs, gross_regular_revenue, profit,
units_received, cost_received,
calculation_timestamp
)
SELECT
_current_processing_date AS snapshot_date,
p.pid,
p.sku,
-- Estimated EOD Stock (using historical daily units)
-- Handle potential NULL from joins with COALESCE 0
COALESCE(ps.eod_stock_quantity, 0) + COALESCE(dh.units_received_today, 0) - COALESCE(dh.units_sold_today, 0) AS estimated_eod_stock,
-- Valued Stock (using estimated stock and CURRENT prices/costs - APPROXIMATION)
GREATEST(0, COALESCE(ps.eod_stock_quantity, 0) + COALESCE(dh.units_received_today, 0) - COALESCE(dh.units_sold_today, 0)) * p.effective_cost_price AS eod_stock_cost,
GREATEST(0, COALESCE(ps.eod_stock_quantity, 0) + COALESCE(dh.units_received_today, 0) - COALESCE(dh.units_sold_today, 0)) * p.current_price AS eod_stock_retail, -- Stock retail uses current price
GREATEST(0, COALESCE(ps.eod_stock_quantity, 0) + COALESCE(dh.units_received_today, 0) - COALESCE(dh.units_sold_today, 0)) * p.current_regular_price AS eod_stock_gross, -- Stock gross uses current regular price
-- Stockout Flag (based on estimated stock)
(COALESCE(ps.eod_stock_quantity, 0) + COALESCE(dh.units_received_today, 0) - COALESCE(dh.units_sold_today, 0)) <= 0 AS stockout_flag,
-- Today's Unit Aggregates from History
COALESCE(dh.units_sold_today, 0) as units_sold,
0 AS units_returned, -- Placeholder: Cannot determine returns from daily summary
-- Monetary Values using looked-up Historical Price and CURRENT Cost/RegPrice
COALESCE(dh.units_sold_today, 0) * COALESCE(hp.price_each, p.current_price) AS gross_revenue, -- Approx Revenue
0 AS discounts, -- Placeholder
0 AS returns_revenue, -- Placeholder
COALESCE(dh.units_sold_today, 0) * COALESCE(hp.price_each, p.current_price) AS net_revenue, -- Approx Net Revenue
COALESCE(dh.units_sold_today, 0) * p.effective_cost_price AS cogs, -- Approx COGS (uses CURRENT cost)
COALESCE(dh.units_sold_today, 0) * p.current_regular_price AS gross_regular_revenue, -- Approx Gross Regular Revenue
-- Approx Profit
(COALESCE(dh.units_sold_today, 0) * COALESCE(hp.price_each, p.current_price)) - (COALESCE(dh.units_sold_today, 0) * p.effective_cost_price) AS profit,
COALESCE(dh.units_received_today, 0) as units_received,
-- Estimate received cost using CURRENT product cost
COALESCE(dh.units_received_today, 0) * p.effective_cost_price AS cost_received, -- Approx
clock_timestamp() -- Timestamp of this specific calculation
FROM temp_product_info p -- Use the temp table for better performance
LEFT JOIN PreviousStock ps ON p.pid = ps.pid
LEFT JOIN DailyHistoryUnits dh ON p.pid = dh.pid -- Join today's historical activity
LEFT JOIN HistoricalPrice hp ON p.pid = hp.pid -- Join the looked-up historical price
-- Optimization: Only process products with activity or previous stock
WHERE (dh.units_sold_today > 0 OR dh.units_received_today > 0 OR COALESCE(ps.eod_stock_quantity, 0) > 0)
ON CONFLICT (snapshot_date, pid) DO NOTHING; -- Avoid errors if rerunning parts, but prefer clean runs
GET DIAGNOSTICS _row_count = ROW_COUNT;
RAISE NOTICE 'Processed %: Inserted/Skipped % rows. Duration: %',
_current_processing_date,
_row_count,
clock_timestamp() - _batch_start_time;
_current_processing_date := _current_processing_date + INTERVAL '1 day';
END LOOP;
-- Clean up temporary tables
DROP TABLE IF EXISTS temp_product_info;
RAISE NOTICE 'Finished FINAL historical snapshot backfill.';
END;
$$ LANGUAGE plpgsql;
-- Example usage:
-- SELECT backfill_daily_snapshots_range_final('2023-01-01'::date, '2023-12-31'::date);

View File

@@ -0,0 +1,396 @@
const path = require('path');
const fs = require('fs');
const os = require('os'); // For detecting CPU cores
// Get the base directory (the directory containing the inventory-server folder)
const baseDir = path.resolve(__dirname, '../../..');
// Load environment variables from the inventory-server directory
require('dotenv').config({ path: path.resolve(__dirname, '../..', '.env') });
// Configure statement timeout (30 minutes)
const PG_STATEMENT_TIMEOUT_MS = 1800000;
// Add error handler for uncaught exceptions
process.on('uncaughtException', (error) => {
console.error('Uncaught Exception:', error);
process.exit(1);
});
// Add error handler for unhandled promise rejections
process.on('unhandledRejection', (reason, promise) => {
console.error('Unhandled Rejection at:', promise, 'reason:', reason);
process.exit(1);
});
// Load progress module
const progress = require('./utils/progress');
// Store progress functions in global scope to ensure availability
global.formatElapsedTime = progress.formatElapsedTime;
global.estimateRemaining = progress.estimateRemaining;
global.calculateRate = progress.calculateRate;
global.outputProgress = progress.outputProgress;
global.clearProgress = progress.clearProgress;
global.getProgress = progress.getProgress;
global.logError = progress.logError;
// Load database module
const { getConnection, closePool } = require('./utils/db');
// Add cancel handler
let isCancelled = false;
let runningQueryPromise = null;
function cancelCalculation() {
if (!isCancelled) {
isCancelled = true;
console.log('Calculation has been cancelled by user');
// Store the query promise to potentially cancel it
const queryToCancel = runningQueryPromise;
if (queryToCancel) {
console.log('Attempting to cancel the running query...');
}
// Force-terminate any query that's been running for more than 5 seconds
try {
const connection = getConnection();
connection.then(async (conn) => {
try {
// Identify and terminate long-running queries from our application
await conn.query(`
SELECT pg_cancel_backend(pid)
FROM pg_stat_activity
WHERE query_start < now() - interval '5 seconds'
AND application_name = 'populate_metrics'
AND query NOT LIKE '%pg_cancel_backend%'
`);
// Release connection
conn.release();
} catch (err) {
console.error('Error during force cancellation:', err);
conn.release();
}
}).catch(err => {
console.error('Could not get connection for cancellation:', err);
});
} catch (err) {
console.error('Failed to terminate running queries:', err);
}
}
return {
success: true,
message: 'Calculation has been cancelled'
};
}
// Handle SIGTERM signal for cancellation
process.on('SIGTERM', cancelCalculation);
process.on('SIGINT', cancelCalculation);
async function populateInitialMetrics() {
let connection;
const startTime = Date.now();
let calculateHistoryId;
try {
// Clean up any previously running calculations
connection = await getConnection({
// Add performance-related settings
application_name: 'populate_metrics',
statement_timeout: PG_STATEMENT_TIMEOUT_MS, // 30 min timeout per statement
});
// Ensure the calculate_status table exists and has the correct structure
await connection.query(`
CREATE TABLE IF NOT EXISTS calculate_status (
module_name TEXT PRIMARY KEY,
last_calculation_timestamp TIMESTAMP WITH TIME ZONE NOT NULL DEFAULT CURRENT_TIMESTAMP
)
`);
await connection.query(`
UPDATE calculate_history
SET
status = 'cancelled',
end_time = NOW(),
duration_seconds = EXTRACT(EPOCH FROM (NOW() - start_time))::INTEGER,
error_message = 'Previous calculation was not completed properly'
WHERE status = 'running' AND additional_info->>'type' = 'populate_initial_metrics'
`);
// Create history record for this calculation
const historyResult = await connection.query(`
INSERT INTO calculate_history (
start_time,
status,
additional_info
) VALUES (
NOW(),
'running',
jsonb_build_object(
'type', 'populate_initial_metrics',
'sql_file', 'populate_initial_product_metrics.sql'
)
) RETURNING id
`);
calculateHistoryId = historyResult.rows[0].id;
// Initialize progress
global.outputProgress({
status: 'running',
operation: 'Starting initial product metrics population',
current: 0,
total: 100,
elapsed: '0s',
remaining: 'Calculating... (this may take a while)',
rate: 0,
percentage: '0',
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
},
historyId: calculateHistoryId
});
// Prepare the database - analyze tables
global.outputProgress({
status: 'running',
operation: 'Analyzing database tables for better query performance',
current: 2,
total: 100,
elapsed: global.formatElapsedTime(startTime),
remaining: 'Analyzing...',
rate: 0,
percentage: '2',
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
},
historyId: calculateHistoryId
});
// Enable better query planning and parallel operations
await connection.query(`
-- Analyze tables for better query planning
ANALYZE public.products;
ANALYZE public.purchase_orders;
ANALYZE public.daily_product_snapshots;
ANALYZE public.orders;
-- Enable parallel operations
SET LOCAL enable_parallel_append = on;
SET LOCAL enable_parallel_hash = on;
SET LOCAL max_parallel_workers_per_gather = 4;
-- Larger work memory for complex sorts/joins
SET LOCAL work_mem = '128MB';
`).catch(err => {
// Non-fatal if analyze fails
console.warn('Failed to analyze tables (non-fatal):', err.message);
});
// Execute the SQL query
global.outputProgress({
status: 'running',
operation: 'Executing initial metrics SQL query',
current: 5,
total: 100,
elapsed: global.formatElapsedTime(startTime),
remaining: 'Calculating... (this could take several hours with 150M+ records)',
rate: 0,
percentage: '5',
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
},
historyId: calculateHistoryId
});
// Read the SQL file
const sqlFilePath = path.resolve(__dirname, 'populate_initial_product_metrics.sql');
console.log('Base directory:', baseDir);
console.log('Script directory:', __dirname);
console.log('SQL file path:', sqlFilePath);
console.log('Current working directory:', process.cwd());
if (!fs.existsSync(sqlFilePath)) {
throw new Error(`SQL file not found at ${sqlFilePath}`);
}
// Read and clean up the SQL (Slightly more robust cleaning)
const sqlQuery = fs.readFileSync(sqlFilePath, 'utf8')
.replace(/\r\n/g, '\n') // Handle Windows endings
.replace(/\r/g, '\n') // Handle old Mac endings
.trim(); // Remove leading/trailing whitespace VERY IMPORTANT
// Log details again AFTER cleaning
console.log('SQL Query length (cleaned):', sqlQuery.length);
console.log('SQL Query structure validation:');
console.log('- Contains DO block:', sqlQuery.includes('DO $$') || sqlQuery.includes('DO $')); // Check both types of tag start
console.log('- Contains BEGIN:', sqlQuery.includes('BEGIN'));
console.log('- Contains END:', sqlQuery.includes('END $$;') || sqlQuery.includes('END $')); // Check both types of tag end
console.log('- First 50 chars:', JSON.stringify(sqlQuery.slice(0, 50)));
console.log('- Last 100 chars (cleaned):', JSON.stringify(sqlQuery.slice(-100)));
// Final check to ensure clean SQL ending
if (!sqlQuery.endsWith('END $$;')) {
console.warn('WARNING: SQL does not end with "END $$;". This might cause issues.');
console.log('Exact ending:', JSON.stringify(sqlQuery.slice(-20)));
}
// Execute the script
console.log('Starting initial product metrics population...');
// Track the query promise for potential cancellation
runningQueryPromise = connection.query({
text: sqlQuery,
rowMode: 'array'
});
await runningQueryPromise;
runningQueryPromise = null;
// Update progress to 100%
global.outputProgress({
status: 'complete',
operation: 'Initial product metrics population complete',
current: 100,
total: 100,
elapsed: global.formatElapsedTime(startTime),
remaining: '0s',
rate: 0,
percentage: '100',
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
},
historyId: calculateHistoryId
});
// Update history with completion
await connection.query(`
UPDATE calculate_history
SET
end_time = NOW(),
duration_seconds = $1,
status = 'completed'
WHERE id = $2
`, [Math.round((Date.now() - startTime) / 1000), calculateHistoryId]);
// Clear progress file on successful completion
global.clearProgress();
return {
success: true,
message: 'Initial product metrics population completed successfully',
duration: Math.round((Date.now() - startTime) / 1000)
};
} catch (error) {
const endTime = Date.now();
const totalElapsedSeconds = Math.round((endTime - startTime) / 1000);
// Enhanced error logging
console.error('Error details:', {
message: error.message,
code: error.code,
hint: error.hint,
position: error.position,
detail: error.detail,
where: error.where ? error.where.substring(0, 500) + '...' : undefined, // Truncate to avoid huge logs
severity: error.severity,
file: error.file,
line: error.line,
routine: error.routine
});
// Update history with error
if (connection && calculateHistoryId) {
await connection.query(`
UPDATE calculate_history
SET
end_time = NOW(),
duration_seconds = $1,
status = $2,
error_message = $3
WHERE id = $4
`, [
totalElapsedSeconds,
isCancelled ? 'cancelled' : 'failed',
error.message,
calculateHistoryId
]);
}
if (isCancelled) {
global.outputProgress({
status: 'cancelled',
operation: 'Calculation cancelled',
current: 50,
total: 100,
elapsed: global.formatElapsedTime(startTime),
remaining: null,
rate: 0,
percentage: '50',
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: totalElapsedSeconds
},
historyId: calculateHistoryId
});
} else {
global.outputProgress({
status: 'error',
operation: 'Error during initial product metrics population',
message: error.message,
current: 0,
total: 100,
elapsed: global.formatElapsedTime(startTime),
remaining: null,
rate: 0,
percentage: '0',
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: totalElapsedSeconds
},
historyId: calculateHistoryId
});
}
console.error('Error during initial product metrics population:', error);
return {
success: false,
error: error.message,
duration: totalElapsedSeconds
};
} finally {
if (connection) {
connection.release();
}
await closePool();
}
}
// Start population process
populateInitialMetrics()
.then(result => {
if (result.success) {
console.log(`Initial product metrics population completed successfully in ${result.duration} seconds`);
process.exit(0);
} else {
console.error(`Initial product metrics population failed: ${result.error}`);
process.exit(1);
}
})
.catch(err => {
console.error('Unexpected error:', err);
process.exit(1);
});

View File

@@ -0,0 +1,448 @@
-- Description: Performs the first population OR full recalculation of the product_metrics table based on
-- historically backfilled daily_product_snapshots and current product/PO data.
-- Calculates all metrics considering the full available history up to 'yesterday'.
-- Run ONCE after backfill_historical_snapshots_final.sql completes successfully.
-- Dependencies: Core import tables (products, purchase_orders), daily_product_snapshots (historically populated),
-- configuration tables (settings_*), product_metrics table must exist.
-- Frequency: Run ONCE.
DO $$
DECLARE
_module_name VARCHAR := 'product_metrics_population'; -- Generic name
_start_time TIMESTAMPTZ := clock_timestamp();
-- Calculate metrics AS OF the end of the last fully completed day
_calculation_date DATE := CURRENT_DATE - INTERVAL '1 day';
BEGIN
RAISE NOTICE 'Running % module. Calculating AS OF: %. Start Time: %', _module_name, _calculation_date, _start_time;
-- Optional: Consider TRUNCATE if you want a completely fresh start,
-- otherwise ON CONFLICT will update existing rows if this is rerun.
-- TRUNCATE TABLE public.product_metrics;
RAISE NOTICE 'Populating product_metrics table. This may take some time...';
-- CTEs to gather necessary information AS OF _calculation_date
WITH CurrentInfo AS (
-- Fetches current product details, including costs/prices used for forecasting & fallbacks
SELECT
p.pid, p.sku, p.title, p.brand, p.vendor, COALESCE(p.image_175, p.image) as image_url,
p.visible as is_visible, p.replenishable,
COALESCE(p.price, 0.00) as current_price, COALESCE(p.regular_price, 0.00) as current_regular_price,
COALESCE(p.cost_price, 0.00) as current_cost_price,
COALESCE(p.landing_cost_price, p.cost_price, 0.00) as current_effective_cost, -- Use landing if available, else cost
p.stock_quantity as current_stock, -- Use actual current stock for forecast base
p.created_at, p.first_received, p.date_last_sold,
p.moq,
p.uom
FROM public.products p
),
OnOrderInfo AS (
-- Calculates current on-order quantities and costs
SELECT
pid,
COALESCE(SUM(ordered - received), 0) AS on_order_qty,
COALESCE(SUM((ordered - received) * cost_price), 0.00) AS on_order_cost,
MIN(expected_date) AS earliest_expected_date
FROM public.purchase_orders
-- Use the most common statuses representing active, unfulfilled POs
WHERE status IN ('open', 'partially_received', 'ordered', 'preordered', 'receiving_started', 'electronically_sent', 'electronically_ready_send')
AND (ordered - received) > 0
GROUP BY pid
),
HistoricalDates AS (
-- Determines key historical dates from orders and PO history (receiving_history)
SELECT
p.pid,
MIN(o.date)::date AS date_first_sold,
MAX(o.date)::date AS max_order_date, -- Used as fallback for date_last_sold
MIN(rh.first_receipt_date) AS date_first_received_calc,
MAX(rh.last_receipt_date) AS date_last_received_calc
FROM public.products p
LEFT JOIN public.orders o ON p.pid = o.pid AND o.quantity > 0 AND o.status NOT IN ('canceled', 'returned')
LEFT JOIN (
SELECT
po.pid,
MIN((rh.item->>'received_at')::date) as first_receipt_date,
MAX((rh.item->>'received_at')::date) as last_receipt_date
FROM public.purchase_orders po
CROSS JOIN LATERAL jsonb_array_elements(po.receiving_history) AS rh(item)
WHERE jsonb_typeof(po.receiving_history) = 'array' AND jsonb_array_length(po.receiving_history) > 0
GROUP BY po.pid
) rh ON p.pid = rh.pid
GROUP BY p.pid
),
SnapshotAggregates AS (
-- Aggregates metrics from historical snapshots up to the _calculation_date
SELECT
pid,
-- Rolling periods relative to _calculation_date
SUM(CASE WHEN snapshot_date BETWEEN _calculation_date - INTERVAL '6 days' AND _calculation_date THEN units_sold ELSE 0 END) AS sales_7d,
SUM(CASE WHEN snapshot_date BETWEEN _calculation_date - INTERVAL '6 days' AND _calculation_date THEN net_revenue ELSE 0 END) AS revenue_7d,
SUM(CASE WHEN snapshot_date BETWEEN _calculation_date - INTERVAL '13 days' AND _calculation_date THEN units_sold ELSE 0 END) AS sales_14d,
SUM(CASE WHEN snapshot_date BETWEEN _calculation_date - INTERVAL '13 days' AND _calculation_date THEN net_revenue ELSE 0 END) AS revenue_14d,
SUM(CASE WHEN snapshot_date BETWEEN _calculation_date - INTERVAL '29 days' AND _calculation_date THEN units_sold ELSE 0 END) AS sales_30d,
SUM(CASE WHEN snapshot_date BETWEEN _calculation_date - INTERVAL '29 days' AND _calculation_date THEN net_revenue ELSE 0 END) AS revenue_30d,
SUM(CASE WHEN snapshot_date BETWEEN _calculation_date - INTERVAL '29 days' AND _calculation_date THEN cogs ELSE 0 END) AS cogs_30d,
SUM(CASE WHEN snapshot_date BETWEEN _calculation_date - INTERVAL '29 days' AND _calculation_date THEN profit ELSE 0 END) AS profit_30d,
SUM(CASE WHEN snapshot_date BETWEEN _calculation_date - INTERVAL '29 days' AND _calculation_date THEN units_returned ELSE 0 END) AS returns_units_30d,
SUM(CASE WHEN snapshot_date BETWEEN _calculation_date - INTERVAL '29 days' AND _calculation_date THEN returns_revenue ELSE 0 END) AS returns_revenue_30d,
SUM(CASE WHEN snapshot_date BETWEEN _calculation_date - INTERVAL '29 days' AND _calculation_date THEN discounts ELSE 0 END) AS discounts_30d,
SUM(CASE WHEN snapshot_date BETWEEN _calculation_date - INTERVAL '29 days' AND _calculation_date THEN gross_revenue ELSE 0 END) AS gross_revenue_30d,
SUM(CASE WHEN snapshot_date BETWEEN _calculation_date - INTERVAL '29 days' AND _calculation_date THEN gross_regular_revenue ELSE 0 END) AS gross_regular_revenue_30d,
SUM(CASE WHEN snapshot_date BETWEEN _calculation_date - INTERVAL '29 days' AND _calculation_date AND stockout_flag THEN 1 ELSE 0 END) AS stockout_days_30d,
-- Add 90-day aggregates if needed
SUM(CASE WHEN snapshot_date BETWEEN _calculation_date - INTERVAL '89 days' AND _calculation_date THEN units_sold ELSE 0 END) AS sales_90d,
SUM(CASE WHEN snapshot_date BETWEEN _calculation_date - INTERVAL '89 days' AND _calculation_date THEN net_revenue ELSE 0 END) AS revenue_90d,
SUM(CASE WHEN snapshot_date BETWEEN _calculation_date - INTERVAL '89 days' AND _calculation_date THEN cogs ELSE 0 END) AS cogs_90d,
SUM(CASE WHEN snapshot_date BETWEEN _calculation_date - INTERVAL '89 days' AND _calculation_date THEN profit ELSE 0 END) AS profit_90d,
-- Add 60-day aggregates if needed
SUM(CASE WHEN snapshot_date BETWEEN _calculation_date - INTERVAL '59 days' AND _calculation_date THEN units_sold ELSE 0 END) AS sales_60d,
SUM(CASE WHEN snapshot_date BETWEEN _calculation_date - INTERVAL '59 days' AND _calculation_date THEN net_revenue ELSE 0 END) AS revenue_60d,
SUM(CASE WHEN snapshot_date BETWEEN _calculation_date - INTERVAL '59 days' AND _calculation_date THEN cogs ELSE 0 END) AS cogs_60d,
SUM(CASE WHEN snapshot_date BETWEEN _calculation_date - INTERVAL '59 days' AND _calculation_date THEN profit ELSE 0 END) AS profit_60d,
SUM(CASE WHEN snapshot_date BETWEEN _calculation_date - INTERVAL '364 days' AND _calculation_date THEN units_sold ELSE 0 END) AS sales_365d,
SUM(CASE WHEN snapshot_date BETWEEN _calculation_date - INTERVAL '364 days' AND _calculation_date THEN net_revenue ELSE 0 END) AS revenue_365d,
SUM(CASE WHEN snapshot_date BETWEEN _calculation_date - INTERVAL '29 days' AND _calculation_date THEN units_received ELSE 0 END) AS received_qty_30d,
SUM(CASE WHEN snapshot_date BETWEEN _calculation_date - INTERVAL '29 days' AND _calculation_date THEN cost_received ELSE 0 END) AS received_cost_30d,
-- Averages over the last 30 days ending _calculation_date
AVG(CASE WHEN snapshot_date BETWEEN _calculation_date - INTERVAL '29 days' AND _calculation_date THEN eod_stock_quantity END) AS avg_stock_units_30d,
AVG(CASE WHEN snapshot_date BETWEEN _calculation_date - INTERVAL '29 days' AND _calculation_date THEN eod_stock_cost END) AS avg_stock_cost_30d,
AVG(CASE WHEN snapshot_date BETWEEN _calculation_date - INTERVAL '29 days' AND _calculation_date THEN eod_stock_retail END) AS avg_stock_retail_30d,
AVG(CASE WHEN snapshot_date BETWEEN _calculation_date - INTERVAL '29 days' AND _calculation_date THEN eod_stock_gross END) AS avg_stock_gross_30d,
-- Lifetime (Sum over ALL available snapshots up to calculation date)
SUM(units_sold) AS lifetime_sales,
SUM(net_revenue) AS lifetime_revenue,
-- Yesterday (Sales for the specific _calculation_date)
SUM(CASE WHEN snapshot_date = _calculation_date THEN units_sold ELSE 0 END) as yesterday_sales
FROM public.daily_product_snapshots
WHERE snapshot_date <= _calculation_date -- Ensure we only use data up to the calculation point
GROUP BY pid
),
FirstPeriodMetrics AS (
-- Calculates sales/revenue for first X days after first sale date
-- Uses HistoricalDates CTE to get the first sale date
SELECT
pid, date_first_sold,
SUM(CASE WHEN snapshot_date BETWEEN date_first_sold AND date_first_sold + INTERVAL '6 days' THEN units_sold ELSE 0 END) AS first_7_days_sales,
SUM(CASE WHEN snapshot_date BETWEEN date_first_sold AND date_first_sold + INTERVAL '6 days' THEN net_revenue ELSE 0 END) AS first_7_days_revenue,
SUM(CASE WHEN snapshot_date BETWEEN date_first_sold AND date_first_sold + INTERVAL '29 days' THEN units_sold ELSE 0 END) AS first_30_days_sales,
SUM(CASE WHEN snapshot_date BETWEEN date_first_sold AND date_first_sold + INTERVAL '29 days' THEN net_revenue ELSE 0 END) AS first_30_days_revenue,
SUM(CASE WHEN snapshot_date BETWEEN date_first_sold AND date_first_sold + INTERVAL '59 days' THEN units_sold ELSE 0 END) AS first_60_days_sales,
SUM(CASE WHEN snapshot_date BETWEEN date_first_sold AND date_first_sold + INTERVAL '59 days' THEN net_revenue ELSE 0 END) AS first_60_days_revenue,
SUM(CASE WHEN snapshot_date BETWEEN date_first_sold AND date_first_sold + INTERVAL '89 days' THEN units_sold ELSE 0 END) AS first_90_days_sales,
SUM(CASE WHEN snapshot_date BETWEEN date_first_sold AND date_first_sold + INTERVAL '89 days' THEN net_revenue ELSE 0 END) AS first_90_days_revenue
FROM public.daily_product_snapshots ds
JOIN HistoricalDates hd USING(pid)
WHERE date_first_sold IS NOT NULL
AND snapshot_date >= date_first_sold -- Only consider snapshots after first sale
AND snapshot_date <= _calculation_date -- Only up to the overall calculation date
GROUP BY pid, date_first_sold
),
Settings AS (
-- Fetches effective configuration settings (Product > Vendor > Global)
SELECT
p.pid,
COALESCE(sp.lead_time_days, sv.default_lead_time_days, (SELECT setting_value FROM settings_global WHERE setting_key = 'default_lead_time_days')::int, 14) AS effective_lead_time,
COALESCE(sp.days_of_stock, sv.default_days_of_stock, (SELECT setting_value FROM settings_global WHERE setting_key = 'default_days_of_stock')::int, 30) AS effective_days_of_stock,
COALESCE(sp.safety_stock, (SELECT setting_value::int FROM settings_global WHERE setting_key = 'default_safety_stock_units'), 0) AS effective_safety_stock,
COALESCE(sp.exclude_from_forecast, FALSE) AS exclude_forecast
FROM public.products p
LEFT JOIN public.settings_product sp ON p.pid = sp.pid
LEFT JOIN public.settings_vendor sv ON p.vendor = sv.vendor
),
AvgLeadTime AS (
-- Calculate Average Lead Time from historical POs
SELECT
pid,
AVG(GREATEST(1,
CASE
WHEN last_received_date IS NOT NULL AND date IS NOT NULL
THEN (last_received_date::date - date::date)
ELSE 1
END
))::int AS avg_lead_time_days_calc
FROM public.purchase_orders
WHERE status = 'received' -- Assumes 'received' marks full receipt
AND last_received_date IS NOT NULL
AND date IS NOT NULL
AND last_received_date >= date
GROUP BY pid
),
RankedForABC AS (
-- Ranks products based on the configured ABC metric (using historical data)
SELECT
p.pid,
CASE COALESCE((SELECT setting_value FROM settings_global WHERE setting_key = 'abc_calculation_basis'), 'revenue_30d')
WHEN 'sales_30d' THEN COALESCE(sa.sales_30d, 0)
WHEN 'lifetime_revenue' THEN COALESCE(sa.lifetime_revenue, 0)::numeric
ELSE COALESCE(sa.revenue_30d, 0) -- Default to revenue_30d
END AS metric_value
FROM public.products p -- Use products as the base
JOIN SnapshotAggregates sa ON p.pid = sa.pid
WHERE p.replenishable = TRUE -- Only rank replenishable products
AND (CASE COALESCE((SELECT setting_value FROM settings_global WHERE setting_key = 'abc_calculation_basis'), 'revenue_30d')
WHEN 'sales_30d' THEN COALESCE(sa.sales_30d, 0)
WHEN 'lifetime_revenue' THEN COALESCE(sa.lifetime_revenue, 0)::numeric
ELSE COALESCE(sa.revenue_30d, 0)
END) > 0 -- Exclude zero-value products from ranking
),
CumulativeABC AS (
-- Calculates cumulative metric values for ABC ranking
SELECT
pid, metric_value,
SUM(metric_value) OVER (ORDER BY metric_value DESC NULLS LAST ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW) as cumulative_metric,
SUM(metric_value) OVER () as total_metric
FROM RankedForABC
),
FinalABC AS (
-- Assigns A, B, or C class based on thresholds
SELECT
pid,
CASE
WHEN cumulative_metric / NULLIF(total_metric, 0) <= COALESCE((SELECT setting_value::numeric FROM settings_global WHERE setting_key = 'abc_revenue_threshold_a'), 0.8) THEN 'A'::char(1)
WHEN cumulative_metric / NULLIF(total_metric, 0) <= COALESCE((SELECT setting_value::numeric FROM settings_global WHERE setting_key = 'abc_revenue_threshold_b'), 0.95) THEN 'B'::char(1)
ELSE 'C'::char(1)
END AS abc_class_calc
FROM CumulativeABC
)
-- Final INSERT/UPDATE statement using all the prepared CTEs
INSERT INTO public.product_metrics (
pid, last_calculated, sku, title, brand, vendor, image_url, is_visible, is_replenishable,
current_price, current_regular_price, current_cost_price, current_landing_cost_price,
current_stock, current_stock_cost, current_stock_retail, current_stock_gross,
on_order_qty, on_order_cost, on_order_retail, earliest_expected_date,
date_created, date_first_received, date_last_received, date_first_sold, date_last_sold, age_days,
sales_7d, revenue_7d, sales_14d, revenue_14d, sales_30d, revenue_30d, cogs_30d, profit_30d,
sales_60d, revenue_60d, cogs_60d, profit_60d,
sales_90d, revenue_90d, cogs_90d, profit_90d,
returns_units_30d, returns_revenue_30d, discounts_30d, gross_revenue_30d, gross_regular_revenue_30d,
stockout_days_30d, sales_365d, revenue_365d,
avg_stock_units_30d, avg_stock_cost_30d, avg_stock_retail_30d, avg_stock_gross_30d,
received_qty_30d, received_cost_30d,
lifetime_sales, lifetime_revenue,
first_7_days_sales, first_7_days_revenue, first_30_days_sales, first_30_days_revenue,
first_60_days_sales, first_60_days_revenue, first_90_days_sales, first_90_days_revenue,
asp_30d, acp_30d, avg_ros_30d, avg_sales_per_day_30d,
margin_30d, markup_30d, gmroi_30d, stockturn_30d, return_rate_30d, discount_rate_30d,
stockout_rate_30d, markdown_30d, markdown_rate_30d, sell_through_30d,
avg_lead_time_days, abc_class,
sales_velocity_daily, config_lead_time, config_days_of_stock, config_safety_stock,
planning_period_days, lead_time_forecast_units, days_of_stock_forecast_units,
planning_period_forecast_units, lead_time_closing_stock, days_of_stock_closing_stock,
replenishment_needed_raw, replenishment_units, replenishment_cost, replenishment_retail, replenishment_profit,
to_order_units, forecast_lost_sales_units, forecast_lost_revenue,
stock_cover_in_days, po_cover_in_days, sells_out_in_days, replenish_date,
overstocked_units, overstocked_cost, overstocked_retail, is_old_stock,
yesterday_sales
)
SELECT
-- Select columns in order, joining all CTEs by pid
ci.pid, _start_time, ci.sku, ci.title, ci.brand, ci.vendor, ci.image_url, ci.is_visible, ci.replenishable,
ci.current_price, ci.current_regular_price, ci.current_cost_price, ci.current_effective_cost,
ci.current_stock, (ci.current_stock * COALESCE(ci.current_effective_cost, 0.00))::numeric(12,2), (ci.current_stock * COALESCE(ci.current_price, 0.00))::numeric(12,2), (ci.current_stock * COALESCE(ci.current_regular_price, 0.00))::numeric(12,2),
COALESCE(ooi.on_order_qty, 0), COALESCE(ooi.on_order_cost, 0.00)::numeric(12,2), (COALESCE(ooi.on_order_qty, 0) * COALESCE(ci.current_price, 0.00))::numeric(12,2), ooi.earliest_expected_date,
-- Fix type issue with date calculation - properly cast timestamps to dates before arithmetic
ci.created_at::date,
COALESCE(ci.first_received::date, hd.date_first_received_calc),
hd.date_last_received_calc,
hd.date_first_sold,
COALESCE(ci.date_last_sold, hd.max_order_date),
-- Fix timestamp + integer error by ensuring we work only with dates
CASE
WHEN LEAST(ci.created_at::date, COALESCE(hd.date_first_sold, ci.created_at::date)) IS NOT NULL
THEN (_calculation_date::date - LEAST(ci.created_at::date, COALESCE(hd.date_first_sold, ci.created_at::date)))::int
ELSE NULL
END,
COALESCE(sa.sales_7d, 0), COALESCE(sa.revenue_7d, 0), COALESCE(sa.sales_14d, 0), COALESCE(sa.revenue_14d, 0), COALESCE(sa.sales_30d, 0), COALESCE(sa.revenue_30d, 0), COALESCE(sa.cogs_30d, 0), COALESCE(sa.profit_30d, 0),
COALESCE(sa.sales_60d, 0), COALESCE(sa.revenue_60d, 0), COALESCE(sa.cogs_60d, 0), COALESCE(sa.profit_60d, 0),
COALESCE(sa.sales_90d, 0), COALESCE(sa.revenue_90d, 0), COALESCE(sa.cogs_90d, 0), COALESCE(sa.profit_90d, 0),
COALESCE(sa.returns_units_30d, 0), COALESCE(sa.returns_revenue_30d, 0), COALESCE(sa.discounts_30d, 0), COALESCE(sa.gross_revenue_30d, 0), COALESCE(sa.gross_regular_revenue_30d, 0),
COALESCE(sa.stockout_days_30d, 0), COALESCE(sa.sales_365d, 0), COALESCE(sa.revenue_365d, 0),
sa.avg_stock_units_30d, sa.avg_stock_cost_30d, sa.avg_stock_retail_30d, sa.avg_stock_gross_30d, -- Averages can be NULL if no data
COALESCE(sa.received_qty_30d, 0), COALESCE(sa.received_cost_30d, 0),
COALESCE(sa.lifetime_sales, 0), COALESCE(sa.lifetime_revenue, 0),
fpm.first_7_days_sales, fpm.first_7_days_revenue, fpm.first_30_days_sales, fpm.first_30_days_revenue,
fpm.first_60_days_sales, fpm.first_60_days_revenue, fpm.first_90_days_sales, fpm.first_90_days_revenue,
-- Calculated KPIs (using COALESCE on inputs where appropriate)
sa.revenue_30d / NULLIF(sa.sales_30d, 0) AS asp_30d,
sa.cogs_30d / NULLIF(sa.sales_30d, 0) AS acp_30d,
sa.profit_30d / NULLIF(sa.sales_30d, 0) AS avg_ros_30d,
COALESCE(sa.sales_30d, 0) / 30.0 AS avg_sales_per_day_30d,
-- Fix for percentages - cast to numeric with appropriate precision
((sa.profit_30d / NULLIF(sa.revenue_30d, 0)) * 100)::numeric(8,2) AS margin_30d,
((sa.profit_30d / NULLIF(sa.cogs_30d, 0)) * 100)::numeric(8,2) AS markup_30d,
sa.profit_30d / NULLIF(sa.avg_stock_cost_30d, 0) AS gmroi_30d,
sa.sales_30d / NULLIF(sa.avg_stock_units_30d, 0) AS stockturn_30d,
((sa.returns_units_30d / NULLIF(COALESCE(sa.sales_30d, 0) + COALESCE(sa.returns_units_30d, 0), 0)) * 100)::numeric(8,2) AS return_rate_30d,
((sa.discounts_30d / NULLIF(sa.gross_revenue_30d, 0)) * 100)::numeric(8,2) AS discount_rate_30d,
((COALESCE(sa.stockout_days_30d, 0) / 30.0) * 100)::numeric(8,2) AS stockout_rate_30d,
GREATEST(0, sa.gross_regular_revenue_30d - sa.gross_revenue_30d) AS markdown_30d, -- Ensure markdown isn't negative
((GREATEST(0, sa.gross_regular_revenue_30d - sa.gross_revenue_30d) / NULLIF(sa.gross_regular_revenue_30d, 0)) * 100)::numeric(8,2) AS markdown_rate_30d,
-- Sell Through Rate: Sales / (Stock at end of period + Sales). This is one definition proxying for Sales / Beginning Stock.
((sa.sales_30d / NULLIF(
(SELECT eod_stock_quantity FROM daily_product_snapshots WHERE snapshot_date = _calculation_date AND pid = ci.pid LIMIT 1) + COALESCE(sa.sales_30d, 0)
, 0)) * 100)::numeric(8,2) AS sell_through_30d,
-- Use calculated periodic metrics
alt.avg_lead_time_days_calc,
CASE
WHEN ci.replenishable = FALSE THEN NULL -- Non-replenishable don't get a class
ELSE COALESCE(fa.abc_class_calc, 'C') -- Default ranked replenishable but non-contributing to C
END,
-- Forecasting intermediate values (based on historical aggregates ending _calculation_date)
(COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) AS sales_velocity_daily, -- Ensure divisor > 0
s.effective_lead_time AS config_lead_time, s.effective_days_of_stock AS config_days_of_stock, s.effective_safety_stock AS config_safety_stock,
(s.effective_lead_time + s.effective_days_of_stock) AS planning_period_days,
-- Calculate raw forecast need components (using safe velocity)
(COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) * s.effective_lead_time AS lead_time_forecast_units,
(COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) * s.effective_days_of_stock AS days_of_stock_forecast_units,
-- Planning period forecast units (sum of lead time and DOS units)
CEILING((COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) * s.effective_lead_time)
+ CEILING((COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) * s.effective_days_of_stock) AS planning_period_forecast_units,
-- Closing stock calculations (using raw forecast components for accuracy before rounding)
(ci.current_stock + COALESCE(ooi.on_order_qty, 0) - ((COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) * s.effective_lead_time)) AS lead_time_closing_stock,
((ci.current_stock + COALESCE(ooi.on_order_qty, 0) - ((COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) * s.effective_lead_time)))
- ((COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) * s.effective_days_of_stock) AS days_of_stock_closing_stock,
-- Raw replenishment needed
(CEILING((COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) * s.effective_lead_time) -- Use rounded forecast units
+ CEILING((COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) * s.effective_days_of_stock))
+ s.effective_safety_stock - ci.current_stock - COALESCE(ooi.on_order_qty, 0) AS replenishment_needed_raw,
-- Final Forecasting Metrics
-- Replenishment Units (calculated need, before MOQ)
CEILING(GREATEST(0,
(CEILING((COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) * s.effective_lead_time)
+ CEILING((COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) * s.effective_days_of_stock))
+ s.effective_safety_stock - ci.current_stock - COALESCE(ooi.on_order_qty, 0)
))::int AS replenishment_units,
-- Replenishment Cost/Retail/Profit (based on replenishment_units)
(CEILING(GREATEST(0,
(CEILING((COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) * s.effective_lead_time)
+ CEILING((COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) * s.effective_days_of_stock))
+ s.effective_safety_stock - ci.current_stock - COALESCE(ooi.on_order_qty, 0)
))::int) * COALESCE(ci.current_effective_cost, 0.00)::numeric(12,2) AS replenishment_cost,
(CEILING(GREATEST(0,
(CEILING((COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) * s.effective_lead_time)
+ CEILING((COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) * s.effective_days_of_stock))
+ s.effective_safety_stock - ci.current_stock - COALESCE(ooi.on_order_qty, 0)
))::int) * COALESCE(ci.current_price, 0.00)::numeric(12,2) AS replenishment_retail,
(CEILING(GREATEST(0,
(CEILING((COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) * s.effective_lead_time)
+ CEILING((COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) * s.effective_days_of_stock))
+ s.effective_safety_stock - ci.current_stock - COALESCE(ooi.on_order_qty, 0)
))::int) * (COALESCE(ci.current_price, 0.00) - COALESCE(ci.current_effective_cost, 0.00))::numeric(12,2) AS replenishment_profit,
-- *** FIX: To Order Units (Apply MOQ rounding) ***
CASE
WHEN COALESCE(ci.moq, 0) <= 1 THEN -- Treat no/invalid MOQ or MOQ=1 as no rounding needed
CEILING(GREATEST(0,
(CEILING((COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) * s.effective_lead_time)
+ CEILING((COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) * s.effective_days_of_stock))
+ s.effective_safety_stock - ci.current_stock - COALESCE(ooi.on_order_qty, 0)
))::int
ELSE -- Apply MOQ rounding: Round UP to nearest multiple of MOQ
(CEILING(GREATEST(0,
(CEILING((COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) * s.effective_lead_time)
+ CEILING((COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) * s.effective_days_of_stock))
+ s.effective_safety_stock - ci.current_stock - COALESCE(ooi.on_order_qty, 0)
) / NULLIF(ci.moq::numeric, 0)) * COALESCE(ci.moq, 1))::int
END AS to_order_units,
-- Forecast Lost Sales (Units occurring during lead time if current+on_order is insufficient)
CEILING(GREATEST(0,
((COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) * s.effective_lead_time) -- Demand during lead time
- (ci.current_stock + COALESCE(ooi.on_order_qty, 0)) -- Supply available before order arrives
))::int AS forecast_lost_sales_units,
-- Forecast Lost Revenue
(CEILING(GREATEST(0,
((COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) * s.effective_lead_time)
- (ci.current_stock + COALESCE(ooi.on_order_qty, 0))
))::int) * COALESCE(ci.current_price, 0.00)::numeric(12,2) AS forecast_lost_revenue,
-- Stock Cover etc (using safe velocity)
ci.current_stock / NULLIF((COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)), 0) AS stock_cover_in_days,
COALESCE(ooi.on_order_qty, 0) / NULLIF((COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)), 0) AS po_cover_in_days,
(ci.current_stock + COALESCE(ooi.on_order_qty, 0)) / NULLIF((COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)), 0) AS sells_out_in_days,
-- Replenish Date (Project forward from 'today', which is _calculation_date + 1 day)
CASE
WHEN (COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) > 0 -- Check for positive velocity
THEN
_calculation_date + INTERVAL '1 day' -- Today
+ FLOOR(GREATEST(0, ci.current_stock - s.effective_safety_stock) -- Stock above safety
/ (COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) -- divided by velocity
)::integer * INTERVAL '1 day' -- Gives date safety stock is hit
- s.effective_lead_time * INTERVAL '1 day' -- Subtract lead time
ELSE NULL -- Cannot calculate if no sales velocity
END AS replenish_date,
-- Overstocked Units (Stock above safety + planning period demand)
GREATEST(0, ci.current_stock - s.effective_safety_stock -
(CEILING((COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) * s.effective_lead_time) -- Demand during lead time
+ CEILING((COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) * s.effective_days_of_stock)) -- Demand during DOS
)::int AS overstocked_units,
(GREATEST(0, ci.current_stock - s.effective_safety_stock -
(CEILING((COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) * s.effective_lead_time)
+ CEILING((COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) * s.effective_days_of_stock))
)::int) * COALESCE(ci.current_effective_cost, 0.00)::numeric(12,2) AS overstocked_cost,
(GREATEST(0, ci.current_stock - s.effective_safety_stock -
(CEILING((COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) * s.effective_lead_time)
+ CEILING((COALESCE(sa.sales_30d, 0) / NULLIF(GREATEST(1.0, 30.0 - COALESCE(sa.stockout_days_30d, 0)), 0)) * s.effective_days_of_stock))
)::int) * COALESCE(ci.current_price, 0.00)::numeric(12,2) AS overstocked_retail,
-- Old Stock Flag
(ci.created_at::date < (_calculation_date - INTERVAL '60 day')::date) AND
(COALESCE(ci.date_last_sold, hd.max_order_date) IS NULL OR COALESCE(ci.date_last_sold, hd.max_order_date) < (_calculation_date - INTERVAL '60 day')::date) AND
(hd.date_last_received_calc IS NULL OR hd.date_last_received_calc < (_calculation_date - INTERVAL '60 day')::date) AND
COALESCE(ooi.on_order_qty, 0) = 0 AS is_old_stock,
COALESCE(sa.yesterday_sales, 0) -- Sales for _calculation_date
FROM CurrentInfo ci
LEFT JOIN OnOrderInfo ooi ON ci.pid = ooi.pid
LEFT JOIN HistoricalDates hd ON ci.pid = hd.pid
LEFT JOIN SnapshotAggregates sa ON ci.pid = sa.pid
LEFT JOIN FirstPeriodMetrics fpm ON ci.pid = fpm.pid
LEFT JOIN Settings s ON ci.pid = s.pid
LEFT JOIN AvgLeadTime alt ON ci.pid = alt.pid -- Join calculated avg lead time
LEFT JOIN FinalABC fa ON ci.pid = fa.pid -- Join calculated ABC class
WHERE s.exclude_forecast IS FALSE OR s.exclude_forecast IS NULL
ON CONFLICT (pid) DO UPDATE SET
-- *** IMPORTANT: List ALL columns here, ensuring order matches INSERT list ***
-- Update ALL columns to ensure entire row is refreshed
last_calculated = EXCLUDED.last_calculated, sku = EXCLUDED.sku, title = EXCLUDED.title, brand = EXCLUDED.brand, vendor = EXCLUDED.vendor, image_url = EXCLUDED.image_url, is_visible = EXCLUDED.is_visible, is_replenishable = EXCLUDED.is_replenishable,
current_price = EXCLUDED.current_price, current_regular_price = EXCLUDED.current_regular_price, current_cost_price = EXCLUDED.current_cost_price, current_landing_cost_price = EXCLUDED.current_landing_cost_price,
current_stock = EXCLUDED.current_stock, current_stock_cost = EXCLUDED.current_stock_cost, current_stock_retail = EXCLUDED.current_stock_retail, current_stock_gross = EXCLUDED.current_stock_gross,
on_order_qty = EXCLUDED.on_order_qty, on_order_cost = EXCLUDED.on_order_cost, on_order_retail = EXCLUDED.on_order_retail, earliest_expected_date = EXCLUDED.earliest_expected_date,
date_created = EXCLUDED.date_created, date_first_received = EXCLUDED.date_first_received, date_last_received = EXCLUDED.date_last_received, date_first_sold = EXCLUDED.date_first_sold, date_last_sold = EXCLUDED.date_last_sold, age_days = EXCLUDED.age_days,
sales_7d = EXCLUDED.sales_7d, revenue_7d = EXCLUDED.revenue_7d, sales_14d = EXCLUDED.sales_14d, revenue_14d = EXCLUDED.revenue_14d, sales_30d = EXCLUDED.sales_30d, revenue_30d = EXCLUDED.revenue_30d, cogs_30d = EXCLUDED.cogs_30d, profit_30d = EXCLUDED.profit_30d,
-- Add 60d/90d columns
sales_60d = EXCLUDED.sales_60d, revenue_60d = EXCLUDED.revenue_60d, cogs_60d = EXCLUDED.cogs_60d, profit_60d = EXCLUDED.profit_60d,
sales_90d = EXCLUDED.sales_90d, revenue_90d = EXCLUDED.revenue_90d, cogs_90d = EXCLUDED.cogs_90d, profit_90d = EXCLUDED.profit_90d,
returns_units_30d = EXCLUDED.returns_units_30d, returns_revenue_30d = EXCLUDED.returns_revenue_30d, discounts_30d = EXCLUDED.discounts_30d, gross_revenue_30d = EXCLUDED.gross_revenue_30d, gross_regular_revenue_30d = EXCLUDED.gross_regular_revenue_30d,
stockout_days_30d = EXCLUDED.stockout_days_30d, sales_365d = EXCLUDED.sales_365d, revenue_365d = EXCLUDED.revenue_365d,
avg_stock_units_30d = EXCLUDED.avg_stock_units_30d, avg_stock_cost_30d = EXCLUDED.avg_stock_cost_30d, avg_stock_retail_30d = EXCLUDED.avg_stock_retail_30d, avg_stock_gross_30d = EXCLUDED.avg_stock_gross_30d,
received_qty_30d = EXCLUDED.received_qty_30d, received_cost_30d = EXCLUDED.received_cost_30d,
lifetime_sales = EXCLUDED.lifetime_sales, lifetime_revenue = EXCLUDED.lifetime_revenue,
first_7_days_sales = EXCLUDED.first_7_days_sales, first_7_days_revenue = EXCLUDED.first_7_days_revenue, first_30_days_sales = EXCLUDED.first_30_days_sales, first_30_days_revenue = EXCLUDED.first_30_days_revenue,
first_60_days_sales = EXCLUDED.first_60_days_sales, first_60_days_revenue = EXCLUDED.first_60_days_revenue, first_90_days_sales = EXCLUDED.first_90_days_sales, first_90_days_revenue = EXCLUDED.first_90_days_revenue,
asp_30d = EXCLUDED.asp_30d, acp_30d = EXCLUDED.acp_30d, avg_ros_30d = EXCLUDED.avg_ros_30d, avg_sales_per_day_30d = EXCLUDED.avg_sales_per_day_30d,
-- *** REMOVED avg_sales_per_month_30d ***
margin_30d = EXCLUDED.margin_30d, markup_30d = EXCLUDED.markup_30d, gmroi_30d = EXCLUDED.gmroi_30d, stockturn_30d = EXCLUDED.stockturn_30d, return_rate_30d = EXCLUDED.return_rate_30d, discount_rate_30d = EXCLUDED.discount_rate_30d,
stockout_rate_30d = EXCLUDED.stockout_rate_30d, markdown_30d = EXCLUDED.markdown_30d, markdown_rate_30d = EXCLUDED.markdown_rate_30d, sell_through_30d = EXCLUDED.sell_through_30d,
avg_lead_time_days = EXCLUDED.avg_lead_time_days, abc_class = EXCLUDED.abc_class,
sales_velocity_daily = EXCLUDED.sales_velocity_daily, config_lead_time = EXCLUDED.config_lead_time, config_days_of_stock = EXCLUDED.config_days_of_stock, config_safety_stock = EXCLUDED.config_safety_stock,
planning_period_days = EXCLUDED.planning_period_days, lead_time_forecast_units = EXCLUDED.lead_time_forecast_units, days_of_stock_forecast_units = EXCLUDED.days_of_stock_forecast_units,
planning_period_forecast_units = EXCLUDED.planning_period_forecast_units, lead_time_closing_stock = EXCLUDED.lead_time_closing_stock, days_of_stock_closing_stock = EXCLUDED.days_of_stock_closing_stock,
replenishment_needed_raw = EXCLUDED.replenishment_needed_raw, replenishment_units = EXCLUDED.replenishment_units, replenishment_cost = EXCLUDED.replenishment_cost, replenishment_retail = EXCLUDED.replenishment_retail, replenishment_profit = EXCLUDED.replenishment_profit,
to_order_units = EXCLUDED.to_order_units, -- *** Update to use EXCLUDED ***
forecast_lost_sales_units = EXCLUDED.forecast_lost_sales_units, forecast_lost_revenue = EXCLUDED.forecast_lost_revenue,
stock_cover_in_days = EXCLUDED.stock_cover_in_days, po_cover_in_days = EXCLUDED.po_cover_in_days, sells_out_in_days = EXCLUDED.sells_out_in_days, replenish_date = EXCLUDED.replenish_date,
overstocked_units = EXCLUDED.overstocked_units, overstocked_cost = EXCLUDED.overstocked_cost, overstocked_retail = EXCLUDED.overstocked_retail, is_old_stock = EXCLUDED.is_old_stock,
yesterday_sales = EXCLUDED.yesterday_sales;
RAISE NOTICE 'Finished % module. Duration: %', _module_name, clock_timestamp() - _start_time;
END $$;

View File

@@ -1,306 +0,0 @@
const path = require('path');
const fs = require('fs');
// Change working directory to script directory
process.chdir(path.dirname(__filename));
// Load environment variables
require('dotenv').config({ path: path.resolve(__dirname, '../..', '.env') });
// Add error handler for uncaught exceptions
process.on('uncaughtException', (error) => {
console.error('Uncaught Exception:', error);
process.exit(1);
});
// Add error handler for unhandled promise rejections
process.on('unhandledRejection', (reason, promise) => {
console.error('Unhandled Rejection at:', promise, 'reason:', reason);
process.exit(1);
});
// Load progress module
const progress = require('./utils/progress');
// Store progress functions in global scope to ensure availability
global.formatElapsedTime = progress.formatElapsedTime;
global.estimateRemaining = progress.estimateRemaining;
global.calculateRate = progress.calculateRate;
global.outputProgress = progress.outputProgress;
global.clearProgress = progress.clearProgress;
global.getProgress = progress.getProgress;
global.logError = progress.logError;
// Load database module
const { getConnection, closePool } = require('./utils/db');
// Add cancel handler
let isCancelled = false;
function cancelCalculation() {
isCancelled = true;
console.log('Calculation has been cancelled by user');
// Force-terminate any query that's been running for more than 5 seconds
try {
const connection = getConnection();
connection.then(async (conn) => {
try {
// Identify and terminate long-running queries from our application
await conn.query(`
SELECT pg_cancel_backend(pid)
FROM pg_stat_activity
WHERE query_start < now() - interval '5 seconds'
AND application_name LIKE '%node%'
AND query NOT LIKE '%pg_cancel_backend%'
`);
// Release connection
conn.release();
} catch (err) {
console.error('Error during force cancellation:', err);
conn.release();
}
}).catch(err => {
console.error('Could not get connection for cancellation:', err);
});
} catch (err) {
console.error('Failed to terminate running queries:', err);
}
return {
success: true,
message: 'Calculation has been cancelled'
};
}
// Handle SIGTERM signal for cancellation
process.on('SIGTERM', cancelCalculation);
async function updateDailySnapshots() {
let connection;
const startTime = Date.now();
let calculateHistoryId;
// Set a maximum execution time (30 minutes)
const MAX_EXECUTION_TIME = 30 * 60 * 1000;
const timeout = setTimeout(() => {
console.error(`Calculation timed out after ${MAX_EXECUTION_TIME/1000} seconds, forcing termination`);
// Call cancel and force exit
cancelCalculation();
process.exit(1);
}, MAX_EXECUTION_TIME);
try {
// Read the SQL file
const sqlFilePath = path.resolve(__dirname, 'update_daily_snapshots.sql');
const sqlQuery = fs.readFileSync(sqlFilePath, 'utf8');
// Clean up any previously running calculations
connection = await getConnection();
// Ensure the calculate_status table exists and has the correct structure
await connection.query(`
CREATE TABLE IF NOT EXISTS calculate_status (
module_name TEXT PRIMARY KEY,
last_calculation_timestamp TIMESTAMP WITH TIME ZONE NOT NULL DEFAULT CURRENT_TIMESTAMP
)
`);
await connection.query(`
UPDATE calculate_history
SET
status = 'cancelled',
end_time = NOW(),
duration_seconds = EXTRACT(EPOCH FROM (NOW() - start_time))::INTEGER,
error_message = 'Previous calculation was not completed properly'
WHERE status = 'running' AND additional_info->>'type' = 'daily_snapshots'
`);
// Create history record for this calculation
const historyResult = await connection.query(`
INSERT INTO calculate_history (
start_time,
status,
additional_info
) VALUES (
NOW(),
'running',
jsonb_build_object(
'type', 'daily_snapshots',
'sql_file', 'update_daily_snapshots.sql'
)
) RETURNING id
`);
calculateHistoryId = historyResult.rows[0].id;
// Initialize progress
global.outputProgress({
status: 'running',
operation: 'Starting daily snapshots calculation',
current: 0,
total: 100,
elapsed: '0s',
remaining: 'Calculating...',
rate: 0,
percentage: '0',
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
// Execute the SQL query
global.outputProgress({
status: 'running',
operation: 'Executing daily snapshots SQL query',
current: 25,
total: 100,
elapsed: global.formatElapsedTime(startTime),
remaining: 'Calculating...',
rate: 0,
percentage: '25',
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
await connection.query(sqlQuery);
// Update calculate_status table
await connection.query(`
INSERT INTO calculate_status (module_name, last_calculation_timestamp)
VALUES ($1, $2)
ON CONFLICT (module_name) DO UPDATE
SET last_calculation_timestamp = EXCLUDED.last_calculation_timestamp
`, ['daily_snapshots', new Date()]);
// Update progress to 100%
global.outputProgress({
status: 'complete',
operation: 'Daily snapshots calculation complete',
current: 100,
total: 100,
elapsed: global.formatElapsedTime(startTime),
remaining: '0s',
rate: 0,
percentage: '100',
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
// Update history with completion
await connection.query(`
UPDATE calculate_history
SET
end_time = NOW(),
duration_seconds = $1,
status = 'completed'
WHERE id = $2
`, [Math.round((Date.now() - startTime) / 1000), calculateHistoryId]);
// Clear progress file on successful completion
global.clearProgress();
return {
success: true,
message: 'Daily snapshots calculation completed successfully',
duration: Math.round((Date.now() - startTime) / 1000)
};
} catch (error) {
const endTime = Date.now();
const totalElapsedSeconds = Math.round((endTime - startTime) / 1000);
// Update history with error
if (connection && calculateHistoryId) {
await connection.query(`
UPDATE calculate_history
SET
end_time = NOW(),
duration_seconds = $1,
status = $2,
error_message = $3
WHERE id = $4
`, [
totalElapsedSeconds,
isCancelled ? 'cancelled' : 'failed',
error.message,
calculateHistoryId
]);
}
if (isCancelled) {
global.outputProgress({
status: 'cancelled',
operation: 'Calculation cancelled',
current: 50,
total: 100,
elapsed: global.formatElapsedTime(startTime),
remaining: null,
rate: 0,
percentage: '50',
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
} else {
global.outputProgress({
status: 'error',
operation: 'Error: ' + error.message,
current: 50,
total: 100,
elapsed: global.formatElapsedTime(startTime),
remaining: null,
rate: 0,
percentage: '50',
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
}
throw error;
} finally {
// Clear the timeout to prevent forced termination
clearTimeout(timeout);
// Always release connection
if (connection) {
try {
connection.release();
} catch (err) {
console.error('Error in final cleanup:', err);
}
}
}
}
// Export as a module with all necessary functions
module.exports = {
updateDailySnapshots,
cancelCalculation,
getProgress: global.getProgress
};
// Run directly if called from command line
if (require.main === module) {
updateDailySnapshots().then(() => {
closePool().then(() => {
process.exit(0);
});
}).catch(error => {
console.error('Error:', error);
closePool().then(() => {
process.exit(1);
});
});
}

View File

@@ -1,306 +0,0 @@
const path = require('path');
const fs = require('fs');
// Change working directory to script directory
process.chdir(path.dirname(__filename));
// Load environment variables
require('dotenv').config({ path: path.resolve(__dirname, '../..', '.env') });
// Add error handler for uncaught exceptions
process.on('uncaughtException', (error) => {
console.error('Uncaught Exception:', error);
process.exit(1);
});
// Add error handler for unhandled promise rejections
process.on('unhandledRejection', (reason, promise) => {
console.error('Unhandled Rejection at:', promise, 'reason:', reason);
process.exit(1);
});
// Load progress module
const progress = require('./utils/progress');
// Store progress functions in global scope to ensure availability
global.formatElapsedTime = progress.formatElapsedTime;
global.estimateRemaining = progress.estimateRemaining;
global.calculateRate = progress.calculateRate;
global.outputProgress = progress.outputProgress;
global.clearProgress = progress.clearProgress;
global.getProgress = progress.getProgress;
global.logError = progress.logError;
// Load database module
const { getConnection, closePool } = require('./utils/db');
// Add cancel handler
let isCancelled = false;
function cancelCalculation() {
isCancelled = true;
console.log('Calculation has been cancelled by user');
// Force-terminate any query that's been running for more than 5 seconds
try {
const connection = getConnection();
connection.then(async (conn) => {
try {
// Identify and terminate long-running queries from our application
await conn.query(`
SELECT pg_cancel_backend(pid)
FROM pg_stat_activity
WHERE query_start < now() - interval '5 seconds'
AND application_name LIKE '%node%'
AND query NOT LIKE '%pg_cancel_backend%'
`);
// Release connection
conn.release();
} catch (err) {
console.error('Error during force cancellation:', err);
conn.release();
}
}).catch(err => {
console.error('Could not get connection for cancellation:', err);
});
} catch (err) {
console.error('Failed to terminate running queries:', err);
}
return {
success: true,
message: 'Calculation has been cancelled'
};
}
// Handle SIGTERM signal for cancellation
process.on('SIGTERM', cancelCalculation);
async function updatePeriodicMetrics() {
let connection;
const startTime = Date.now();
let calculateHistoryId;
// Set a maximum execution time (30 minutes)
const MAX_EXECUTION_TIME = 30 * 60 * 1000;
const timeout = setTimeout(() => {
console.error(`Calculation timed out after ${MAX_EXECUTION_TIME/1000} seconds, forcing termination`);
// Call cancel and force exit
cancelCalculation();
process.exit(1);
}, MAX_EXECUTION_TIME);
try {
// Read the SQL file
const sqlFilePath = path.resolve(__dirname, 'update_periodic_metrics.sql');
const sqlQuery = fs.readFileSync(sqlFilePath, 'utf8');
// Clean up any previously running calculations
connection = await getConnection();
// Ensure the calculate_status table exists and has the correct structure
await connection.query(`
CREATE TABLE IF NOT EXISTS calculate_status (
module_name TEXT PRIMARY KEY,
last_calculation_timestamp TIMESTAMP WITH TIME ZONE NOT NULL DEFAULT CURRENT_TIMESTAMP
)
`);
await connection.query(`
UPDATE calculate_history
SET
status = 'cancelled',
end_time = NOW(),
duration_seconds = EXTRACT(EPOCH FROM (NOW() - start_time))::INTEGER,
error_message = 'Previous calculation was not completed properly'
WHERE status = 'running' AND additional_info->>'type' = 'periodic_metrics'
`);
// Create history record for this calculation
const historyResult = await connection.query(`
INSERT INTO calculate_history (
start_time,
status,
additional_info
) VALUES (
NOW(),
'running',
jsonb_build_object(
'type', 'periodic_metrics',
'sql_file', 'update_periodic_metrics.sql'
)
) RETURNING id
`);
calculateHistoryId = historyResult.rows[0].id;
// Initialize progress
global.outputProgress({
status: 'running',
operation: 'Starting periodic metrics calculation',
current: 0,
total: 100,
elapsed: '0s',
remaining: 'Calculating...',
rate: 0,
percentage: '0',
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
// Execute the SQL query
global.outputProgress({
status: 'running',
operation: 'Executing periodic metrics SQL query',
current: 25,
total: 100,
elapsed: global.formatElapsedTime(startTime),
remaining: 'Calculating...',
rate: 0,
percentage: '25',
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
await connection.query(sqlQuery);
// Update calculate_status table
await connection.query(`
INSERT INTO calculate_status (module_name, last_calculation_timestamp)
VALUES ($1, $2)
ON CONFLICT (module_name) DO UPDATE
SET last_calculation_timestamp = EXCLUDED.last_calculation_timestamp
`, ['periodic_metrics', new Date()]);
// Update progress to 100%
global.outputProgress({
status: 'complete',
operation: 'Periodic metrics calculation complete',
current: 100,
total: 100,
elapsed: global.formatElapsedTime(startTime),
remaining: '0s',
rate: 0,
percentage: '100',
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
// Update history with completion
await connection.query(`
UPDATE calculate_history
SET
end_time = NOW(),
duration_seconds = $1,
status = 'completed'
WHERE id = $2
`, [Math.round((Date.now() - startTime) / 1000), calculateHistoryId]);
// Clear progress file on successful completion
global.clearProgress();
return {
success: true,
message: 'Periodic metrics calculation completed successfully',
duration: Math.round((Date.now() - startTime) / 1000)
};
} catch (error) {
const endTime = Date.now();
const totalElapsedSeconds = Math.round((endTime - startTime) / 1000);
// Update history with error
if (connection && calculateHistoryId) {
await connection.query(`
UPDATE calculate_history
SET
end_time = NOW(),
duration_seconds = $1,
status = $2,
error_message = $3
WHERE id = $4
`, [
totalElapsedSeconds,
isCancelled ? 'cancelled' : 'failed',
error.message,
calculateHistoryId
]);
}
if (isCancelled) {
global.outputProgress({
status: 'cancelled',
operation: 'Calculation cancelled',
current: 50,
total: 100,
elapsed: global.formatElapsedTime(startTime),
remaining: null,
rate: 0,
percentage: '50',
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
} else {
global.outputProgress({
status: 'error',
operation: 'Error: ' + error.message,
current: 50,
total: 100,
elapsed: global.formatElapsedTime(startTime),
remaining: null,
rate: 0,
percentage: '50',
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
}
throw error;
} finally {
// Clear the timeout to prevent forced termination
clearTimeout(timeout);
// Always release connection
if (connection) {
try {
connection.release();
} catch (err) {
console.error('Error in final cleanup:', err);
}
}
}
}
// Export as a module with all necessary functions
module.exports = {
updatePeriodicMetrics,
cancelCalculation,
getProgress: global.getProgress
};
// Run directly if called from command line
if (require.main === module) {
updatePeriodicMetrics().then(() => {
closePool().then(() => {
process.exit(0);
});
}).catch(error => {
console.error('Error:', error);
closePool().then(() => {
process.exit(1);
});
});
}

View File

@@ -1,306 +0,0 @@
const path = require('path');
const fs = require('fs');
// Change working directory to script directory
process.chdir(path.dirname(__filename));
// Load environment variables
require('dotenv').config({ path: path.resolve(__dirname, '../..', '.env') });
// Add error handler for uncaught exceptions
process.on('uncaughtException', (error) => {
console.error('Uncaught Exception:', error);
process.exit(1);
});
// Add error handler for unhandled promise rejections
process.on('unhandledRejection', (reason, promise) => {
console.error('Unhandled Rejection at:', promise, 'reason:', reason);
process.exit(1);
});
// Load progress module
const progress = require('./utils/progress');
// Store progress functions in global scope to ensure availability
global.formatElapsedTime = progress.formatElapsedTime;
global.estimateRemaining = progress.estimateRemaining;
global.calculateRate = progress.calculateRate;
global.outputProgress = progress.outputProgress;
global.clearProgress = progress.clearProgress;
global.getProgress = progress.getProgress;
global.logError = progress.logError;
// Load database module
const { getConnection, closePool } = require('./utils/db');
// Add cancel handler
let isCancelled = false;
function cancelCalculation() {
isCancelled = true;
console.log('Calculation has been cancelled by user');
// Force-terminate any query that's been running for more than 5 seconds
try {
const connection = getConnection();
connection.then(async (conn) => {
try {
// Identify and terminate long-running queries from our application
await conn.query(`
SELECT pg_cancel_backend(pid)
FROM pg_stat_activity
WHERE query_start < now() - interval '5 seconds'
AND application_name LIKE '%node%'
AND query NOT LIKE '%pg_cancel_backend%'
`);
// Release connection
conn.release();
} catch (err) {
console.error('Error during force cancellation:', err);
conn.release();
}
}).catch(err => {
console.error('Could not get connection for cancellation:', err);
});
} catch (err) {
console.error('Failed to terminate running queries:', err);
}
return {
success: true,
message: 'Calculation has been cancelled'
};
}
// Handle SIGTERM signal for cancellation
process.on('SIGTERM', cancelCalculation);
async function updateProductMetrics() {
let connection;
const startTime = Date.now();
let calculateHistoryId;
// Set a maximum execution time (30 minutes)
const MAX_EXECUTION_TIME = 30 * 60 * 1000;
const timeout = setTimeout(() => {
console.error(`Calculation timed out after ${MAX_EXECUTION_TIME/1000} seconds, forcing termination`);
// Call cancel and force exit
cancelCalculation();
process.exit(1);
}, MAX_EXECUTION_TIME);
try {
// Read the SQL file
const sqlFilePath = path.resolve(__dirname, 'update_product_metrics.sql');
const sqlQuery = fs.readFileSync(sqlFilePath, 'utf8');
// Clean up any previously running calculations
connection = await getConnection();
// Ensure the calculate_status table exists and has the correct structure
await connection.query(`
CREATE TABLE IF NOT EXISTS calculate_status (
module_name TEXT PRIMARY KEY,
last_calculation_timestamp TIMESTAMP WITH TIME ZONE NOT NULL DEFAULT CURRENT_TIMESTAMP
)
`);
await connection.query(`
UPDATE calculate_history
SET
status = 'cancelled',
end_time = NOW(),
duration_seconds = EXTRACT(EPOCH FROM (NOW() - start_time))::INTEGER,
error_message = 'Previous calculation was not completed properly'
WHERE status = 'running' AND additional_info->>'type' = 'product_metrics'
`);
// Create history record for this calculation
const historyResult = await connection.query(`
INSERT INTO calculate_history (
start_time,
status,
additional_info
) VALUES (
NOW(),
'running',
jsonb_build_object(
'type', 'product_metrics',
'sql_file', 'update_product_metrics.sql'
)
) RETURNING id
`);
calculateHistoryId = historyResult.rows[0].id;
// Initialize progress
global.outputProgress({
status: 'running',
operation: 'Starting product metrics calculation',
current: 0,
total: 100,
elapsed: '0s',
remaining: 'Calculating...',
rate: 0,
percentage: '0',
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
// Execute the SQL query
global.outputProgress({
status: 'running',
operation: 'Executing product metrics SQL query',
current: 25,
total: 100,
elapsed: global.formatElapsedTime(startTime),
remaining: 'Calculating...',
rate: 0,
percentage: '25',
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
await connection.query(sqlQuery);
// Update calculate_status table
await connection.query(`
INSERT INTO calculate_status (module_name, last_calculation_timestamp)
VALUES ($1, $2)
ON CONFLICT (module_name) DO UPDATE
SET last_calculation_timestamp = EXCLUDED.last_calculation_timestamp
`, ['product_metrics', new Date()]);
// Update progress to 100%
global.outputProgress({
status: 'complete',
operation: 'Product metrics calculation complete',
current: 100,
total: 100,
elapsed: global.formatElapsedTime(startTime),
remaining: '0s',
rate: 0,
percentage: '100',
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
// Update history with completion
await connection.query(`
UPDATE calculate_history
SET
end_time = NOW(),
duration_seconds = $1,
status = 'completed'
WHERE id = $2
`, [Math.round((Date.now() - startTime) / 1000), calculateHistoryId]);
// Clear progress file on successful completion
global.clearProgress();
return {
success: true,
message: 'Product metrics calculation completed successfully',
duration: Math.round((Date.now() - startTime) / 1000)
};
} catch (error) {
const endTime = Date.now();
const totalElapsedSeconds = Math.round((endTime - startTime) / 1000);
// Update history with error
if (connection && calculateHistoryId) {
await connection.query(`
UPDATE calculate_history
SET
end_time = NOW(),
duration_seconds = $1,
status = $2,
error_message = $3
WHERE id = $4
`, [
totalElapsedSeconds,
isCancelled ? 'cancelled' : 'failed',
error.message,
calculateHistoryId
]);
}
if (isCancelled) {
global.outputProgress({
status: 'cancelled',
operation: 'Calculation cancelled',
current: 50,
total: 100,
elapsed: global.formatElapsedTime(startTime),
remaining: null,
rate: 0,
percentage: '50',
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
} else {
global.outputProgress({
status: 'error',
operation: 'Error: ' + error.message,
current: 50,
total: 100,
elapsed: global.formatElapsedTime(startTime),
remaining: null,
rate: 0,
percentage: '50',
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
}
throw error;
} finally {
// Clear the timeout to prevent forced termination
clearTimeout(timeout);
// Always release connection
if (connection) {
try {
connection.release();
} catch (err) {
console.error('Error in final cleanup:', err);
}
}
}
}
// Export as a module with all necessary functions
module.exports = {
updateProductMetrics,
cancelCalculation,
getProgress: global.getProgress
};
// Run directly if called from command line
if (require.main === module) {
updateProductMetrics().then(() => {
closePool().then(() => {
process.exit(0);
});
}).catch(error => {
console.error('Error:', error);
closePool().then(() => {
process.exit(1);
});
});
}

View File

@@ -35,18 +35,59 @@ BEGIN
FROM public.products p -- Start from products to include those with no orders today
LEFT JOIN public.orders o
ON p.pid = o.pid
AND o.date >= _target_date -- Filter orders for the target date
AND o.date < _target_date + INTERVAL '1 day'
AND o.date::date = _target_date -- Cast to date to ensure compatibility regardless of original type
GROUP BY p.pid, p.sku
),
ReceivingData AS (
SELECT
po.pid,
COALESCE(SUM((rh.item->>'qty')::numeric), 0) AS units_received,
COALESCE(SUM((rh.item->>'qty')::numeric * COALESCE((rh.item->>'cost')::numeric, po.cost_price)), 0.00) AS cost_received
-- Prioritize the actual table fields over the JSON data
COALESCE(
-- First try the received field from purchase_orders table
SUM(CASE WHEN po.date::date = _target_date THEN po.received ELSE 0 END),
-- Otherwise fall back to the receiving_history JSON as secondary source
SUM(
CASE
WHEN (rh.item->>'date')::date = _target_date THEN (rh.item->>'qty')::numeric
WHEN (rh.item->>'received_at')::date = _target_date THEN (rh.item->>'qty')::numeric
WHEN (rh.item->>'receipt_date')::date = _target_date THEN (rh.item->>'qty')::numeric
ELSE 0
END
),
0
) AS units_received,
COALESCE(
-- First try the actual cost_price from purchase_orders
SUM(CASE WHEN po.date::date = _target_date THEN po.received * po.cost_price ELSE 0 END),
-- Otherwise fall back to receiving_history JSON
SUM(
CASE
WHEN (rh.item->>'date')::date = _target_date THEN (rh.item->>'qty')::numeric
WHEN (rh.item->>'received_at')::date = _target_date THEN (rh.item->>'qty')::numeric
WHEN (rh.item->>'receipt_date')::date = _target_date THEN (rh.item->>'qty')::numeric
ELSE 0
END
* COALESCE((rh.item->>'cost')::numeric, po.cost_price)
),
0.00
) AS cost_received
FROM public.purchase_orders po
CROSS JOIN LATERAL jsonb_array_elements(po.receiving_history) AS rh(item)
WHERE (rh.item->>'received_at')::date = _target_date -- Filter receipts for the target date
LEFT JOIN LATERAL jsonb_array_elements(po.receiving_history) AS rh(item) ON
jsonb_typeof(po.receiving_history) = 'array' AND
jsonb_array_length(po.receiving_history) > 0 AND
(
(rh.item->>'date')::date = _target_date OR
(rh.item->>'received_at')::date = _target_date OR
(rh.item->>'receipt_date')::date = _target_date
)
-- Include POs with the current date or relevant receiving_history
WHERE
po.date::date = _target_date OR
jsonb_typeof(po.receiving_history) = 'array' AND
jsonb_array_length(po.receiving_history) > 0
GROUP BY po.pid
),
CurrentStock AS (

View File

@@ -58,15 +58,41 @@ BEGIN
p.pid,
MIN(o.date)::date AS date_first_sold,
MAX(o.date)::date AS max_order_date, -- Use MAX for potential recalc of date_last_sold
MIN(rh.first_receipt_date) AS date_first_received_calc,
MAX(rh.last_receipt_date) AS date_last_received_calc
-- For first received date, try table data first then fall back to JSON
COALESCE(
MIN(po.date)::date, -- Try purchase_order date first
MIN(rh.first_receipt_date) -- Fall back to JSON data if needed
) AS date_first_received_calc,
-- If we only have one receipt date (first = last), use that for last_received too
COALESCE(
MAX(po.date)::date, -- Try purchase_order date first
NULLIF(MAX(rh.last_receipt_date), NULL),
MIN(rh.first_receipt_date)
) AS date_last_received_calc
FROM public.products p
LEFT JOIN public.orders o ON p.pid = o.pid AND o.quantity > 0 AND o.status NOT IN ('canceled', 'returned')
LEFT JOIN public.purchase_orders po ON p.pid = po.pid AND po.received > 0
LEFT JOIN (
SELECT
po.pid,
MIN((rh.item->>'received_at')::date) as first_receipt_date,
MAX((rh.item->>'received_at')::date) as last_receipt_date
MIN(
CASE
WHEN rh.item->>'date' IS NOT NULL THEN (rh.item->>'date')::date
WHEN rh.item->>'received_at' IS NOT NULL THEN (rh.item->>'received_at')::date
WHEN rh.item->>'receipt_date' IS NOT NULL THEN (rh.item->>'receipt_date')::date
ELSE NULL
END
) as first_receipt_date,
MAX(
CASE
WHEN rh.item->>'date' IS NOT NULL THEN (rh.item->>'date')::date
WHEN rh.item->>'received_at' IS NOT NULL THEN (rh.item->>'received_at')::date
WHEN rh.item->>'receipt_date' IS NOT NULL THEN (rh.item->>'receipt_date')::date
ELSE NULL
END
) as last_receipt_date
FROM public.purchase_orders po
CROSS JOIN LATERAL jsonb_array_elements(po.receiving_history) AS rh(item)
WHERE jsonb_typeof(po.receiving_history) = 'array' AND jsonb_array_length(po.receiving_history) > 0
@@ -77,7 +103,14 @@ BEGIN
SnapshotAggregates AS (
SELECT
pid,
-- Rolling periods (ensure dates are inclusive/exclusive as needed)
-- Get the counts of all available data
COUNT(DISTINCT snapshot_date) AS available_days,
-- Rolling periods with no time constraint - just sum everything we have
SUM(units_sold) AS total_units_sold,
SUM(net_revenue) AS total_net_revenue,
-- Specific time windows if we have enough data
SUM(CASE WHEN snapshot_date >= _current_date - INTERVAL '6 days' THEN units_sold ELSE 0 END) AS sales_7d,
SUM(CASE WHEN snapshot_date >= _current_date - INTERVAL '6 days' THEN net_revenue ELSE 0 END) AS revenue_7d,
SUM(CASE WHEN snapshot_date >= _current_date - INTERVAL '13 days' THEN units_sold ELSE 0 END) AS sales_14d,
@@ -103,7 +136,7 @@ BEGIN
AVG(CASE WHEN snapshot_date >= _current_date - INTERVAL '29 days' THEN eod_stock_retail END) AS avg_stock_retail_30d,
AVG(CASE WHEN snapshot_date >= _current_date - INTERVAL '29 days' THEN eod_stock_gross END) AS avg_stock_gross_30d,
-- Lifetime
-- Lifetime - should match total values above
SUM(units_sold) AS lifetime_sales,
SUM(net_revenue) AS lifetime_revenue,
@@ -111,8 +144,6 @@ BEGIN
SUM(CASE WHEN snapshot_date = _current_date - INTERVAL '1 day' THEN units_sold ELSE 0 END) as yesterday_sales
FROM public.daily_product_snapshots
WHERE snapshot_date <= _current_date -- Include today's snapshot
AND snapshot_date >= _current_date - INTERVAL '365 days' -- Limit history scan slightly
GROUP BY pid
),
FirstPeriodMetrics AS (
@@ -191,7 +222,9 @@ BEGIN
sa.stockout_days_30d, sa.sales_365d, sa.revenue_365d,
sa.avg_stock_units_30d, sa.avg_stock_cost_30d, sa.avg_stock_retail_30d, sa.avg_stock_gross_30d,
sa.received_qty_30d, sa.received_cost_30d,
sa.lifetime_sales, sa.lifetime_revenue,
-- Use total counts for lifetime values to ensure we have data even with limited history
COALESCE(sa.total_units_sold, sa.lifetime_sales) AS lifetime_sales,
COALESCE(sa.total_net_revenue, sa.lifetime_revenue) AS lifetime_revenue,
fpm.first_7_days_sales, fpm.first_7_days_revenue, fpm.first_30_days_sales, fpm.first_30_days_revenue,
fpm.first_60_days_sales, fpm.first_60_days_revenue, fpm.first_90_days_sales, fpm.first_90_days_revenue,

View File

@@ -1,321 +0,0 @@
const { outputProgress, formatElapsedTime, estimateRemaining, calculateRate, logError } = require('./utils/progress');
const { getConnection } = require('./utils/db');
async function calculateBrandMetrics(startTime, totalProducts, processedCount = 0, isCancelled = false) {
const connection = await getConnection();
let success = false;
let processedOrders = 0;
try {
if (isCancelled) {
outputProgress({
status: 'cancelled',
operation: 'Brand metrics calculation cancelled',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: null,
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
return {
processedProducts: processedCount,
processedOrders: 0,
processedPurchaseOrders: 0,
success
};
}
// Get order count that will be processed
const orderCount = await connection.query(`
SELECT COUNT(*) as count
FROM orders o
WHERE o.canceled = false
`);
processedOrders = parseInt(orderCount.rows[0].count);
outputProgress({
status: 'running',
operation: 'Starting brand metrics calculation',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount, totalProducts),
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
// Calculate brand metrics with optimized queries
await connection.query(`
INSERT INTO brand_metrics (
brand,
product_count,
active_products,
total_stock_units,
total_stock_cost,
total_stock_retail,
total_revenue,
avg_margin,
growth_rate
)
WITH filtered_products AS (
SELECT
p.*,
CASE
WHEN p.stock_quantity <= 5000 AND p.stock_quantity >= 0
THEN p.pid
END as valid_pid,
CASE
WHEN p.visible = true
AND p.stock_quantity <= 5000
AND p.stock_quantity >= 0
THEN p.pid
END as active_pid,
CASE
WHEN p.stock_quantity IS NULL
OR p.stock_quantity < 0
OR p.stock_quantity > 5000
THEN 0
ELSE p.stock_quantity
END as valid_stock
FROM products p
WHERE p.brand IS NOT NULL
),
sales_periods AS (
SELECT
p.brand,
SUM(o.quantity * (o.price - COALESCE(o.discount, 0))) as period_revenue,
SUM(o.quantity * (o.price - COALESCE(o.discount, 0) - p.cost_price)) as period_margin,
COUNT(DISTINCT DATE(o.date)) as period_days,
CASE
WHEN o.date >= CURRENT_DATE - INTERVAL '3 months' THEN 'current'
WHEN o.date BETWEEN CURRENT_DATE - INTERVAL '15 months'
AND CURRENT_DATE - INTERVAL '12 months' THEN 'previous'
END as period_type
FROM filtered_products p
JOIN orders o ON p.pid = o.pid
WHERE o.canceled = false
AND o.date >= CURRENT_DATE - INTERVAL '15 months'
GROUP BY p.brand, period_type
),
brand_data AS (
SELECT
p.brand,
COUNT(DISTINCT p.valid_pid) as product_count,
COUNT(DISTINCT p.active_pid) as active_products,
SUM(p.valid_stock) as total_stock_units,
SUM(p.valid_stock * p.cost_price) as total_stock_cost,
SUM(p.valid_stock * p.price) as total_stock_retail,
COALESCE(SUM(o.quantity * (o.price - COALESCE(o.discount, 0))), 0) as total_revenue,
CASE
WHEN SUM(o.quantity * o.price) > 0
THEN GREATEST(
-100.0,
LEAST(
100.0,
(
SUM(o.quantity * o.price) - -- Use gross revenue (before discounts)
SUM(o.quantity * COALESCE(p.cost_price, 0)) -- Total costs
) * 100.0 /
NULLIF(SUM(o.quantity * o.price), 0) -- Divide by gross revenue
)
)
ELSE 0
END as avg_margin
FROM filtered_products p
LEFT JOIN orders o ON p.pid = o.pid AND o.canceled = false
GROUP BY p.brand
)
SELECT
bd.brand,
bd.product_count,
bd.active_products,
bd.total_stock_units,
bd.total_stock_cost,
bd.total_stock_retail,
bd.total_revenue,
bd.avg_margin,
CASE
WHEN MAX(CASE WHEN sp.period_type = 'previous' THEN sp.period_revenue END) = 0
AND MAX(CASE WHEN sp.period_type = 'current' THEN sp.period_revenue END) > 0
THEN 100.0
WHEN MAX(CASE WHEN sp.period_type = 'previous' THEN sp.period_revenue END) = 0
THEN 0.0
ELSE GREATEST(
-100.0,
LEAST(
((MAX(CASE WHEN sp.period_type = 'current' THEN sp.period_revenue END) -
MAX(CASE WHEN sp.period_type = 'previous' THEN sp.period_revenue END)) /
NULLIF(ABS(MAX(CASE WHEN sp.period_type = 'previous' THEN sp.period_revenue END)), 0)) * 100.0,
999.99
)
)
END as growth_rate
FROM brand_data bd
LEFT JOIN sales_periods sp ON bd.brand = sp.brand
GROUP BY bd.brand, bd.product_count, bd.active_products, bd.total_stock_units,
bd.total_stock_cost, bd.total_stock_retail, bd.total_revenue, bd.avg_margin
ON CONFLICT (brand) DO UPDATE
SET
product_count = EXCLUDED.product_count,
active_products = EXCLUDED.active_products,
total_stock_units = EXCLUDED.total_stock_units,
total_stock_cost = EXCLUDED.total_stock_cost,
total_stock_retail = EXCLUDED.total_stock_retail,
total_revenue = EXCLUDED.total_revenue,
avg_margin = EXCLUDED.avg_margin,
growth_rate = EXCLUDED.growth_rate,
last_calculated_at = CURRENT_TIMESTAMP
`);
processedCount = Math.floor(totalProducts * 0.97);
outputProgress({
status: 'running',
operation: 'Brand metrics calculated, starting time-based metrics',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount, totalProducts),
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
if (isCancelled) return {
processedProducts: processedCount,
processedOrders,
processedPurchaseOrders: 0,
success
};
// Calculate brand time-based metrics with optimized query
await connection.query(`
INSERT INTO brand_time_metrics (
brand,
year,
month,
product_count,
active_products,
total_stock_units,
total_stock_cost,
total_stock_retail,
total_revenue,
avg_margin
)
WITH filtered_products AS (
SELECT
p.*,
CASE WHEN p.stock_quantity <= 5000 THEN p.pid END as valid_pid,
CASE WHEN p.visible = true AND p.stock_quantity <= 5000 THEN p.pid END as active_pid,
CASE
WHEN p.stock_quantity IS NULL OR p.stock_quantity < 0 OR p.stock_quantity > 5000 THEN 0
ELSE p.stock_quantity
END as valid_stock
FROM products p
WHERE p.brand IS NOT NULL
),
monthly_metrics AS (
SELECT
p.brand,
EXTRACT(YEAR FROM o.date::timestamp with time zone) as year,
EXTRACT(MONTH FROM o.date::timestamp with time zone) as month,
COUNT(DISTINCT p.valid_pid) as product_count,
COUNT(DISTINCT p.active_pid) as active_products,
SUM(p.valid_stock) as total_stock_units,
SUM(p.valid_stock * p.cost_price) as total_stock_cost,
SUM(p.valid_stock * p.price) as total_stock_retail,
SUM(o.quantity * o.price) as total_revenue,
CASE
WHEN SUM(o.quantity * o.price) > 0
THEN GREATEST(
-100.0,
LEAST(
100.0,
(
SUM(o.quantity * o.price) - -- Use gross revenue (before discounts)
SUM(o.quantity * COALESCE(p.cost_price, 0)) -- Total costs
) * 100.0 /
NULLIF(SUM(o.quantity * o.price), 0) -- Divide by gross revenue
)
)
ELSE 0
END as avg_margin
FROM filtered_products p
LEFT JOIN orders o ON p.pid = o.pid AND o.canceled = false
WHERE o.date >= CURRENT_DATE - INTERVAL '12 months'
GROUP BY p.brand, EXTRACT(YEAR FROM o.date::timestamp with time zone), EXTRACT(MONTH FROM o.date::timestamp with time zone)
)
SELECT *
FROM monthly_metrics
ON CONFLICT (brand, year, month) DO UPDATE
SET
product_count = EXCLUDED.product_count,
active_products = EXCLUDED.active_products,
total_stock_units = EXCLUDED.total_stock_units,
total_stock_cost = EXCLUDED.total_stock_cost,
total_stock_retail = EXCLUDED.total_stock_retail,
total_revenue = EXCLUDED.total_revenue,
avg_margin = EXCLUDED.avg_margin
`);
processedCount = Math.floor(totalProducts * 0.99);
outputProgress({
status: 'running',
operation: 'Brand time-based metrics calculated',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount, totalProducts),
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
// If we get here, everything completed successfully
success = true;
// Update calculate_status
await connection.query(`
INSERT INTO calculate_status (module_name, last_calculation_timestamp)
VALUES ('brand_metrics', NOW())
ON CONFLICT (module_name) DO UPDATE
SET last_calculation_timestamp = NOW()
`);
return {
processedProducts: processedCount,
processedOrders,
processedPurchaseOrders: 0,
success
};
} catch (error) {
success = false;
logError(error, 'Error calculating brand metrics');
throw error;
} finally {
if (connection) {
connection.release();
}
}
}
module.exports = calculateBrandMetrics;

View File

@@ -1,554 +0,0 @@
const { outputProgress, formatElapsedTime, estimateRemaining, calculateRate, logError } = require('./utils/progress');
const { getConnection } = require('./utils/db');
async function calculateCategoryMetrics(startTime, totalProducts, processedCount = 0, isCancelled = false) {
const connection = await getConnection();
let success = false;
let processedOrders = 0;
try {
if (isCancelled) {
outputProgress({
status: 'cancelled',
operation: 'Category metrics calculation cancelled',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: null,
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
return {
processedProducts: processedCount,
processedOrders: 0,
processedPurchaseOrders: 0,
success
};
}
// Get order count that will be processed
const orderCount = await connection.query(`
SELECT COUNT(*) as count
FROM orders o
WHERE o.canceled = false
`);
processedOrders = parseInt(orderCount.rows[0].count);
outputProgress({
status: 'running',
operation: 'Starting category metrics calculation',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount, totalProducts),
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
// First, calculate base category metrics
await connection.query(`
INSERT INTO category_metrics (
category_id,
product_count,
active_products,
total_value,
status,
last_calculated_at
)
SELECT
c.cat_id,
COUNT(DISTINCT p.pid) as product_count,
COUNT(DISTINCT CASE WHEN p.visible = true THEN p.pid END) as active_products,
COALESCE(SUM(p.stock_quantity * p.cost_price), 0) as total_value,
c.status,
NOW() as last_calculated_at
FROM categories c
LEFT JOIN product_categories pc ON c.cat_id = pc.cat_id
LEFT JOIN products p ON pc.pid = p.pid
GROUP BY c.cat_id, c.status
ON CONFLICT (category_id) DO UPDATE
SET
product_count = EXCLUDED.product_count,
active_products = EXCLUDED.active_products,
total_value = EXCLUDED.total_value,
status = EXCLUDED.status,
last_calculated_at = EXCLUDED.last_calculated_at
`);
processedCount = Math.floor(totalProducts * 0.90);
outputProgress({
status: 'running',
operation: 'Base category metrics calculated, updating with margin data',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount, totalProducts),
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
if (isCancelled) return {
processedProducts: processedCount,
processedOrders,
processedPurchaseOrders: 0,
success
};
// Then update with margin and turnover data
await connection.query(`
WITH category_sales AS (
SELECT
pc.cat_id,
SUM(o.quantity * o.price) as total_sales,
SUM(o.quantity * (o.price - p.cost_price)) as total_margin,
SUM(o.quantity) as units_sold,
AVG(GREATEST(p.stock_quantity, 0)) as avg_stock,
COUNT(DISTINCT DATE(o.date)) as active_days
FROM product_categories pc
JOIN products p ON pc.pid = p.pid
JOIN orders o ON p.pid = o.pid
LEFT JOIN turnover_config tc ON
(tc.category_id = pc.cat_id AND tc.vendor = p.vendor) OR
(tc.category_id = pc.cat_id AND tc.vendor IS NULL) OR
(tc.category_id IS NULL AND tc.vendor = p.vendor) OR
(tc.category_id IS NULL AND tc.vendor IS NULL)
WHERE o.canceled = false
AND o.date >= CURRENT_DATE - (COALESCE(tc.calculation_period_days, 30) || ' days')::INTERVAL
GROUP BY pc.cat_id
)
UPDATE category_metrics
SET
avg_margin = COALESCE(cs.total_margin * 100.0 / NULLIF(cs.total_sales, 0), 0),
turnover_rate = CASE
WHEN cs.avg_stock > 0 AND cs.active_days > 0
THEN LEAST(
(cs.units_sold / cs.avg_stock) * (365.0 / cs.active_days),
999.99
)
ELSE 0
END,
last_calculated_at = NOW()
FROM category_sales cs
WHERE category_id = cs.cat_id
`);
processedCount = Math.floor(totalProducts * 0.95);
outputProgress({
status: 'running',
operation: 'Margin data updated, calculating growth rates',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount, totalProducts),
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
if (isCancelled) return {
processedProducts: processedCount,
processedOrders,
processedPurchaseOrders: 0,
success
};
// Finally update growth rates
await connection.query(`
WITH current_period AS (
SELECT
pc.cat_id,
SUM(o.quantity * (o.price - COALESCE(o.discount, 0)) /
(1 + COALESCE(ss.seasonality_factor, 0))) as revenue,
SUM(o.quantity * (o.price - COALESCE(o.discount, 0) - p.cost_price)) as gross_profit,
COUNT(DISTINCT DATE(o.date)) as days
FROM product_categories pc
JOIN products p ON pc.pid = p.pid
JOIN orders o ON p.pid = o.pid
LEFT JOIN sales_seasonality ss ON EXTRACT(MONTH FROM o.date) = ss.month
WHERE o.canceled = false
AND o.date >= CURRENT_DATE - INTERVAL '3 months'
GROUP BY pc.cat_id
),
previous_period AS (
SELECT
pc.cat_id,
SUM(o.quantity * (o.price - COALESCE(o.discount, 0)) /
(1 + COALESCE(ss.seasonality_factor, 0))) as revenue,
COUNT(DISTINCT DATE(o.date)) as days
FROM product_categories pc
JOIN products p ON pc.pid = p.pid
JOIN orders o ON p.pid = o.pid
LEFT JOIN sales_seasonality ss ON EXTRACT(MONTH FROM o.date) = ss.month
WHERE o.canceled = false
AND o.date BETWEEN CURRENT_DATE - INTERVAL '15 months'
AND CURRENT_DATE - INTERVAL '12 months'
GROUP BY pc.cat_id
),
trend_data AS (
SELECT
pc.cat_id,
EXTRACT(MONTH FROM o.date) as month,
SUM(o.quantity * (o.price - COALESCE(o.discount, 0)) /
(1 + COALESCE(ss.seasonality_factor, 0))) as revenue,
COUNT(DISTINCT DATE(o.date)) as days_in_month
FROM product_categories pc
JOIN products p ON pc.pid = p.pid
JOIN orders o ON p.pid = o.pid
LEFT JOIN sales_seasonality ss ON EXTRACT(MONTH FROM o.date) = ss.month
WHERE o.canceled = false
AND o.date >= CURRENT_DATE - INTERVAL '15 months'
GROUP BY pc.cat_id, EXTRACT(MONTH FROM o.date)
),
trend_stats AS (
SELECT
cat_id,
COUNT(*) as n,
AVG(month) as avg_x,
AVG(revenue / NULLIF(days_in_month, 0)) as avg_y,
SUM(month * (revenue / NULLIF(days_in_month, 0))) as sum_xy,
SUM(month * month) as sum_xx
FROM trend_data
GROUP BY cat_id
HAVING COUNT(*) >= 6
),
trend_analysis AS (
SELECT
cat_id,
((n * sum_xy) - (avg_x * n * avg_y)) /
NULLIF((n * sum_xx) - (n * avg_x * avg_x), 0) as trend_slope,
avg_y as avg_daily_revenue
FROM trend_stats
),
margin_calc AS (
SELECT
pc.cat_id,
CASE
WHEN SUM(o.quantity * o.price) > 0 THEN
GREATEST(
-100.0,
LEAST(
100.0,
(
SUM(o.quantity * o.price) - -- Use gross revenue (before discounts)
SUM(o.quantity * COALESCE(p.cost_price, 0)) -- Total costs
) * 100.0 /
NULLIF(SUM(o.quantity * o.price), 0) -- Divide by gross revenue
)
)
ELSE NULL
END as avg_margin
FROM product_categories pc
JOIN products p ON pc.pid = p.pid
JOIN orders o ON p.pid = o.pid
WHERE o.canceled = false
AND o.date >= CURRENT_DATE - INTERVAL '3 months'
GROUP BY pc.cat_id
),
combined_metrics AS (
SELECT
COALESCE(cp.cat_id, pp.cat_id) as category_id,
CASE
WHEN pp.revenue = 0 AND COALESCE(cp.revenue, 0) > 0 THEN 100.0
WHEN pp.revenue = 0 OR cp.revenue IS NULL THEN 0.0
WHEN ta.trend_slope IS NOT NULL THEN
GREATEST(
-100.0,
LEAST(
(ta.trend_slope / NULLIF(ta.avg_daily_revenue, 0)) * 365 * 100,
999.99
)
)
ELSE
GREATEST(
-100.0,
LEAST(
((COALESCE(cp.revenue, 0) - pp.revenue) /
NULLIF(ABS(pp.revenue), 0)) * 100.0,
999.99
)
)
END as growth_rate,
mc.avg_margin
FROM current_period cp
FULL OUTER JOIN previous_period pp ON cp.cat_id = pp.cat_id
LEFT JOIN trend_analysis ta ON COALESCE(cp.cat_id, pp.cat_id) = ta.cat_id
LEFT JOIN margin_calc mc ON COALESCE(cp.cat_id, pp.cat_id) = mc.cat_id
)
UPDATE category_metrics cm
SET
growth_rate = CASE
WHEN pp.revenue = 0 AND COALESCE(cp.revenue, 0) > 0 THEN 100.0
WHEN pp.revenue = 0 OR cp.revenue IS NULL THEN 0.0
WHEN ta.trend_slope IS NOT NULL THEN
GREATEST(
-100.0,
LEAST(
(ta.trend_slope / NULLIF(ta.avg_daily_revenue, 0)) * 365 * 100,
999.99
)
)
ELSE
GREATEST(
-100.0,
LEAST(
((COALESCE(cp.revenue, 0) - pp.revenue) /
NULLIF(ABS(pp.revenue), 0)) * 100.0,
999.99
)
)
END,
avg_margin = COALESCE(mc.avg_margin, cm.avg_margin),
last_calculated_at = NOW()
FROM current_period cp
FULL OUTER JOIN previous_period pp ON cp.cat_id = pp.cat_id
LEFT JOIN trend_analysis ta ON COALESCE(cp.cat_id, pp.cat_id) = ta.cat_id
LEFT JOIN margin_calc mc ON COALESCE(cp.cat_id, pp.cat_id) = mc.cat_id
WHERE cm.category_id = COALESCE(cp.cat_id, pp.cat_id)
`);
processedCount = Math.floor(totalProducts * 0.97);
outputProgress({
status: 'running',
operation: 'Growth rates calculated, updating time-based metrics',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount, totalProducts),
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
if (isCancelled) return {
processedProducts: processedCount,
processedOrders,
processedPurchaseOrders: 0,
success
};
// Calculate time-based metrics
await connection.query(`
INSERT INTO category_time_metrics (
category_id,
year,
month,
product_count,
active_products,
total_value,
total_revenue,
avg_margin,
turnover_rate
)
SELECT
pc.cat_id,
EXTRACT(YEAR FROM o.date::timestamp with time zone) as year,
EXTRACT(MONTH FROM o.date::timestamp with time zone) as month,
COUNT(DISTINCT p.pid) as product_count,
COUNT(DISTINCT CASE WHEN p.visible = true THEN p.pid END) as active_products,
SUM(p.stock_quantity * p.cost_price) as total_value,
SUM(o.quantity * o.price) as total_revenue,
CASE
WHEN SUM(o.quantity * o.price) > 0 THEN
LEAST(
GREATEST(
SUM(o.quantity * (o.price - GREATEST(p.cost_price, 0))) * 100.0 /
SUM(o.quantity * o.price),
-100
),
100
)
ELSE 0
END as avg_margin,
COALESCE(
LEAST(
SUM(o.quantity) / NULLIF(AVG(GREATEST(p.stock_quantity, 0)), 0),
999.99
),
0
) as turnover_rate
FROM product_categories pc
JOIN products p ON pc.pid = p.pid
JOIN orders o ON p.pid = o.pid
WHERE o.canceled = false
AND o.date >= CURRENT_DATE - INTERVAL '12 months'
GROUP BY pc.cat_id, EXTRACT(YEAR FROM o.date::timestamp with time zone), EXTRACT(MONTH FROM o.date::timestamp with time zone)
ON CONFLICT (category_id, year, month) DO UPDATE
SET
product_count = EXCLUDED.product_count,
active_products = EXCLUDED.active_products,
total_value = EXCLUDED.total_value,
total_revenue = EXCLUDED.total_revenue,
avg_margin = EXCLUDED.avg_margin,
turnover_rate = EXCLUDED.turnover_rate
`);
processedCount = Math.floor(totalProducts * 0.99);
outputProgress({
status: 'running',
operation: 'Time-based metrics calculated, updating category-sales metrics',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount, totalProducts),
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
if (isCancelled) return {
processedProducts: processedCount,
processedOrders,
processedPurchaseOrders: 0,
success
};
// Calculate category-sales metrics
await connection.query(`
INSERT INTO category_sales_metrics (
category_id,
brand,
period_start,
period_end,
avg_daily_sales,
total_sold,
num_products,
avg_price,
last_calculated_at
)
WITH date_ranges AS (
SELECT
CURRENT_DATE - INTERVAL '30 days' as period_start,
CURRENT_DATE as period_end
UNION ALL
SELECT
CURRENT_DATE - INTERVAL '90 days',
CURRENT_DATE - INTERVAL '31 days'
UNION ALL
SELECT
CURRENT_DATE - INTERVAL '180 days',
CURRENT_DATE - INTERVAL '91 days'
UNION ALL
SELECT
CURRENT_DATE - INTERVAL '365 days',
CURRENT_DATE - INTERVAL '181 days'
),
sales_data AS (
SELECT
pc.cat_id,
COALESCE(p.brand, 'Unknown') as brand,
dr.period_start,
dr.period_end,
COUNT(DISTINCT p.pid) as num_products,
SUM(o.quantity) as total_sold,
SUM(o.quantity * o.price) as total_revenue,
COUNT(DISTINCT DATE(o.date)) as num_days
FROM products p
JOIN product_categories pc ON p.pid = pc.pid
JOIN orders o ON p.pid = o.pid
CROSS JOIN date_ranges dr
WHERE o.canceled = false
AND o.date BETWEEN dr.period_start AND dr.period_end
GROUP BY pc.cat_id, p.brand, dr.period_start, dr.period_end
)
SELECT
cat_id as category_id,
brand,
period_start,
period_end,
CASE
WHEN num_days > 0
THEN total_sold / num_days
ELSE 0
END as avg_daily_sales,
total_sold,
num_products,
CASE
WHEN total_sold > 0
THEN total_revenue / total_sold
ELSE 0
END as avg_price,
NOW() as last_calculated_at
FROM sales_data
ON CONFLICT (category_id, brand, period_start, period_end) DO UPDATE
SET
avg_daily_sales = EXCLUDED.avg_daily_sales,
total_sold = EXCLUDED.total_sold,
num_products = EXCLUDED.num_products,
avg_price = EXCLUDED.avg_price,
last_calculated_at = EXCLUDED.last_calculated_at
`);
processedCount = Math.floor(totalProducts * 1.0);
outputProgress({
status: 'running',
operation: 'Category-sales metrics calculated',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount, totalProducts),
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
// If we get here, everything completed successfully
success = true;
// Update calculate_status
await connection.query(`
INSERT INTO calculate_status (module_name, last_calculation_timestamp)
VALUES ('category_metrics', NOW())
ON CONFLICT (module_name) DO UPDATE
SET last_calculation_timestamp = NOW()
`);
return {
processedProducts: processedCount,
processedOrders,
processedPurchaseOrders: 0,
success
};
} catch (error) {
success = false;
logError(error, 'Error calculating category metrics');
throw error;
} finally {
if (connection) {
connection.release();
}
}
}
module.exports = calculateCategoryMetrics;

View File

@@ -1,214 +0,0 @@
const { outputProgress, formatElapsedTime, estimateRemaining, calculateRate, logError } = require('./utils/progress');
const { getConnection } = require('./utils/db');
async function calculateFinancialMetrics(startTime, totalProducts, processedCount = 0, isCancelled = false) {
const connection = await getConnection();
let success = false;
let processedOrders = 0;
try {
if (isCancelled) {
outputProgress({
status: 'cancelled',
operation: 'Financial metrics calculation cancelled',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: null,
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
return {
processedProducts: processedCount,
processedOrders: 0,
processedPurchaseOrders: 0,
success
};
}
// Get order count that will be processed
const orderCount = await connection.query(`
SELECT COUNT(*) as count
FROM orders o
WHERE o.canceled = false
AND DATE(o.date) >= CURRENT_DATE - INTERVAL '12 months'
`);
processedOrders = parseInt(orderCount.rows[0].count);
outputProgress({
status: 'running',
operation: 'Starting financial metrics calculation',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount, totalProducts),
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
// First, calculate beginning inventory values (12 months ago)
await connection.query(`
CREATE TEMPORARY TABLE IF NOT EXISTS temp_beginning_inventory AS
WITH beginning_inventory_calc AS (
SELECT
p.pid,
p.stock_quantity as current_quantity,
COALESCE(SUM(o.quantity), 0) as sold_quantity,
COALESCE(SUM(po.received), 0) as received_quantity,
GREATEST(0, (p.stock_quantity + COALESCE(SUM(o.quantity), 0) - COALESCE(SUM(po.received), 0))) as beginning_quantity,
p.cost_price
FROM
products p
LEFT JOIN
orders o ON p.pid = o.pid
AND o.canceled = false
AND o.date >= CURRENT_DATE - INTERVAL '12 months'::interval
LEFT JOIN
purchase_orders po ON p.pid = po.pid
AND po.received_date IS NOT NULL
AND po.received_date >= CURRENT_DATE - INTERVAL '12 months'::interval
GROUP BY
p.pid, p.stock_quantity, p.cost_price
)
SELECT
pid,
beginning_quantity,
beginning_quantity * cost_price as beginning_value,
current_quantity * cost_price as current_value,
((beginning_quantity * cost_price) + (current_quantity * cost_price)) / 2 as average_inventory_value
FROM
beginning_inventory_calc
`);
processedCount = Math.floor(totalProducts * 0.60);
outputProgress({
status: 'running',
operation: 'Beginning inventory values calculated, computing financial metrics',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount, totalProducts),
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
// Calculate financial metrics with optimized query and standard formulas
await connection.query(`
WITH product_financials AS (
SELECT
p.pid,
COALESCE(bi.average_inventory_value, p.cost_price * p.stock_quantity) as avg_inventory_value,
p.cost_price * p.stock_quantity as current_inventory_value,
SUM(o.quantity * (o.price - COALESCE(o.discount, 0))) as total_revenue,
SUM(o.quantity * COALESCE(o.costeach, 0)) as cost_of_goods_sold,
SUM(o.quantity * (o.price - COALESCE(o.discount, 0) - COALESCE(o.costeach, 0))) as gross_profit,
MIN(o.date) as first_sale_date,
MAX(o.date) as last_sale_date,
EXTRACT(DAY FROM (MAX(o.date)::timestamp with time zone - MIN(o.date)::timestamp with time zone)) + 1 as calculation_period_days,
COUNT(DISTINCT DATE(o.date)) as active_days
FROM products p
LEFT JOIN orders o ON p.pid = o.pid
LEFT JOIN temp_beginning_inventory bi ON p.pid = bi.pid
WHERE o.canceled = false
AND DATE(o.date) >= CURRENT_DATE - INTERVAL '12 months'::interval
GROUP BY p.pid, p.cost_price, p.stock_quantity, bi.average_inventory_value
)
UPDATE product_metrics pm
SET
inventory_value = COALESCE(pf.current_inventory_value, 0)::decimal(10,3),
total_revenue = COALESCE(pf.total_revenue, 0)::decimal(10,3),
cost_of_goods_sold = COALESCE(pf.cost_of_goods_sold, 0)::decimal(10,3),
gross_profit = COALESCE(pf.gross_profit, 0)::decimal(10,3),
turnover_rate = CASE
WHEN COALESCE(pf.avg_inventory_value, 0) > 0 THEN
COALESCE(pf.cost_of_goods_sold, 0) / NULLIF(pf.avg_inventory_value, 0)
ELSE 0
END::decimal(12,3),
gmroi = CASE
WHEN COALESCE(pf.avg_inventory_value, 0) > 0 THEN
COALESCE(pf.gross_profit, 0) / NULLIF(pf.avg_inventory_value, 0)
ELSE 0
END::decimal(10,3),
last_calculated_at = CURRENT_TIMESTAMP
FROM product_financials pf
WHERE pm.pid = pf.pid
`);
processedCount = Math.floor(totalProducts * 0.65);
outputProgress({
status: 'running',
operation: 'Base financial metrics calculated, updating time aggregates',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount, totalProducts),
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
if (isCancelled) return {
processedProducts: processedCount,
processedOrders,
processedPurchaseOrders: 0,
success
};
// Clean up temporary tables
await connection.query('DROP TABLE IF EXISTS temp_beginning_inventory');
// If we get here, everything completed successfully
success = true;
// Update calculate_status
await connection.query(`
INSERT INTO calculate_status (module_name, last_calculation_timestamp)
VALUES ('financial_metrics', NOW())
ON CONFLICT (module_name) DO UPDATE
SET last_calculation_timestamp = NOW()
`);
return {
processedProducts: processedCount,
processedOrders,
processedPurchaseOrders: 0,
success
};
} catch (error) {
success = false;
logError(error, 'Error calculating financial metrics');
throw error;
} finally {
if (connection) {
try {
// Make sure temporary tables are always cleaned up
await connection.query('DROP TABLE IF EXISTS temp_beginning_inventory');
} catch (err) {
console.error('Error cleaning up temp tables:', err);
}
connection.release();
}
}
}
module.exports = calculateFinancialMetrics;

View File

@@ -1,736 +0,0 @@
const { outputProgress, formatElapsedTime, estimateRemaining, calculateRate, logError } = require('./utils/progress');
const { getConnection } = require('./utils/db');
// Helper function to handle NaN and undefined values
function sanitizeValue(value) {
if (value === undefined || value === null || Number.isNaN(value)) {
return null;
}
return value;
}
async function calculateProductMetrics(startTime, totalProducts, processedCount = 0, isCancelled = false) {
let connection;
let success = false;
let processedOrders = 0;
const BATCH_SIZE = 5000;
try {
connection = await getConnection();
// Skip flags are inherited from the parent scope
const SKIP_PRODUCT_BASE_METRICS = 0;
const SKIP_PRODUCT_TIME_AGGREGATES = 0;
// Get total product count if not provided
if (!totalProducts) {
const productCount = await connection.query('SELECT COUNT(*) as count FROM products');
totalProducts = parseInt(productCount.rows[0].count);
}
if (isCancelled) {
outputProgress({
status: 'cancelled',
operation: 'Product metrics calculation cancelled',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: null,
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
return {
processedProducts: processedCount,
processedOrders,
processedPurchaseOrders: 0,
success
};
}
// First ensure all products have a metrics record
await connection.query(`
INSERT INTO product_metrics (pid, last_calculated_at)
SELECT pid, NOW()
FROM products
ON CONFLICT (pid) DO NOTHING
`);
// Get threshold settings once
const thresholds = await connection.query(`
SELECT critical_days, reorder_days, overstock_days, low_stock_threshold
FROM stock_thresholds
WHERE category_id IS NULL AND vendor IS NULL
LIMIT 1
`);
// Check if threshold data was returned
if (!thresholds.rows || thresholds.rows.length === 0) {
console.warn('No default thresholds found in the database. Using explicit type casting in the query.');
}
const defaultThresholds = thresholds.rows[0];
// Get financial calculation configuration parameters
const financialConfig = await connection.query(`
SELECT
order_cost,
holding_rate,
service_level_z_score,
min_reorder_qty,
default_reorder_qty,
default_safety_stock
FROM financial_calc_config
WHERE id = 1
LIMIT 1
`);
const finConfig = financialConfig.rows[0] || {
order_cost: 25.00,
holding_rate: 0.25,
service_level_z_score: 1.96,
min_reorder_qty: 1,
default_reorder_qty: 5,
default_safety_stock: 5
};
// Calculate base product metrics
if (!SKIP_PRODUCT_BASE_METRICS) {
outputProgress({
status: 'running',
operation: 'Starting base product metrics calculation',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount, totalProducts),
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
// Get order count that will be processed
const orderCount = await connection.query(`
SELECT COUNT(*) as count
FROM orders o
WHERE o.canceled = false
`);
processedOrders = parseInt(orderCount.rows[0].count);
// Clear temporary tables
await connection.query('DROP TABLE IF EXISTS temp_sales_metrics');
await connection.query('DROP TABLE IF EXISTS temp_purchase_metrics');
// Create temp_sales_metrics
await connection.query(`
CREATE TEMPORARY TABLE temp_sales_metrics (
pid BIGINT NOT NULL,
daily_sales_avg DECIMAL(10,3),
weekly_sales_avg DECIMAL(10,3),
monthly_sales_avg DECIMAL(10,3),
total_revenue DECIMAL(10,3),
avg_margin_percent DECIMAL(10,3),
first_sale_date DATE,
last_sale_date DATE,
stddev_daily_sales DECIMAL(10,3),
PRIMARY KEY (pid)
)
`);
// Create temp_purchase_metrics
await connection.query(`
CREATE TEMPORARY TABLE temp_purchase_metrics (
pid BIGINT NOT NULL,
avg_lead_time_days DECIMAL(10,2),
last_purchase_date DATE,
first_received_date DATE,
last_received_date DATE,
stddev_lead_time_days DECIMAL(10,2),
PRIMARY KEY (pid)
)
`);
// Populate temp_sales_metrics with base stats and sales averages
await connection.query(`
INSERT INTO temp_sales_metrics
SELECT
p.pid,
COALESCE(SUM(o.quantity) / NULLIF(COUNT(DISTINCT DATE(o.date)), 0), 0) as daily_sales_avg,
COALESCE(SUM(o.quantity) / NULLIF(CEIL(COUNT(DISTINCT DATE(o.date)) / 7), 0), 0) as weekly_sales_avg,
COALESCE(SUM(o.quantity) / NULLIF(CEIL(COUNT(DISTINCT DATE(o.date)) / 30), 0), 0) as monthly_sales_avg,
COALESCE(SUM(o.quantity * o.price), 0) as total_revenue,
CASE
WHEN SUM(o.quantity * o.price) > 0
THEN ((SUM(o.quantity * o.price) - SUM(o.quantity * p.cost_price)) / SUM(o.quantity * o.price)) * 100
ELSE 0
END as avg_margin_percent,
MIN(o.date) as first_sale_date,
MAX(o.date) as last_sale_date,
COALESCE(STDDEV_SAMP(daily_qty.quantity), 0) as stddev_daily_sales
FROM products p
LEFT JOIN orders o ON p.pid = o.pid
AND o.canceled = false
AND o.date >= CURRENT_DATE - INTERVAL '90 days'
LEFT JOIN (
SELECT
pid,
DATE(date) as sale_date,
SUM(quantity) as quantity
FROM orders
WHERE canceled = false
AND date >= CURRENT_DATE - INTERVAL '90 days'
GROUP BY pid, DATE(date)
) daily_qty ON p.pid = daily_qty.pid
GROUP BY p.pid
`);
// Populate temp_purchase_metrics with timeout protection
await Promise.race([
connection.query(`
INSERT INTO temp_purchase_metrics
SELECT
p.pid,
AVG(
CASE
WHEN po.received_date IS NOT NULL AND po.date IS NOT NULL
THEN EXTRACT(EPOCH FROM (po.received_date::timestamp with time zone - po.date::timestamp with time zone)) / 86400.0
ELSE NULL
END
) as avg_lead_time_days,
MAX(po.date) as last_purchase_date,
MIN(po.received_date) as first_received_date,
MAX(po.received_date) as last_received_date,
STDDEV_SAMP(
CASE
WHEN po.received_date IS NOT NULL AND po.date IS NOT NULL
THEN EXTRACT(EPOCH FROM (po.received_date::timestamp with time zone - po.date::timestamp with time zone)) / 86400.0
ELSE NULL
END
) as stddev_lead_time_days
FROM products p
LEFT JOIN purchase_orders po ON p.pid = po.pid
AND po.received_date IS NOT NULL
AND po.date IS NOT NULL
AND po.date >= CURRENT_DATE - INTERVAL '365 days'
GROUP BY p.pid
`),
new Promise((_, reject) =>
setTimeout(() => reject(new Error('Timeout: temp_purchase_metrics query took too long')), 60000)
)
]).catch(async (err) => {
logError(err, 'Error populating temp_purchase_metrics, continuing with empty table');
// Create an empty fallback to continue processing
await connection.query(`
INSERT INTO temp_purchase_metrics
SELECT
p.pid,
30.0 as avg_lead_time_days,
NULL as last_purchase_date,
NULL as first_received_date,
NULL as last_received_date,
0.0 as stddev_lead_time_days
FROM products p
LEFT JOIN temp_purchase_metrics tpm ON p.pid = tpm.pid
WHERE tpm.pid IS NULL
`);
});
// Process updates in batches
let lastPid = 0;
let batchCount = 0;
const MAX_BATCHES = 1000; // Safety limit for number of batches to prevent infinite loops
while (batchCount < MAX_BATCHES) {
if (isCancelled) break;
batchCount++;
const batch = await connection.query(
'SELECT pid FROM products WHERE pid > $1 ORDER BY pid LIMIT $2',
[lastPid, BATCH_SIZE]
);
if (batch.rows.length === 0) break;
// Process the entire batch in a single efficient query
const lowStockThreshold = parseInt(defaultThresholds?.low_stock_threshold) || 5;
const criticalDays = parseInt(defaultThresholds?.critical_days) || 7;
const reorderDays = parseInt(defaultThresholds?.reorder_days) || 14;
const overstockDays = parseInt(defaultThresholds?.overstock_days) || 90;
const serviceLevel = parseFloat(finConfig?.service_level_z_score) || 1.96;
const defaultSafetyStock = parseInt(finConfig?.default_safety_stock) || 5;
const defaultReorderQty = parseInt(finConfig?.default_reorder_qty) || 5;
const orderCost = parseFloat(finConfig?.order_cost) || 25.00;
const holdingRate = parseFloat(finConfig?.holding_rate) || 0.25;
const minReorderQty = parseInt(finConfig?.min_reorder_qty) || 1;
await connection.query(`
UPDATE product_metrics pm
SET
inventory_value = p.stock_quantity * NULLIF(p.cost_price, 0),
daily_sales_avg = COALESCE(sm.daily_sales_avg, 0),
weekly_sales_avg = COALESCE(sm.weekly_sales_avg, 0),
monthly_sales_avg = COALESCE(sm.monthly_sales_avg, 0),
total_revenue = COALESCE(sm.total_revenue, 0),
avg_margin_percent = COALESCE(sm.avg_margin_percent, 0),
first_sale_date = sm.first_sale_date,
last_sale_date = sm.last_sale_date,
avg_lead_time_days = COALESCE(lm.avg_lead_time_days, 30.0),
days_of_inventory = CASE
WHEN COALESCE(sm.daily_sales_avg, 0) > 0
THEN FLOOR(p.stock_quantity / NULLIF(sm.daily_sales_avg, 0))
ELSE NULL
END,
weeks_of_inventory = CASE
WHEN COALESCE(sm.weekly_sales_avg, 0) > 0
THEN FLOOR(p.stock_quantity / NULLIF(sm.weekly_sales_avg, 0))
ELSE NULL
END,
stock_status = CASE
WHEN p.stock_quantity <= 0 THEN 'Out of Stock'
WHEN COALESCE(sm.daily_sales_avg, 0) = 0 AND p.stock_quantity <= ${lowStockThreshold} THEN 'Low Stock'
WHEN COALESCE(sm.daily_sales_avg, 0) = 0 THEN 'In Stock'
WHEN p.stock_quantity / NULLIF(sm.daily_sales_avg, 0) <= ${criticalDays} THEN 'Critical'
WHEN p.stock_quantity / NULLIF(sm.daily_sales_avg, 0) <= ${reorderDays} THEN 'Reorder'
WHEN p.stock_quantity / NULLIF(sm.daily_sales_avg, 0) > ${overstockDays} THEN 'Overstocked'
ELSE 'Healthy'
END,
safety_stock = CASE
WHEN COALESCE(sm.daily_sales_avg, 0) > 0 AND COALESCE(lm.avg_lead_time_days, 0) > 0 THEN
CEIL(
${serviceLevel} * SQRT(
GREATEST(0, COALESCE(lm.avg_lead_time_days, 0)) * POWER(COALESCE(sm.stddev_daily_sales, 0), 2) +
POWER(COALESCE(sm.daily_sales_avg, 0), 2) * POWER(COALESCE(lm.stddev_lead_time_days, 0), 2)
)
)
ELSE ${defaultSafetyStock}
END,
reorder_point = CASE
WHEN COALESCE(sm.daily_sales_avg, 0) > 0 THEN
CEIL(sm.daily_sales_avg * GREATEST(0, COALESCE(lm.avg_lead_time_days, 30.0))) +
(CASE
WHEN COALESCE(sm.daily_sales_avg, 0) > 0 AND COALESCE(lm.avg_lead_time_days, 0) > 0 THEN
CEIL(
${serviceLevel} * SQRT(
GREATEST(0, COALESCE(lm.avg_lead_time_days, 0)) * POWER(COALESCE(sm.stddev_daily_sales, 0), 2) +
POWER(COALESCE(sm.daily_sales_avg, 0), 2) * POWER(COALESCE(lm.stddev_lead_time_days, 0), 2)
)
)
ELSE ${defaultSafetyStock}
END)
ELSE ${lowStockThreshold}
END,
reorder_qty = CASE
WHEN COALESCE(sm.daily_sales_avg, 0) > 0 AND NULLIF(p.cost_price, 0) IS NOT NULL AND NULLIF(p.cost_price, 0) > 0 THEN
GREATEST(
CEIL(SQRT(
(2 * (sm.daily_sales_avg * 365) * ${orderCost}) /
NULLIF(p.cost_price * ${holdingRate}, 0)
)),
${minReorderQty}
)
ELSE ${defaultReorderQty}
END,
overstocked_amt = CASE
WHEN p.stock_quantity / NULLIF(sm.daily_sales_avg, 0) > ${overstockDays}
THEN GREATEST(0, p.stock_quantity - CEIL(sm.daily_sales_avg * ${overstockDays}))
ELSE 0
END,
last_calculated_at = NOW()
FROM products p
LEFT JOIN temp_sales_metrics sm ON p.pid = sm.pid
LEFT JOIN temp_purchase_metrics lm ON p.pid = lm.pid
WHERE p.pid = ANY($1::BIGINT[])
AND pm.pid = p.pid
`, [batch.rows.map(row => row.pid)]);
lastPid = batch.rows[batch.rows.length - 1].pid;
processedCount += batch.rows.length;
outputProgress({
status: 'running',
operation: 'Processing base metrics batch',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount, totalProducts),
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
}
// Add safety check if the loop processed MAX_BATCHES
if (batchCount >= MAX_BATCHES) {
logError(new Error(`Reached maximum batch count (${MAX_BATCHES}). Process may have entered an infinite loop.`), 'Batch processing safety limit reached');
}
}
// Calculate forecast accuracy and bias in batches
let forecastPid = 0;
while (true) {
if (isCancelled) break;
const forecastBatch = await connection.query(
'SELECT pid FROM products WHERE pid > $1 ORDER BY pid LIMIT $2',
[forecastPid, BATCH_SIZE]
);
if (forecastBatch.rows.length === 0) break;
const forecastPidArray = forecastBatch.rows.map(row => row.pid);
// Use array_to_string to convert the array to a string of comma-separated values
await connection.query(`
WITH forecast_metrics AS (
SELECT
sf.pid,
AVG(CASE
WHEN o.quantity > 0
THEN ABS(sf.forecast_quantity - o.quantity) / o.quantity * 100
ELSE 100
END) as avg_forecast_error,
AVG(CASE
WHEN o.quantity > 0
THEN (sf.forecast_quantity - o.quantity) / o.quantity * 100
ELSE 0
END) as avg_forecast_bias,
MAX(sf.forecast_date) as last_forecast_date
FROM sales_forecasts sf
JOIN orders o ON sf.pid = o.pid
AND DATE(o.date) = sf.forecast_date
WHERE o.canceled = false
AND sf.forecast_date >= CURRENT_DATE - INTERVAL '90 days'
AND sf.pid = ANY('{${forecastPidArray.join(',')}}'::BIGINT[])
GROUP BY sf.pid
)
UPDATE product_metrics pm
SET
forecast_accuracy = GREATEST(0, 100 - LEAST(fm.avg_forecast_error, 100)),
forecast_bias = GREATEST(-100, LEAST(fm.avg_forecast_bias, 100)),
last_forecast_date = fm.last_forecast_date,
last_calculated_at = NOW()
FROM forecast_metrics fm
WHERE pm.pid = fm.pid
`);
forecastPid = forecastBatch.rows[forecastBatch.rows.length - 1].pid;
}
// Calculate product time aggregates
if (!SKIP_PRODUCT_TIME_AGGREGATES) {
outputProgress({
status: 'running',
operation: 'Starting product time aggregates calculation',
current: processedCount || 0,
total: totalProducts || 0,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount || 0, totalProducts || 0),
rate: calculateRate(startTime, processedCount || 0),
percentage: (((processedCount || 0) / (totalProducts || 1)) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
// Note: The time-aggregates calculation has been moved to time-aggregates.js
// This module will not duplicate that functionality
processedCount = Math.floor(totalProducts * 0.6);
outputProgress({
status: 'running',
operation: 'Product time aggregates calculation delegated to time-aggregates module',
current: processedCount || 0,
total: totalProducts || 0,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount || 0, totalProducts || 0),
rate: calculateRate(startTime, processedCount || 0),
percentage: (((processedCount || 0) / (totalProducts || 1)) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
} else {
processedCount = Math.floor(totalProducts * 0.6);
outputProgress({
status: 'running',
operation: 'Skipping product time aggregates calculation',
current: processedCount || 0,
total: totalProducts || 0,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount || 0, totalProducts || 0),
rate: calculateRate(startTime, processedCount || 0),
percentage: (((processedCount || 0) / (totalProducts || 1)) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
}
// Calculate ABC classification
outputProgress({
status: 'running',
operation: 'Starting ABC classification',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount, totalProducts),
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
if (isCancelled) return {
processedProducts: processedCount,
processedOrders,
processedPurchaseOrders: 0, // This module doesn't process POs
success
};
const abcConfig = await connection.query('SELECT a_threshold, b_threshold FROM abc_classification_config WHERE id = 1');
const abcThresholds = abcConfig.rows[0] || { a_threshold: 20, b_threshold: 50 };
// Extract values and ensure they are valid numbers
const aThreshold = parseFloat(abcThresholds.a_threshold) || 20;
const bThreshold = parseFloat(abcThresholds.b_threshold) || 50;
// First, create and populate the rankings table with an index
await connection.query('DROP TABLE IF EXISTS temp_revenue_ranks');
await connection.query(`
CREATE TEMPORARY TABLE temp_revenue_ranks (
pid BIGINT NOT NULL,
total_revenue DECIMAL(10,3),
rank_num INT,
dense_rank_num INT,
percentile DECIMAL(5,2),
total_count INT,
PRIMARY KEY (pid)
)
`);
await connection.query('CREATE INDEX ON temp_revenue_ranks (rank_num)');
await connection.query('CREATE INDEX ON temp_revenue_ranks (dense_rank_num)');
await connection.query('CREATE INDEX ON temp_revenue_ranks (percentile)');
// Calculate rankings with proper tie handling
await connection.query(`
INSERT INTO temp_revenue_ranks
WITH revenue_data AS (
SELECT
pid,
total_revenue,
COUNT(*) OVER () as total_count,
PERCENT_RANK() OVER (ORDER BY total_revenue DESC) * 100 as percentile,
RANK() OVER (ORDER BY total_revenue DESC) as rank_num,
DENSE_RANK() OVER (ORDER BY total_revenue DESC) as dense_rank_num
FROM product_metrics
WHERE total_revenue > 0
)
SELECT
pid,
total_revenue,
rank_num,
dense_rank_num,
percentile,
total_count
FROM revenue_data
`);
// Get total count for percentage calculation
const rankingCount = await connection.query('SELECT MAX(rank_num) as total_count FROM temp_revenue_ranks');
const totalCount = parseInt(rankingCount.rows[0].total_count) || 1;
// Process updates in batches
let abcProcessedCount = 0;
const batchSize = 5000;
const maxPid = await connection.query('SELECT MAX(pid) as max_pid FROM products');
const maxProductId = parseInt(maxPid.rows[0].max_pid);
while (abcProcessedCount < maxProductId) {
if (isCancelled) return {
processedProducts: processedCount,
processedOrders,
processedPurchaseOrders: 0,
success
};
// Get a batch of PIDs that need updating
const pids = await connection.query(`
SELECT pm.pid
FROM product_metrics pm
LEFT JOIN temp_revenue_ranks tr ON pm.pid = tr.pid
WHERE pm.pid > $1
AND (pm.abc_class IS NULL
OR pm.abc_class !=
CASE
WHEN tr.pid IS NULL THEN 'C'
WHEN tr.percentile <= ${aThreshold} THEN 'A'
WHEN tr.percentile <= ${bThreshold} THEN 'B'
ELSE 'C'
END)
ORDER BY pm.pid
LIMIT $2
`, [abcProcessedCount, batchSize]);
if (pids.rows.length === 0) break;
const pidValues = pids.rows.map(row => row.pid);
await connection.query(`
UPDATE product_metrics pm
SET abc_class =
CASE
WHEN tr.pid IS NULL THEN 'C'
WHEN tr.percentile <= ${aThreshold} THEN 'A'
WHEN tr.percentile <= ${bThreshold} THEN 'B'
ELSE 'C'
END,
last_calculated_at = NOW()
FROM (SELECT pid, percentile FROM temp_revenue_ranks) tr
WHERE pm.pid = tr.pid AND pm.pid = ANY($1::BIGINT[])
OR (pm.pid = ANY($1::BIGINT[]) AND tr.pid IS NULL)
`, [pidValues]);
// Now update turnover rate with proper handling of zero inventory periods
await connection.query(`
UPDATE product_metrics pm
SET
turnover_rate = CASE
WHEN sales.avg_nonzero_stock > 0 AND sales.active_days > 0
THEN LEAST(
(sales.total_sold / sales.avg_nonzero_stock) * (365.0 / sales.active_days),
999.99
)
ELSE 0
END,
last_calculated_at = NOW()
FROM (
SELECT
o.pid,
SUM(o.quantity) as total_sold,
COUNT(DISTINCT DATE(o.date)) as active_days,
AVG(CASE
WHEN p.stock_quantity > 0 THEN p.stock_quantity
ELSE NULL
END) as avg_nonzero_stock
FROM orders o
JOIN products p ON o.pid = p.pid
WHERE o.canceled = false
AND o.date >= CURRENT_DATE - INTERVAL '90 days'
AND o.pid = ANY($1::BIGINT[])
GROUP BY o.pid
) sales
WHERE pm.pid = sales.pid
`, [pidValues]);
abcProcessedCount = pids.rows[pids.rows.length - 1].pid;
// Calculate progress proportionally to total products
processedCount = Math.floor(totalProducts * (0.60 + (abcProcessedCount / maxProductId) * 0.2));
outputProgress({
status: 'running',
operation: 'ABC classification progress',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount, totalProducts),
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
}
// If we get here, everything completed successfully
success = true;
// Update calculate_status
await connection.query(`
INSERT INTO calculate_status (module_name, last_calculation_timestamp)
VALUES ('product_metrics', NOW())
ON CONFLICT (module_name) DO UPDATE
SET last_calculation_timestamp = NOW()
`);
return {
processedProducts: processedCount || 0,
processedOrders: processedOrders || 0,
processedPurchaseOrders: 0, // This module doesn't process POs
success
};
} catch (error) {
success = false;
logError(error, 'Error calculating product metrics');
throw error;
} finally {
// Always clean up temporary tables, even if an error occurred
if (connection) {
try {
await connection.query('DROP TABLE IF EXISTS temp_sales_metrics');
await connection.query('DROP TABLE IF EXISTS temp_purchase_metrics');
} catch (err) {
console.error('Error cleaning up temporary tables:', err);
}
// Make sure to release the connection
connection.release();
}
}
}
function calculateStockStatus(stock, config, daily_sales_avg, weekly_sales_avg, monthly_sales_avg) {
if (stock <= 0) {
return 'Out of Stock';
}
// Use the most appropriate sales average based on data quality
let sales_avg = daily_sales_avg;
if (sales_avg === 0) {
sales_avg = weekly_sales_avg / 7;
}
if (sales_avg === 0) {
sales_avg = monthly_sales_avg / 30;
}
if (sales_avg === 0) {
return stock <= config.low_stock_threshold ? 'Low Stock' : 'In Stock';
}
const days_of_stock = stock / sales_avg;
if (days_of_stock <= config.critical_days) {
return 'Critical';
} else if (days_of_stock <= config.reorder_days) {
return 'Reorder';
} else if (days_of_stock > config.overstock_days) {
return 'Overstocked';
}
return 'Healthy';
}
// Note: calculateReorderQuantities function has been removed as its logic has been incorporated
// in the main SQL query with configurable parameters
module.exports = calculateProductMetrics;

View File

@@ -1,440 +0,0 @@
const { outputProgress, formatElapsedTime, estimateRemaining, calculateRate, logError } = require('./utils/progress');
const { getConnection } = require('./utils/db');
async function calculateSalesForecasts(startTime, totalProducts, processedCount = 0, isCancelled = false) {
const connection = await getConnection();
let success = false;
let processedOrders = 0;
try {
if (isCancelled) {
outputProgress({
status: 'cancelled',
operation: 'Sales forecasts calculation cancelled',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: null,
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
return {
processedProducts: processedCount,
processedOrders: 0,
processedPurchaseOrders: 0,
success
};
}
// Get order count that will be processed
const orderCount = await connection.query(`
SELECT COUNT(*) as count
FROM orders o
WHERE o.canceled = false
AND o.date >= CURRENT_DATE - INTERVAL '90 days'
`);
processedOrders = parseInt(orderCount.rows[0].count);
outputProgress({
status: 'running',
operation: 'Starting sales forecasts calculation',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount, totalProducts),
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
// First, create a temporary table for forecast dates
await connection.query(`
CREATE TEMPORARY TABLE IF NOT EXISTS temp_forecast_dates (
forecast_date DATE,
day_of_week INT,
month INT,
PRIMARY KEY (forecast_date)
)
`);
await connection.query(`
INSERT INTO temp_forecast_dates
SELECT
CURRENT_DATE + (n || ' days')::INTERVAL as forecast_date,
EXTRACT(DOW FROM CURRENT_DATE + (n || ' days')::INTERVAL) + 1 as day_of_week,
EXTRACT(MONTH FROM CURRENT_DATE + (n || ' days')::INTERVAL) as month
FROM (
SELECT a.n + b.n * 10 as n
FROM
(SELECT 0 as n UNION SELECT 1 UNION SELECT 2 UNION SELECT 3 UNION SELECT 4 UNION
SELECT 5 UNION SELECT 6 UNION SELECT 7 UNION SELECT 8 UNION SELECT 9) a,
(SELECT 0 as n UNION SELECT 1 UNION SELECT 2) b
ORDER BY n
LIMIT 31
) numbers
`);
processedCount = Math.floor(totalProducts * 0.92);
outputProgress({
status: 'running',
operation: 'Forecast dates prepared, calculating daily sales stats',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount, totalProducts),
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
if (isCancelled) return {
processedProducts: processedCount,
processedOrders,
processedPurchaseOrders: 0,
success
};
// Create temporary table for daily sales stats
await connection.query(`
CREATE TEMPORARY TABLE temp_daily_sales AS
SELECT
o.pid,
EXTRACT(DOW FROM o.date) + 1 as day_of_week,
SUM(o.quantity) as daily_quantity,
SUM(o.price * o.quantity) as daily_revenue,
COUNT(DISTINCT DATE(o.date)) as day_count
FROM orders o
WHERE o.canceled = false
AND o.date >= CURRENT_DATE - INTERVAL '90 days'
GROUP BY o.pid, EXTRACT(DOW FROM o.date) + 1
`);
processedCount = Math.floor(totalProducts * 0.94);
outputProgress({
status: 'running',
operation: 'Daily sales stats calculated, preparing product stats',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount, totalProducts),
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
if (isCancelled) return {
processedProducts: processedCount,
processedOrders,
processedPurchaseOrders: 0,
success
};
// Create temporary table for product stats
await connection.query(`
CREATE TEMPORARY TABLE temp_product_stats AS
SELECT
pid,
AVG(daily_revenue) as overall_avg_revenue,
SUM(day_count) as total_days
FROM temp_daily_sales
GROUP BY pid
`);
processedCount = Math.floor(totalProducts * 0.96);
outputProgress({
status: 'running',
operation: 'Product stats prepared, calculating product-level forecasts',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount, totalProducts),
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
if (isCancelled) return {
processedProducts: processedCount,
processedOrders,
processedPurchaseOrders: 0,
success
};
// Calculate product-level forecasts
await connection.query(`
INSERT INTO sales_forecasts (
pid,
forecast_date,
forecast_quantity,
confidence_level,
created_at
)
WITH daily_stats AS (
SELECT
ds.pid,
AVG(ds.daily_quantity) as avg_daily_qty,
STDDEV(ds.daily_quantity) as std_daily_qty,
COUNT(DISTINCT ds.day_count) as data_points,
SUM(ds.day_count) as total_days,
AVG(ds.daily_revenue) as avg_daily_revenue,
STDDEV(ds.daily_revenue) as std_daily_revenue,
MIN(ds.daily_quantity) as min_daily_qty,
MAX(ds.daily_quantity) as max_daily_qty,
-- Calculate variance without using LAG
COALESCE(
STDDEV(ds.daily_quantity) / NULLIF(AVG(ds.daily_quantity), 0),
0
) as daily_variance_ratio
FROM temp_daily_sales ds
GROUP BY ds.pid
HAVING AVG(ds.daily_quantity) > 0
)
SELECT
ds.pid,
fd.forecast_date,
GREATEST(0,
ROUND(
ds.avg_daily_qty *
(1 + COALESCE(sf.seasonality_factor, 0))
)
) as forecast_quantity,
CASE
WHEN ds.total_days >= 60 AND ds.daily_variance_ratio < 0.5 THEN 90
WHEN ds.total_days >= 60 THEN 85
WHEN ds.total_days >= 30 AND ds.daily_variance_ratio < 0.5 THEN 80
WHEN ds.total_days >= 30 THEN 75
WHEN ds.total_days >= 14 AND ds.daily_variance_ratio < 0.5 THEN 70
WHEN ds.total_days >= 14 THEN 65
ELSE 60
END as confidence_level,
NOW() as created_at
FROM daily_stats ds
JOIN temp_product_stats ps ON ds.pid = ps.pid
CROSS JOIN temp_forecast_dates fd
LEFT JOIN sales_seasonality sf ON fd.month = sf.month
GROUP BY ds.pid, fd.forecast_date, ps.overall_avg_revenue, sf.seasonality_factor,
ds.avg_daily_qty, ds.std_daily_qty, ds.avg_daily_qty, ds.total_days, ds.daily_variance_ratio
ON CONFLICT (pid, forecast_date) DO UPDATE
SET
forecast_quantity = EXCLUDED.forecast_quantity,
confidence_level = EXCLUDED.confidence_level,
created_at = NOW()
`);
processedCount = Math.floor(totalProducts * 0.98);
outputProgress({
status: 'running',
operation: 'Product forecasts calculated, preparing category stats',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount, totalProducts),
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
if (isCancelled) return {
processedProducts: processedCount,
processedOrders,
processedPurchaseOrders: 0,
success
};
// Create temporary table for category stats
await connection.query(`
CREATE TEMPORARY TABLE temp_category_sales AS
SELECT
pc.cat_id,
EXTRACT(DOW FROM o.date) + 1 as day_of_week,
SUM(o.quantity) as daily_quantity,
SUM(o.price * o.quantity) as daily_revenue,
COUNT(DISTINCT DATE(o.date)) as day_count
FROM orders o
JOIN product_categories pc ON o.pid = pc.pid
WHERE o.canceled = false
AND o.date >= CURRENT_DATE - INTERVAL '90 days'
GROUP BY pc.cat_id, EXTRACT(DOW FROM o.date) + 1
`);
await connection.query(`
CREATE TEMPORARY TABLE temp_category_stats AS
SELECT
cat_id,
AVG(daily_revenue) as overall_avg_revenue,
SUM(day_count) as total_days
FROM temp_category_sales
GROUP BY cat_id
`);
processedCount = Math.floor(totalProducts * 0.99);
outputProgress({
status: 'running',
operation: 'Category stats prepared, calculating category-level forecasts',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount, totalProducts),
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
if (isCancelled) return {
processedProducts: processedCount,
processedOrders,
processedPurchaseOrders: 0,
success
};
// Calculate category-level forecasts
await connection.query(`
INSERT INTO category_forecasts (
category_id,
forecast_date,
forecast_units,
forecast_revenue,
confidence_level,
created_at
)
SELECT
cs.cat_id::bigint as category_id,
fd.forecast_date,
GREATEST(0,
ROUND(AVG(cs.daily_quantity) *
(1 + COALESCE(sf.seasonality_factor, 0)))
) as forecast_units,
GREATEST(0,
COALESCE(
CASE
WHEN SUM(cs.day_count) >= 4 THEN AVG(cs.daily_revenue)
ELSE ct.overall_avg_revenue
END *
(1 + COALESCE(sf.seasonality_factor, 0)),
0
)
) as forecast_revenue,
CASE
WHEN ct.total_days >= 60 THEN 90
WHEN ct.total_days >= 30 THEN 80
WHEN ct.total_days >= 14 THEN 70
ELSE 60
END as confidence_level,
NOW() as created_at
FROM temp_category_sales cs
JOIN temp_category_stats ct ON cs.cat_id = ct.cat_id
CROSS JOIN temp_forecast_dates fd
LEFT JOIN sales_seasonality sf ON fd.month = sf.month
GROUP BY
cs.cat_id,
fd.forecast_date,
ct.overall_avg_revenue,
ct.total_days,
sf.seasonality_factor,
sf.month
HAVING AVG(cs.daily_quantity) > 0
ON CONFLICT (category_id, forecast_date) DO UPDATE
SET
forecast_units = EXCLUDED.forecast_units,
forecast_revenue = EXCLUDED.forecast_revenue,
confidence_level = EXCLUDED.confidence_level,
created_at = NOW()
`);
// Clean up temporary tables
await connection.query(`
DROP TABLE IF EXISTS temp_forecast_dates;
DROP TABLE IF EXISTS temp_daily_sales;
DROP TABLE IF EXISTS temp_product_stats;
DROP TABLE IF EXISTS temp_category_sales;
DROP TABLE IF EXISTS temp_category_stats;
`);
processedCount = Math.floor(totalProducts * 1.0);
outputProgress({
status: 'running',
operation: 'Category forecasts calculated and temporary tables cleaned up',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount, totalProducts),
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
// If we get here, everything completed successfully
success = true;
// Update calculate_status
await connection.query(`
INSERT INTO calculate_status (module_name, last_calculation_timestamp)
VALUES ('sales_forecasts', NOW())
ON CONFLICT (module_name) DO UPDATE
SET last_calculation_timestamp = NOW()
`);
return {
processedProducts: processedCount,
processedOrders,
processedPurchaseOrders: 0,
success
};
} catch (error) {
success = false;
logError(error, 'Error calculating sales forecasts');
throw error;
} finally {
if (connection) {
try {
// Ensure temporary tables are cleaned up
await connection.query(`
DROP TABLE IF EXISTS temp_forecast_dates;
DROP TABLE IF EXISTS temp_daily_sales;
DROP TABLE IF EXISTS temp_product_stats;
DROP TABLE IF EXISTS temp_category_sales;
DROP TABLE IF EXISTS temp_category_stats;
`);
} catch (err) {
console.error('Error cleaning up temporary tables:', err);
}
connection.release();
}
}
}
module.exports = calculateSalesForecasts;

View File

@@ -1,344 +0,0 @@
const { outputProgress, formatElapsedTime, estimateRemaining, calculateRate, logError } = require('./utils/progress');
const { getConnection } = require('./utils/db');
async function calculateTimeAggregates(startTime, totalProducts, processedCount = 0, isCancelled = false) {
const connection = await getConnection();
let success = false;
let processedOrders = 0;
try {
if (isCancelled) {
outputProgress({
status: 'cancelled',
operation: 'Time aggregates calculation cancelled',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: null,
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
return {
processedProducts: processedCount,
processedOrders: 0,
processedPurchaseOrders: 0,
success
};
}
// Get order count that will be processed
const orderCount = await connection.query(`
SELECT COUNT(*) as count
FROM orders o
WHERE o.canceled = false
`);
processedOrders = parseInt(orderCount.rows[0].count);
outputProgress({
status: 'running',
operation: 'Starting time aggregates calculation',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount, totalProducts),
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
// Create a temporary table for end-of-month inventory values
await connection.query(`
CREATE TEMPORARY TABLE IF NOT EXISTS temp_monthly_inventory AS
WITH months AS (
-- Generate all year/month combinations for the last 12 months
SELECT
EXTRACT(YEAR FROM month_date)::INTEGER as year,
EXTRACT(MONTH FROM month_date)::INTEGER as month,
month_date as start_date,
(month_date + INTERVAL '1 month'::interval - INTERVAL '1 day'::interval)::DATE as end_date
FROM (
SELECT generate_series(
DATE_TRUNC('month', CURRENT_DATE - INTERVAL '12 months'::interval)::DATE,
DATE_TRUNC('month', CURRENT_DATE)::DATE,
INTERVAL '1 month'::interval
) as month_date
) dates
),
monthly_inventory_calc AS (
SELECT
p.pid,
m.year,
m.month,
m.end_date,
p.stock_quantity as current_quantity,
-- Calculate sold during period (before end_date)
COALESCE(SUM(
CASE
WHEN o.date <= m.end_date THEN o.quantity
ELSE 0
END
), 0) as sold_after_end_date,
-- Calculate received during period (before end_date)
COALESCE(SUM(
CASE
WHEN po.received_date <= m.end_date THEN po.received
ELSE 0
END
), 0) as received_after_end_date,
p.cost_price
FROM
products p
CROSS JOIN
months m
LEFT JOIN
orders o ON p.pid = o.pid
AND o.canceled = false
AND o.date > m.end_date
AND o.date <= CURRENT_DATE
LEFT JOIN
purchase_orders po ON p.pid = po.pid
AND po.received_date IS NOT NULL
AND po.received_date > m.end_date
AND po.received_date <= CURRENT_DATE
GROUP BY
p.pid, m.year, m.month, m.end_date, p.stock_quantity, p.cost_price
)
SELECT
pid,
year,
month,
-- End of month quantity = current quantity - sold after + received after
GREATEST(0, current_quantity - sold_after_end_date + received_after_end_date) as end_of_month_quantity,
-- End of month inventory value
GREATEST(0, current_quantity - sold_after_end_date + received_after_end_date) * cost_price as end_of_month_value,
cost_price
FROM
monthly_inventory_calc
`);
processedCount = Math.floor(totalProducts * 0.40);
outputProgress({
status: 'running',
operation: 'Monthly inventory values calculated, processing time aggregates',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount, totalProducts),
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
// Initial insert of time-based aggregates
await connection.query(`
INSERT INTO product_time_aggregates (
pid,
year,
month,
total_quantity_sold,
total_revenue,
total_cost,
order_count,
stock_received,
stock_ordered,
avg_price,
profit_margin,
inventory_value,
gmroi
)
WITH monthly_sales AS (
SELECT
o.pid,
EXTRACT(YEAR FROM o.date::timestamp with time zone)::INTEGER as year,
EXTRACT(MONTH FROM o.date::timestamp with time zone)::INTEGER as month,
SUM(o.quantity) as total_quantity_sold,
SUM((o.price - COALESCE(o.discount, 0)) * o.quantity) as total_revenue,
SUM(COALESCE(o.costeach, 0) * o.quantity) as total_cost,
COUNT(DISTINCT o.order_number) as order_count,
AVG(o.price - COALESCE(o.discount, 0)) as avg_price,
CASE
WHEN SUM((o.price - COALESCE(o.discount, 0)) * o.quantity) > 0
THEN ((SUM((o.price - COALESCE(o.discount, 0)) * o.quantity) - SUM(COALESCE(o.costeach, 0) * o.quantity))
/ SUM((o.price - COALESCE(o.discount, 0)) * o.quantity)) * 100
ELSE 0
END as profit_margin,
COUNT(DISTINCT DATE(o.date)) as active_days
FROM orders o
JOIN products p ON o.pid = p.pid
WHERE o.canceled = false
GROUP BY o.pid, EXTRACT(YEAR FROM o.date::timestamp with time zone), EXTRACT(MONTH FROM o.date::timestamp with time zone)
),
monthly_stock AS (
SELECT
pid,
EXTRACT(YEAR FROM date::timestamp with time zone)::INTEGER as year,
EXTRACT(MONTH FROM date::timestamp with time zone)::INTEGER as month,
SUM(received) as stock_received,
SUM(ordered) as stock_ordered
FROM purchase_orders
GROUP BY pid, EXTRACT(YEAR FROM date::timestamp with time zone), EXTRACT(MONTH FROM date::timestamp with time zone)
)
SELECT
COALESCE(s.pid, ms.pid, mi.pid) as pid,
COALESCE(s.year, ms.year, mi.year) as year,
COALESCE(s.month, ms.month, mi.month) as month,
COALESCE(s.total_quantity_sold, 0)::INTEGER as total_quantity_sold,
COALESCE(s.total_revenue, 0)::DECIMAL(10,3) as total_revenue,
COALESCE(s.total_cost, 0)::DECIMAL(10,3) as total_cost,
COALESCE(s.order_count, 0)::INTEGER as order_count,
COALESCE(ms.stock_received, 0)::INTEGER as stock_received,
COALESCE(ms.stock_ordered, 0)::INTEGER as stock_ordered,
COALESCE(s.avg_price, 0)::DECIMAL(10,3) as avg_price,
COALESCE(s.profit_margin, 0)::DECIMAL(10,3) as profit_margin,
COALESCE(mi.end_of_month_value, 0)::DECIMAL(10,3) as inventory_value,
CASE
WHEN COALESCE(mi.end_of_month_value, 0) > 0
THEN (COALESCE(s.total_revenue, 0) - COALESCE(s.total_cost, 0))
/ NULLIF(COALESCE(mi.end_of_month_value, 0), 0)
ELSE 0
END::DECIMAL(10,3) as gmroi
FROM (
SELECT * FROM monthly_sales s
UNION ALL
SELECT
pid,
year,
month,
0 as total_quantity_sold,
0 as total_revenue,
0 as total_cost,
0 as order_count,
NULL as avg_price,
0 as profit_margin,
0 as active_days
FROM monthly_stock ms
WHERE NOT EXISTS (
SELECT 1 FROM monthly_sales s2
WHERE s2.pid = ms.pid
AND s2.year = ms.year
AND s2.month = ms.month
)
UNION ALL
SELECT
pid,
year,
month,
0 as total_quantity_sold,
0 as total_revenue,
0 as total_cost,
0 as order_count,
NULL as avg_price,
0 as profit_margin,
0 as active_days
FROM temp_monthly_inventory mi
WHERE NOT EXISTS (
SELECT 1 FROM monthly_sales s3
WHERE s3.pid = mi.pid
AND s3.year = mi.year
AND s3.month = mi.month
)
AND NOT EXISTS (
SELECT 1 FROM monthly_stock ms3
WHERE ms3.pid = mi.pid
AND ms3.year = mi.year
AND ms3.month = mi.month
)
) s
LEFT JOIN monthly_stock ms
ON s.pid = ms.pid
AND s.year = ms.year
AND s.month = ms.month
LEFT JOIN temp_monthly_inventory mi
ON s.pid = mi.pid
AND s.year = mi.year
AND s.month = mi.month
ON CONFLICT (pid, year, month) DO UPDATE
SET
total_quantity_sold = EXCLUDED.total_quantity_sold,
total_revenue = EXCLUDED.total_revenue,
total_cost = EXCLUDED.total_cost,
order_count = EXCLUDED.order_count,
stock_received = EXCLUDED.stock_received,
stock_ordered = EXCLUDED.stock_ordered,
avg_price = EXCLUDED.avg_price,
profit_margin = EXCLUDED.profit_margin,
inventory_value = EXCLUDED.inventory_value,
gmroi = EXCLUDED.gmroi
`);
processedCount = Math.floor(totalProducts * 0.60);
outputProgress({
status: 'running',
operation: 'Base time aggregates calculated',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount, totalProducts),
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
if (isCancelled) return {
processedProducts: processedCount,
processedOrders,
processedPurchaseOrders: 0,
success
};
// Clean up temporary tables
await connection.query('DROP TABLE IF EXISTS temp_monthly_inventory');
// If we get here, everything completed successfully
success = true;
// Update calculate_status
await connection.query(`
INSERT INTO calculate_status (module_name, last_calculation_timestamp)
VALUES ('time_aggregates', NOW())
ON CONFLICT (module_name) DO UPDATE
SET last_calculation_timestamp = NOW()
`);
return {
processedProducts: processedCount,
processedOrders,
processedPurchaseOrders: 0,
success
};
} catch (error) {
success = false;
logError(error, 'Error calculating time aggregates');
throw error;
} finally {
if (connection) {
try {
// Ensure temporary tables are cleaned up
await connection.query('DROP TABLE IF EXISTS temp_monthly_inventory');
} catch (err) {
console.error('Error cleaning up temporary tables:', err);
}
connection.release();
}
}
}
module.exports = calculateTimeAggregates;

View File

@@ -1,39 +0,0 @@
const { Pool } = require('pg');
const path = require('path');
require('dotenv').config({ path: path.resolve(__dirname, '../../..', '.env') });
// Database configuration
const dbConfig = {
host: process.env.DB_HOST,
user: process.env.DB_USER,
password: process.env.DB_PASSWORD,
database: process.env.DB_NAME,
port: process.env.DB_PORT || 5432,
ssl: process.env.DB_SSL === 'true',
// Add performance optimizations
max: 10, // connection pool max size
idleTimeoutMillis: 30000,
connectionTimeoutMillis: 60000
};
// Create a single pool instance to be reused
const pool = new Pool(dbConfig);
// Add event handlers for pool
pool.on('error', (err, client) => {
console.error('Unexpected error on idle client', err);
});
async function getConnection() {
return await pool.connect();
}
async function closePool() {
await pool.end();
}
module.exports = {
dbConfig,
getConnection,
closePool
};

View File

@@ -1,158 +0,0 @@
const fs = require('fs');
const path = require('path');
// Helper function to format elapsed time
function formatElapsedTime(elapsed) {
// If elapsed is a timestamp, convert to elapsed milliseconds
if (elapsed instanceof Date || elapsed > 1000000000000) {
elapsed = Date.now() - elapsed;
} else {
// If elapsed is in seconds, convert to milliseconds
elapsed = elapsed * 1000;
}
const seconds = Math.floor(elapsed / 1000);
const minutes = Math.floor(seconds / 60);
const hours = Math.floor(minutes / 60);
if (hours > 0) {
return `${hours}h ${minutes % 60}m`;
} else if (minutes > 0) {
return `${minutes}m ${seconds % 60}s`;
} else {
return `${seconds}s`;
}
}
// Helper function to estimate remaining time
function estimateRemaining(startTime, current, total) {
if (current === 0) return null;
const elapsed = Date.now() - startTime;
const rate = current / elapsed;
const remaining = (total - current) / rate;
const minutes = Math.floor(remaining / 60000);
const seconds = Math.floor((remaining % 60000) / 1000);
if (minutes > 0) {
return `${minutes}m ${seconds}s`;
} else {
return `${seconds}s`;
}
}
// Helper function to calculate rate
function calculateRate(startTime, current) {
const elapsed = (Date.now() - startTime) / 1000; // Convert to seconds
return elapsed > 0 ? Math.round(current / elapsed) : 0;
}
// Set up logging
const LOG_DIR = path.join(__dirname, '../../../logs');
const ERROR_LOG = path.join(LOG_DIR, 'import-errors.log');
const IMPORT_LOG = path.join(LOG_DIR, 'import.log');
const STATUS_FILE = path.join(LOG_DIR, 'metrics-status.json');
// Ensure log directory exists
if (!fs.existsSync(LOG_DIR)) {
fs.mkdirSync(LOG_DIR, { recursive: true });
}
// Helper function to log errors
function logError(error, context = '') {
const timestamp = new Date().toISOString();
const errorMessage = `[${timestamp}] ${context}\nError: ${error.message}\nStack: ${error.stack}\n\n`;
// Log to error file
fs.appendFileSync(ERROR_LOG, errorMessage);
// Also log to console
console.error(`\n${context}\nError: ${error.message}`);
}
// Helper function to log import progress
function logImport(message) {
const timestamp = new Date().toISOString();
const logMessage = `[${timestamp}] ${message}\n`;
fs.appendFileSync(IMPORT_LOG, logMessage);
}
// Helper function to output progress
function outputProgress(data) {
// Save progress to file for resumption
saveProgress(data);
// Format as SSE event
const event = {
progress: data
};
// Always send to stdout for frontend
process.stdout.write(JSON.stringify(event) + '\n');
// Log significant events to disk
const isSignificant =
// Operation starts
(data.operation && !data.current) ||
// Operation completions and errors
data.status === 'complete' ||
data.status === 'error' ||
// Major phase changes
data.operation?.includes('Starting ABC classification') ||
data.operation?.includes('Starting time-based aggregates') ||
data.operation?.includes('Starting vendor metrics');
if (isSignificant) {
logImport(`${data.operation || 'Operation'}${data.message ? ': ' + data.message : ''}${data.error ? ' Error: ' + data.error : ''}${data.status ? ' Status: ' + data.status : ''}`);
}
}
function saveProgress(progress) {
try {
fs.writeFileSync(STATUS_FILE, JSON.stringify({
...progress,
timestamp: Date.now()
}));
} catch (err) {
console.error('Failed to save progress:', err);
}
}
function clearProgress() {
try {
if (fs.existsSync(STATUS_FILE)) {
fs.unlinkSync(STATUS_FILE);
}
} catch (err) {
console.error('Failed to clear progress:', err);
}
}
function getProgress() {
try {
if (fs.existsSync(STATUS_FILE)) {
const progress = JSON.parse(fs.readFileSync(STATUS_FILE, 'utf8'));
// Check if the progress is still valid (less than 1 hour old)
if (progress.timestamp && Date.now() - progress.timestamp < 3600000) {
return progress;
} else {
// Clear old progress
clearProgress();
}
}
} catch (err) {
console.error('Failed to read progress:', err);
clearProgress();
}
return null;
}
module.exports = {
formatElapsedTime,
estimateRemaining,
calculateRate,
logError,
logImport,
outputProgress,
saveProgress,
clearProgress,
getProgress
};

View File

@@ -1,378 +0,0 @@
const { outputProgress, formatElapsedTime, estimateRemaining, calculateRate, logError } = require('./utils/progress');
const { getConnection } = require('./utils/db');
async function calculateVendorMetrics(startTime, totalProducts, processedCount = 0, isCancelled = false) {
const connection = await getConnection();
let success = false;
let processedOrders = 0;
let processedPurchaseOrders = 0;
try {
if (isCancelled) {
outputProgress({
status: 'cancelled',
operation: 'Vendor metrics calculation cancelled',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: null,
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
return {
processedProducts: processedCount,
processedOrders,
processedPurchaseOrders,
success
};
}
// Get counts of records that will be processed
const [orderCountResult, poCountResult] = await Promise.all([
connection.query(`
SELECT COUNT(*) as count
FROM orders o
WHERE o.canceled = false
`),
connection.query(`
SELECT COUNT(*) as count
FROM purchase_orders po
WHERE po.status != 0
`)
]);
processedOrders = parseInt(orderCountResult.rows[0].count);
processedPurchaseOrders = parseInt(poCountResult.rows[0].count);
outputProgress({
status: 'running',
operation: 'Starting vendor metrics calculation',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount, totalProducts),
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
// First ensure all vendors exist in vendor_details
await connection.query(`
INSERT INTO vendor_details (vendor, status, created_at, updated_at)
SELECT DISTINCT
vendor,
'active' as status,
NOW() as created_at,
NOW() as updated_at
FROM products
WHERE vendor IS NOT NULL
ON CONFLICT (vendor) DO NOTHING
`);
processedCount = Math.floor(totalProducts * 0.8);
outputProgress({
status: 'running',
operation: 'Vendor details updated, calculating metrics',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount, totalProducts),
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
if (isCancelled) return {
processedProducts: processedCount,
processedOrders,
processedPurchaseOrders,
success
};
// Now calculate vendor metrics
await connection.query(`
INSERT INTO vendor_metrics (
vendor,
total_revenue,
total_orders,
total_late_orders,
avg_lead_time_days,
on_time_delivery_rate,
order_fill_rate,
avg_order_value,
active_products,
total_products,
total_purchase_value,
avg_margin_percent,
status,
last_calculated_at
)
WITH vendor_sales AS (
SELECT
p.vendor,
SUM(o.quantity * o.price) as total_revenue,
COUNT(DISTINCT o.id) as total_orders,
COUNT(DISTINCT p.pid) as active_products,
SUM(o.quantity * (o.price - p.cost_price)) as total_margin
FROM products p
JOIN orders o ON p.pid = o.pid
WHERE o.canceled = false
AND o.date >= CURRENT_DATE - INTERVAL '12 months'
GROUP BY p.vendor
),
vendor_po AS (
SELECT
p.vendor,
COUNT(DISTINCT CASE WHEN po.receiving_status = 40 THEN po.id END) as received_orders,
COUNT(DISTINCT po.id) as total_orders,
AVG(CASE
WHEN po.receiving_status = 40
AND po.received_date IS NOT NULL
AND po.date IS NOT NULL
THEN EXTRACT(EPOCH FROM (po.received_date::timestamp with time zone - po.date::timestamp with time zone)) / 86400.0
ELSE NULL
END) as avg_lead_time_days,
SUM(po.ordered * po.po_cost_price) as total_purchase_value
FROM products p
JOIN purchase_orders po ON p.pid = po.pid
WHERE po.date >= CURRENT_DATE - INTERVAL '12 months'
GROUP BY p.vendor
),
vendor_products AS (
SELECT
vendor,
COUNT(DISTINCT pid) as total_products
FROM products
GROUP BY vendor
)
SELECT
vs.vendor,
COALESCE(vs.total_revenue, 0) as total_revenue,
COALESCE(vp.total_orders, 0) as total_orders,
COALESCE(vp.total_orders - vp.received_orders, 0) as total_late_orders,
COALESCE(vp.avg_lead_time_days, 0) as avg_lead_time_days,
CASE
WHEN vp.total_orders > 0
THEN (vp.received_orders / vp.total_orders) * 100
ELSE 0
END as on_time_delivery_rate,
CASE
WHEN vp.total_orders > 0
THEN (vp.received_orders / vp.total_orders) * 100
ELSE 0
END as order_fill_rate,
CASE
WHEN vs.total_orders > 0
THEN vs.total_revenue / vs.total_orders
ELSE 0
END as avg_order_value,
COALESCE(vs.active_products, 0) as active_products,
COALESCE(vpr.total_products, 0) as total_products,
COALESCE(vp.total_purchase_value, 0) as total_purchase_value,
CASE
WHEN vs.total_revenue > 0
THEN (vs.total_margin / vs.total_revenue) * 100
ELSE 0
END as avg_margin_percent,
'active' as status,
NOW() as last_calculated_at
FROM vendor_sales vs
LEFT JOIN vendor_po vp ON vs.vendor = vp.vendor
LEFT JOIN vendor_products vpr ON vs.vendor = vpr.vendor
WHERE vs.vendor IS NOT NULL
ON CONFLICT (vendor) DO UPDATE
SET
total_revenue = EXCLUDED.total_revenue,
total_orders = EXCLUDED.total_orders,
total_late_orders = EXCLUDED.total_late_orders,
avg_lead_time_days = EXCLUDED.avg_lead_time_days,
on_time_delivery_rate = EXCLUDED.on_time_delivery_rate,
order_fill_rate = EXCLUDED.order_fill_rate,
avg_order_value = EXCLUDED.avg_order_value,
active_products = EXCLUDED.active_products,
total_products = EXCLUDED.total_products,
total_purchase_value = EXCLUDED.total_purchase_value,
avg_margin_percent = EXCLUDED.avg_margin_percent,
status = EXCLUDED.status,
last_calculated_at = EXCLUDED.last_calculated_at
`);
processedCount = Math.floor(totalProducts * 0.9);
outputProgress({
status: 'running',
operation: 'Vendor metrics calculated, updating time-based metrics',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount, totalProducts),
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
if (isCancelled) return {
processedProducts: processedCount,
processedOrders,
processedPurchaseOrders,
success
};
// Calculate time-based metrics
await connection.query(`
INSERT INTO vendor_time_metrics (
vendor,
year,
month,
total_orders,
late_orders,
avg_lead_time_days,
total_purchase_value,
total_revenue,
avg_margin_percent
)
WITH monthly_orders AS (
SELECT
p.vendor,
EXTRACT(YEAR FROM o.date::timestamp with time zone) as year,
EXTRACT(MONTH FROM o.date::timestamp with time zone) as month,
COUNT(DISTINCT o.id) as total_orders,
SUM(o.quantity * o.price) as total_revenue,
SUM(o.quantity * (o.price - p.cost_price)) as total_margin
FROM products p
JOIN orders o ON p.pid = o.pid
WHERE o.canceled = false
AND o.date >= CURRENT_DATE - INTERVAL '12 months'
AND p.vendor IS NOT NULL
GROUP BY p.vendor, EXTRACT(YEAR FROM o.date::timestamp with time zone), EXTRACT(MONTH FROM o.date::timestamp with time zone)
),
monthly_po AS (
SELECT
p.vendor,
EXTRACT(YEAR FROM po.date::timestamp with time zone) as year,
EXTRACT(MONTH FROM po.date::timestamp with time zone) as month,
COUNT(DISTINCT po.id) as total_po,
COUNT(DISTINCT CASE
WHEN po.receiving_status = 40 AND po.received_date > po.expected_date
THEN po.id
END) as late_orders,
AVG(CASE
WHEN po.receiving_status = 40
AND po.received_date IS NOT NULL
AND po.date IS NOT NULL
THEN EXTRACT(EPOCH FROM (po.received_date::timestamp with time zone - po.date::timestamp with time zone)) / 86400.0
ELSE NULL
END) as avg_lead_time_days,
SUM(po.ordered * po.po_cost_price) as total_purchase_value
FROM products p
JOIN purchase_orders po ON p.pid = po.pid
WHERE po.date >= CURRENT_DATE - INTERVAL '12 months'
AND p.vendor IS NOT NULL
GROUP BY p.vendor, EXTRACT(YEAR FROM po.date::timestamp with time zone), EXTRACT(MONTH FROM po.date::timestamp with time zone)
)
SELECT
mo.vendor,
mo.year,
mo.month,
COALESCE(mp.total_po, 0) as total_orders,
COALESCE(mp.late_orders, 0) as late_orders,
COALESCE(mp.avg_lead_time_days, 0) as avg_lead_time_days,
COALESCE(mp.total_purchase_value, 0) as total_purchase_value,
mo.total_revenue,
CASE
WHEN mo.total_revenue > 0
THEN (mo.total_margin / mo.total_revenue) * 100
ELSE 0
END as avg_margin_percent
FROM monthly_orders mo
LEFT JOIN monthly_po mp ON mo.vendor = mp.vendor
AND mo.year = mp.year
AND mo.month = mp.month
UNION
SELECT
mp.vendor,
mp.year,
mp.month,
mp.total_po as total_orders,
mp.late_orders,
mp.avg_lead_time_days,
mp.total_purchase_value,
0 as total_revenue,
0 as avg_margin_percent
FROM monthly_po mp
LEFT JOIN monthly_orders mo ON mp.vendor = mo.vendor
AND mp.year = mo.year
AND mp.month = mo.month
WHERE mo.vendor IS NULL
ON CONFLICT (vendor, year, month) DO UPDATE
SET
total_orders = EXCLUDED.total_orders,
late_orders = EXCLUDED.late_orders,
avg_lead_time_days = EXCLUDED.avg_lead_time_days,
total_purchase_value = EXCLUDED.total_purchase_value,
total_revenue = EXCLUDED.total_revenue,
avg_margin_percent = EXCLUDED.avg_margin_percent
`);
processedCount = Math.floor(totalProducts * 0.95);
outputProgress({
status: 'running',
operation: 'Time-based vendor metrics calculated',
current: processedCount,
total: totalProducts,
elapsed: formatElapsedTime(startTime),
remaining: estimateRemaining(startTime, processedCount, totalProducts),
rate: calculateRate(startTime, processedCount),
percentage: ((processedCount / totalProducts) * 100).toFixed(1),
timing: {
start_time: new Date(startTime).toISOString(),
end_time: new Date().toISOString(),
elapsed_seconds: Math.round((Date.now() - startTime) / 1000)
}
});
// If we get here, everything completed successfully
success = true;
// Update calculate_status
await connection.query(`
INSERT INTO calculate_status (module_name, last_calculation_timestamp)
VALUES ('vendor_metrics', NOW())
ON CONFLICT (module_name) DO UPDATE
SET last_calculation_timestamp = NOW()
`);
return {
processedProducts: processedCount,
processedOrders,
processedPurchaseOrders,
success
};
} catch (error) {
success = false;
logError(error, 'Error calculating vendor metrics');
throw error;
} finally {
if (connection) {
connection.release();
}
}
}
module.exports = calculateVendorMetrics;

View File

@@ -0,0 +1,428 @@
#!/bin/bash
# Simple script to import CSV to PostgreSQL using psql
# Usage: ./psql-csv-import.sh <csv-file> <table-name> [start-batch]
# Exit on error
set -e
# Get arguments
CSV_FILE=$1
TABLE_NAME=$2
BATCH_SIZE=500000 # Process 500,000 rows at a time
START_BATCH=${3:-1} # Optional third parameter to start from a specific batch
if [ -z "$CSV_FILE" ] || [ -z "$TABLE_NAME" ]; then
echo "Usage: ./psql-csv-import.sh <csv-file> <table-name> [start-batch]"
exit 1
fi
# Check if file exists (only needed for batch 1)
if [ "$START_BATCH" -eq 1 ] && [ ! -f "$CSV_FILE" ]; then
echo "Error: CSV file '$CSV_FILE' not found"
exit 1
fi
# Load environment variables
if [ -f "../.env" ]; then
source "../.env"
else
echo "Warning: .env file not found, using default connection parameters"
fi
# Set default connection parameters if not from .env
DB_HOST=${DB_HOST:-localhost}
DB_PORT=${DB_PORT:-5432}
DB_NAME=${DB_NAME:-inventory_db}
DB_USER=${DB_USER:-postgres}
export PGPASSWORD=${DB_PASSWORD:-} # Export password for psql
# Common psql parameters
PSQL_OPTS="-h $DB_HOST -p $DB_PORT -U $DB_USER -d $DB_NAME"
# Function to clean up database state
cleanup_and_optimize() {
echo "Cleaning up and optimizing database state..."
# Analyze the target table to update statistics
psql $PSQL_OPTS -c "ANALYZE $TABLE_NAME;"
# Perform vacuum to reclaim space and update stats
psql $PSQL_OPTS -c "VACUUM $TABLE_NAME;"
# Reset connection pool
psql $PSQL_OPTS -c "SELECT pg_terminate_backend(pid) FROM pg_stat_activity WHERE datname = current_database() AND pid <> pg_backend_pid();"
# Clean up shared memory
psql $PSQL_OPTS -c "DISCARD ALL;"
echo "Optimization complete."
}
# Show connection info
echo "Importing $CSV_FILE into $TABLE_NAME"
echo "Database: $DB_NAME on $DB_HOST:$DB_PORT with batch size: $BATCH_SIZE starting at batch $START_BATCH"
# Start timer
START_TIME=$(date +%s)
# Create progress tracking file
PROGRESS_FILE="/tmp/import_progress_${TABLE_NAME}.txt"
touch "$PROGRESS_FILE"
echo "Starting import at $(date), batch $START_BATCH" >> "$PROGRESS_FILE"
# If we're resuming, run cleanup first
if [ "$START_BATCH" -gt 1 ]; then
cleanup_and_optimize
fi
# For imported_product_stat_history, use optimized approach with hardcoded column names
if [ "$TABLE_NAME" = "imported_product_stat_history" ]; then
echo "Using optimized import for $TABLE_NAME"
# Only drop constraints/indexes and create staging table for batch 1
if [ "$START_BATCH" -eq 1 ]; then
# Extract CSV header
CSV_HEADER=$(head -n 1 "$CSV_FILE")
echo "CSV header: $CSV_HEADER"
# Step 1: Drop constraints and indexes
echo "Dropping constraints and indexes..."
psql $PSQL_OPTS -c "
DO \$\$
DECLARE
constraint_name TEXT;
BEGIN
-- Drop primary key constraint if exists
SELECT conname INTO constraint_name
FROM pg_constraint
WHERE conrelid = '$TABLE_NAME'::regclass AND contype = 'p';
IF FOUND THEN
EXECUTE 'ALTER TABLE $TABLE_NAME DROP CONSTRAINT IF EXISTS ' || constraint_name;
RAISE NOTICE 'Dropped primary key constraint: %', constraint_name;
END IF;
END \$\$;
"
# Drop all indexes on the table
psql $PSQL_OPTS -c "
DO \$\$
DECLARE
index_name TEXT;
index_record RECORD;
BEGIN
FOR index_record IN
SELECT indexname
FROM pg_indexes
WHERE tablename = '$TABLE_NAME'
LOOP
EXECUTE 'DROP INDEX IF EXISTS ' || index_record.indexname;
RAISE NOTICE 'Dropped index: %', index_record.indexname;
END LOOP;
END \$\$;
"
# Step 2: Set maintenance_work_mem and disable triggers
echo "Setting maintenance_work_mem and disabling triggers..."
psql $PSQL_OPTS -c "
SET maintenance_work_mem = '1GB';
ALTER TABLE $TABLE_NAME DISABLE TRIGGER ALL;
"
# Step 3: Create staging table
echo "Creating staging table..."
psql $PSQL_OPTS -c "
DROP TABLE IF EXISTS staging_import;
CREATE UNLOGGED TABLE staging_import (
pid TEXT,
date TEXT,
score TEXT,
score2 TEXT,
qty_in_baskets TEXT,
qty_sold TEXT,
notifies_set TEXT,
visibility_score TEXT,
health_score TEXT,
sold_view_score TEXT
);
-- Create an index on staging_import to improve OFFSET performance
CREATE INDEX ON staging_import (pid);
"
# Step 4: Import CSV into staging table
echo "Importing CSV into staging table..."
psql $PSQL_OPTS -c "\copy staging_import FROM '$CSV_FILE' WITH CSV HEADER DELIMITER ','"
else
echo "Resuming import from batch $START_BATCH - skipping table creation and CSV import"
# Check if staging table exists
STAGING_EXISTS=$(psql $PSQL_OPTS -t -c "SELECT EXISTS(SELECT 1 FROM pg_tables WHERE tablename='staging_import');" | tr -d '[:space:]')
if [ "$STAGING_EXISTS" != "t" ]; then
echo "Error: Staging table 'staging_import' does not exist. Run without batch parameter first."
exit 1
fi
# Ensure triggers are disabled
psql $PSQL_OPTS -c "ALTER TABLE $TABLE_NAME DISABLE TRIGGER ALL;"
# Optimize PostgreSQL for better performance
psql $PSQL_OPTS -c "
-- Increase work mem for this session
SET work_mem = '256MB';
SET maintenance_work_mem = '1GB';
"
fi
# Step 5: Get total row count
TOTAL_ROWS=$(psql $PSQL_OPTS -t -c "SELECT COUNT(*) FROM staging_import;" | tr -d '[:space:]')
echo "Total rows to import: $TOTAL_ROWS"
# Calculate starting point
PROCESSED=$(( ($START_BATCH - 1) * $BATCH_SIZE ))
if [ $PROCESSED -ge $TOTAL_ROWS ]; then
echo "Error: Start batch $START_BATCH is beyond the available rows ($TOTAL_ROWS)"
exit 1
fi
# Step 6: Process in batches with shell loop
BATCH_NUM=$(( $START_BATCH - 1 ))
# We'll process batches in chunks of 10 before cleaning up
CHUNKS_SINCE_CLEANUP=0
while [ $PROCESSED -lt $TOTAL_ROWS ]; do
BATCH_NUM=$(( $BATCH_NUM + 1 ))
BATCH_START=$(date +%s)
MAX_ROWS=$(( $PROCESSED + $BATCH_SIZE ))
if [ $MAX_ROWS -gt $TOTAL_ROWS ]; then
MAX_ROWS=$TOTAL_ROWS
fi
echo "Processing batch $BATCH_NUM (rows $PROCESSED to $MAX_ROWS)..."
# Optimize query buffer for this batch
psql $PSQL_OPTS -c "SET work_mem = '256MB';"
# Insert batch with type casts
psql $PSQL_OPTS -c "
INSERT INTO $TABLE_NAME (
pid, date, score, score2, qty_in_baskets, qty_sold,
notifies_set, visibility_score, health_score, sold_view_score
)
SELECT
pid::bigint,
date::date,
score::numeric,
score2::numeric,
qty_in_baskets::smallint,
qty_sold::smallint,
notifies_set::smallint,
visibility_score::numeric,
health_score::varchar,
sold_view_score::numeric
FROM staging_import
LIMIT $BATCH_SIZE
OFFSET $PROCESSED;
"
# Update progress
BATCH_END=$(date +%s)
BATCH_ELAPSED=$(( $BATCH_END - $BATCH_START ))
PROGRESS_PCT=$(echo "scale=2; $MAX_ROWS * 100 / $TOTAL_ROWS" | bc)
echo "Batch $BATCH_NUM committed in ${BATCH_ELAPSED}s, $MAX_ROWS of $TOTAL_ROWS rows processed ($PROGRESS_PCT%)" | tee -a "$PROGRESS_FILE"
# Increment counter
PROCESSED=$(( $PROCESSED + $BATCH_SIZE ))
CHUNKS_SINCE_CLEANUP=$(( $CHUNKS_SINCE_CLEANUP + 1 ))
# Check current row count every 10 batches
if [ $(( $BATCH_NUM % 10 )) -eq 0 ]; then
CURRENT_COUNT=$(psql $PSQL_OPTS -t -c "SELECT COUNT(*) FROM $TABLE_NAME;" | tr -d '[:space:]')
echo "Current row count in $TABLE_NAME: $CURRENT_COUNT" | tee -a "$PROGRESS_FILE"
# Every 10 batches, run an intermediate cleanup
if [ $CHUNKS_SINCE_CLEANUP -ge 10 ]; then
echo "Running intermediate cleanup and optimization..."
psql $PSQL_OPTS -c "VACUUM $TABLE_NAME;"
CHUNKS_SINCE_CLEANUP=0
fi
fi
# Optional - write a checkpoint file to know where to restart
echo "$BATCH_NUM" > "/tmp/import_last_batch_${TABLE_NAME}.txt"
done
# Only recreate indexes if we've completed the import
if [ $PROCESSED -ge $TOTAL_ROWS ]; then
# Step 7: Re-enable triggers and recreate primary key
echo "Re-enabling triggers and recreating primary key..."
psql $PSQL_OPTS -c "
ALTER TABLE $TABLE_NAME ENABLE TRIGGER ALL;
ALTER TABLE $TABLE_NAME ADD PRIMARY KEY (pid, date);
"
# Step 8: Clean up and get final count
echo "Cleaning up and getting final count..."
psql $PSQL_OPTS -c "
DROP TABLE staging_import;
VACUUM ANALYZE $TABLE_NAME;
SELECT COUNT(*) AS \"Total rows in $TABLE_NAME\" FROM $TABLE_NAME;
"
else
echo "Import interrupted at batch $BATCH_NUM. To resume, run:"
echo "./psql-csv-import.sh $CSV_FILE $TABLE_NAME $BATCH_NUM"
fi
else
# Generic approach for other tables
if [ "$START_BATCH" -eq 1 ]; then
# Extract CSV header
CSV_HEADER=$(head -n 1 "$CSV_FILE")
echo "CSV header: $CSV_HEADER"
# Extract CSV header and format it for SQL
CSV_COLUMNS=$(echo "$CSV_HEADER" | tr ',' '\n' | sed 's/^/"/;s/$/"/' | tr '\n' ',' | sed 's/,$//')
TEMP_COLUMNS=$(echo "$CSV_HEADER" | tr ',' '\n' | sed 's/$/ TEXT/' | tr '\n' ',' | sed 's/,$//')
echo "Importing columns: $CSV_COLUMNS"
# Step 1: Set maintenance_work_mem and disable triggers
echo "Setting maintenance_work_mem and disabling triggers..."
psql $PSQL_OPTS -c "
SET maintenance_work_mem = '1GB';
ALTER TABLE $TABLE_NAME DISABLE TRIGGER ALL;
"
# Step 2: Create temp table
echo "Creating temporary table..."
psql $PSQL_OPTS -c "
DROP TABLE IF EXISTS temp_import;
CREATE UNLOGGED TABLE temp_import ($TEMP_COLUMNS);
-- Create an index on temp_import to improve OFFSET performance
CREATE INDEX ON temp_import ((1)); -- Index on first column
"
# Step 3: Import CSV into temp table
echo "Importing CSV into temporary table..."
psql $PSQL_OPTS -c "\copy temp_import FROM '$CSV_FILE' WITH CSV HEADER DELIMITER ','"
else
echo "Resuming import from batch $START_BATCH - skipping table creation and CSV import"
# Check if temp table exists
TEMP_EXISTS=$(psql $PSQL_OPTS -t -c "SELECT EXISTS(SELECT 1 FROM pg_tables WHERE tablename='temp_import');" | tr -d '[:space:]')
if [ "$TEMP_EXISTS" != "t" ]; then
echo "Error: Temporary table 'temp_import' does not exist. Run without batch parameter first."
exit 1
fi
# Ensure triggers are disabled
psql $PSQL_OPTS -c "ALTER TABLE $TABLE_NAME DISABLE TRIGGER ALL;"
# Optimize PostgreSQL for better performance
psql $PSQL_OPTS -c "
-- Increase work mem for this session
SET work_mem = '256MB';
SET maintenance_work_mem = '1GB';
"
# Hard-code columns since we know them
CSV_COLUMNS='"pid","date","score","score2","qty_in_baskets","qty_sold","notifies_set","visibility_score","health_score","sold_view_score"'
echo "Using standard columns: $CSV_COLUMNS"
fi
# Step 4: Get total row count
TOTAL_ROWS=$(psql $PSQL_OPTS -t -c "SELECT COUNT(*) FROM temp_import;" | tr -d '[:space:]')
echo "Total rows to import: $TOTAL_ROWS"
# Calculate starting point
PROCESSED=$(( ($START_BATCH - 1) * $BATCH_SIZE ))
if [ $PROCESSED -ge $TOTAL_ROWS ]; then
echo "Error: Start batch $START_BATCH is beyond the available rows ($TOTAL_ROWS)"
exit 1
fi
# Step 5: Process in batches with shell loop
BATCH_NUM=$(( $START_BATCH - 1 ))
# We'll process batches in chunks of 10 before cleaning up
CHUNKS_SINCE_CLEANUP=0
while [ $PROCESSED -lt $TOTAL_ROWS ]; do
BATCH_NUM=$(( $BATCH_NUM + 1 ))
BATCH_START=$(date +%s)
MAX_ROWS=$(( $PROCESSED + $BATCH_SIZE ))
if [ $MAX_ROWS -gt $TOTAL_ROWS ]; then
MAX_ROWS=$TOTAL_ROWS
fi
echo "Processing batch $BATCH_NUM (rows $PROCESSED to $MAX_ROWS)..."
# Optimize query buffer for this batch
psql $PSQL_OPTS -c "SET work_mem = '256MB';"
# Insert batch
psql $PSQL_OPTS -c "
INSERT INTO $TABLE_NAME ($CSV_COLUMNS)
SELECT $CSV_COLUMNS
FROM temp_import
LIMIT $BATCH_SIZE
OFFSET $PROCESSED;
"
# Update progress
BATCH_END=$(date +%s)
BATCH_ELAPSED=$(( $BATCH_END - $BATCH_START ))
PROGRESS_PCT=$(echo "scale=2; $MAX_ROWS * 100 / $TOTAL_ROWS" | bc)
echo "Batch $BATCH_NUM committed in ${BATCH_ELAPSED}s, $MAX_ROWS of $TOTAL_ROWS rows processed ($PROGRESS_PCT%)" | tee -a "$PROGRESS_FILE"
# Increment counter
PROCESSED=$(( $PROCESSED + $BATCH_SIZE ))
CHUNKS_SINCE_CLEANUP=$(( $CHUNKS_SINCE_CLEANUP + 1 ))
# Check current row count every 10 batches
if [ $(( $BATCH_NUM % 10 )) -eq 0 ]; then
CURRENT_COUNT=$(psql $PSQL_OPTS -t -c "SELECT COUNT(*) FROM $TABLE_NAME;" | tr -d '[:space:]')
echo "Current row count in $TABLE_NAME: $CURRENT_COUNT" | tee -a "$PROGRESS_FILE"
# Every 10 batches, run an intermediate cleanup
if [ $CHUNKS_SINCE_CLEANUP -ge 10 ]; then
echo "Running intermediate cleanup and optimization..."
psql $PSQL_OPTS -c "VACUUM $TABLE_NAME;"
CHUNKS_SINCE_CLEANUP=0
fi
fi
# Optional - write a checkpoint file to know where to restart
echo "$BATCH_NUM" > "/tmp/import_last_batch_${TABLE_NAME}.txt"
done
# Only clean up if we've completed the import
if [ $PROCESSED -ge $TOTAL_ROWS ]; then
# Step 6: Re-enable triggers and clean up
echo "Re-enabling triggers and cleaning up..."
psql $PSQL_OPTS -c "
ALTER TABLE $TABLE_NAME ENABLE TRIGGER ALL;
DROP TABLE temp_import;
VACUUM ANALYZE $TABLE_NAME;
SELECT COUNT(*) AS \"Total rows in $TABLE_NAME\" FROM $TABLE_NAME;
"
else
echo "Import interrupted at batch $BATCH_NUM. To resume, run:"
echo "./psql-csv-import.sh $CSV_FILE $TABLE_NAME $BATCH_NUM"
fi
fi
# Calculate elapsed time
END_TIME=$(date +%s)
ELAPSED=$((END_TIME - START_TIME))
echo "Import completed successfully in ${ELAPSED}s ($(($ELAPSED / 60)) minutes)"
echo "Progress log saved to $PROGRESS_FILE"

View File

@@ -1,378 +0,0 @@
const { Client } = require('pg');
const path = require('path');
const fs = require('fs');
require('dotenv').config({ path: path.resolve(__dirname, '../.env') });
const dbConfig = {
host: process.env.DB_HOST,
user: process.env.DB_USER,
password: process.env.DB_PASSWORD,
database: process.env.DB_NAME,
port: process.env.DB_PORT || 5432
};
function outputProgress(data) {
if (!data.status) {
data = {
status: 'running',
...data
};
}
console.log(JSON.stringify(data));
}
// Explicitly define all metrics-related tables in dependency order
const METRICS_TABLES = [
'brand_metrics',
'brand_time_metrics',
'category_forecasts',
'category_metrics',
'category_sales_metrics',
'category_time_metrics',
'product_metrics',
'product_time_aggregates',
'sales_forecasts',
'temp_purchase_metrics',
'temp_sales_metrics',
'vendor_metrics',
'vendor_time_metrics',
'vendor_details'
];
// Tables to always protect from being dropped
const PROTECTED_TABLES = [
'users',
'permissions',
'user_permissions',
'calculate_history',
'import_history',
'ai_prompts',
'ai_validation_performance',
'templates',
'reusable_images'
];
// Split SQL into individual statements
function splitSQLStatements(sql) {
sql = sql.replace(/\r\n/g, '\n');
let statements = [];
let currentStatement = '';
let inString = false;
let stringChar = '';
for (let i = 0; i < sql.length; i++) {
const char = sql[i];
const nextChar = sql[i + 1] || '';
if ((char === "'" || char === '"') && sql[i - 1] !== '\\') {
if (!inString) {
inString = true;
stringChar = char;
} else if (char === stringChar) {
inString = false;
}
}
if (!inString && char === '-' && nextChar === '-') {
while (i < sql.length && sql[i] !== '\n') i++;
continue;
}
if (!inString && char === '/' && nextChar === '*') {
i += 2;
while (i < sql.length && (sql[i] !== '*' || sql[i + 1] !== '/')) i++;
i++;
continue;
}
if (!inString && char === ';') {
if (currentStatement.trim()) {
statements.push(currentStatement.trim());
}
currentStatement = '';
} else {
currentStatement += char;
}
}
if (currentStatement.trim()) {
statements.push(currentStatement.trim());
}
return statements;
}
async function resetMetrics() {
let client;
try {
outputProgress({
operation: 'Starting metrics reset',
message: 'Connecting to database...'
});
client = new Client(dbConfig);
await client.connect();
// Explicitly begin a transaction
await client.query('BEGIN');
// First verify current state
const initialTables = await client.query(`
SELECT tablename as name
FROM pg_tables
WHERE schemaname = 'public'
AND tablename = ANY($1)
AND tablename NOT IN (SELECT unnest($2::text[]))
`, [METRICS_TABLES, PROTECTED_TABLES]);
outputProgress({
operation: 'Initial state',
message: `Found ${initialTables.rows.length} existing metrics tables: ${initialTables.rows.map(t => t.name).join(', ')}`
});
// Disable foreign key checks at the start
await client.query('SET session_replication_role = \'replica\'');
// Drop all metrics tables in reverse order to handle dependencies
outputProgress({
operation: 'Dropping metrics tables',
message: 'Removing existing metrics tables...'
});
for (const table of [...METRICS_TABLES].reverse()) {
// Skip protected tables
if (PROTECTED_TABLES.includes(table)) {
outputProgress({
operation: 'Protected table',
message: `Skipping protected table: ${table}`
});
continue;
}
try {
// Use NOWAIT to avoid hanging if there's a lock
await client.query(`DROP TABLE IF EXISTS "${table}" CASCADE`);
// Verify the table was actually dropped
const checkDrop = await client.query(`
SELECT COUNT(*) as count
FROM pg_tables
WHERE schemaname = 'public'
AND tablename = $1
`, [table]);
if (parseInt(checkDrop.rows[0].count) > 0) {
throw new Error(`Failed to drop table ${table} - table still exists`);
}
outputProgress({
operation: 'Table dropped',
message: `Successfully dropped table: ${table}`
});
// Commit after each table drop to ensure locks are released
await client.query('COMMIT');
// Start a new transaction for the next table
await client.query('BEGIN');
// Re-disable foreign key constraints for the new transaction
await client.query('SET session_replication_role = \'replica\'');
} catch (err) {
outputProgress({
status: 'error',
operation: 'Drop table error',
message: `Error dropping table ${table}: ${err.message}`
});
await client.query('ROLLBACK');
// Re-start transaction for next table
await client.query('BEGIN');
await client.query('SET session_replication_role = \'replica\'');
}
}
// Verify all tables were dropped
const afterDrop = await client.query(`
SELECT tablename as name
FROM pg_tables
WHERE schemaname = 'public'
AND tablename = ANY($1)
`, [METRICS_TABLES]);
if (afterDrop.rows.length > 0) {
throw new Error(`Failed to drop all tables. Remaining tables: ${afterDrop.rows.map(t => t.name).join(', ')}`);
}
// Make sure we have a fresh transaction here
await client.query('COMMIT');
await client.query('BEGIN');
await client.query('SET session_replication_role = \'replica\'');
// Read metrics schema
outputProgress({
operation: 'Reading schema',
message: 'Loading metrics schema file...'
});
const schemaPath = path.resolve(__dirname, '../db/metrics-schema.sql');
if (!fs.existsSync(schemaPath)) {
throw new Error(`Schema file not found at: ${schemaPath}`);
}
const schemaSQL = fs.readFileSync(schemaPath, 'utf8');
const statements = splitSQLStatements(schemaSQL);
outputProgress({
operation: 'Schema loaded',
message: `Found ${statements.length} SQL statements to execute`
});
// Execute schema statements
for (let i = 0; i < statements.length; i++) {
const stmt = statements[i];
try {
const result = await client.query(stmt);
// If this is a CREATE TABLE statement, verify the table was created
if (stmt.trim().toLowerCase().startsWith('create table')) {
const tableName = stmt.match(/create\s+table\s+(?:if\s+not\s+exists\s+)?["]?(\w+)["]?/i)?.[1];
if (tableName) {
const checkCreate = await client.query(`
SELECT tablename as name
FROM pg_tables
WHERE schemaname = 'public'
AND tablename = $1
`, [tableName]);
if (checkCreate.rows.length === 0) {
throw new Error(`Failed to create table ${tableName} - table does not exist after CREATE statement`);
}
outputProgress({
operation: 'Table created',
message: `Successfully created table: ${tableName}`
});
}
}
outputProgress({
operation: 'SQL Progress',
message: {
statement: i + 1,
total: statements.length,
preview: stmt.substring(0, 100) + (stmt.length > 100 ? '...' : ''),
rowCount: result.rowCount
}
});
// Commit every 10 statements to avoid long-running transactions
if (i > 0 && i % 10 === 0) {
await client.query('COMMIT');
await client.query('BEGIN');
await client.query('SET session_replication_role = \'replica\'');
}
} catch (sqlError) {
outputProgress({
status: 'error',
operation: 'SQL Error',
message: {
error: sqlError.message,
statement: stmt,
statementNumber: i + 1
}
});
await client.query('ROLLBACK');
throw sqlError;
}
}
// Final commit for any pending statements
await client.query('COMMIT');
// Start new transaction for final checks
await client.query('BEGIN');
// Re-enable foreign key checks after all tables are created
await client.query('SET session_replication_role = \'origin\'');
// Verify metrics tables were created
outputProgress({
operation: 'Verifying metrics tables',
message: 'Checking all metrics tables were created...'
});
const metricsTablesResult = await client.query(`
SELECT tablename as name
FROM pg_tables
WHERE schemaname = 'public'
AND tablename = ANY($1)
`, [METRICS_TABLES]);
outputProgress({
operation: 'Tables found',
message: `Found ${metricsTablesResult.rows.length} tables: ${metricsTablesResult.rows.map(t => t.name).join(', ')}`
});
const existingMetricsTables = metricsTablesResult.rows.map(t => t.name);
const missingMetricsTables = METRICS_TABLES.filter(t => !existingMetricsTables.includes(t));
if (missingMetricsTables.length > 0) {
// Do one final check of the actual tables
const finalCheck = await client.query(`
SELECT tablename as name
FROM pg_tables
WHERE schemaname = 'public'
`);
outputProgress({
operation: 'Final table check',
message: `All database tables: ${finalCheck.rows.map(t => t.name).join(', ')}`
});
await client.query('ROLLBACK');
throw new Error(`Failed to create metrics tables: ${missingMetricsTables.join(', ')}`);
}
// Commit final transaction
await client.query('COMMIT');
outputProgress({
status: 'complete',
operation: 'Reset complete',
message: 'All metrics tables have been reset successfully'
});
} catch (error) {
outputProgress({
status: 'error',
operation: 'Reset failed',
message: error.message,
stack: error.stack
});
if (client) {
try {
await client.query('ROLLBACK');
} catch (rollbackError) {
console.error('Error during rollback:', rollbackError);
}
// Make sure to re-enable foreign key checks even if there's an error
await client.query('SET session_replication_role = \'origin\'').catch(() => {});
}
throw error;
} finally {
if (client) {
// One final attempt to ensure foreign key checks are enabled
await client.query('SET session_replication_role = \'origin\'').catch(() => {});
await client.end();
}
}
}
// Export if required as a module
if (typeof module !== 'undefined' && module.exports) {
module.exports = resetMetrics;
}
// Run if called from command line
if (require.main === module) {
resetMetrics().catch(error => {
console.error('Error:', error);
process.exit(1);
});
}

View File

@@ -1,337 +0,0 @@
/**
* This script updates the costeach values for existing orders from the original MySQL database
* without needing to run the full import process.
*/
const dotenv = require("dotenv");
const path = require("path");
const fs = require("fs");
const { setupConnections, closeConnections } = require('./import/utils');
const { outputProgress, formatElapsedTime } = require('./metrics/utils/progress');
dotenv.config({ path: path.join(__dirname, "../.env") });
// SSH configuration
const sshConfig = {
ssh: {
host: process.env.PROD_SSH_HOST,
port: process.env.PROD_SSH_PORT || 22,
username: process.env.PROD_SSH_USER,
privateKey: process.env.PROD_SSH_KEY_PATH
? fs.readFileSync(process.env.PROD_SSH_KEY_PATH)
: undefined,
compress: true, // Enable SSH compression
},
prodDbConfig: {
// MySQL config for production
host: process.env.PROD_DB_HOST || "localhost",
user: process.env.PROD_DB_USER,
password: process.env.PROD_DB_PASSWORD,
database: process.env.PROD_DB_NAME,
port: process.env.PROD_DB_PORT || 3306,
timezone: 'Z',
},
localDbConfig: {
// PostgreSQL config for local
host: process.env.DB_HOST,
user: process.env.DB_USER,
password: process.env.DB_PASSWORD,
database: process.env.DB_NAME,
port: process.env.DB_PORT || 5432,
ssl: process.env.DB_SSL === 'true',
connectionTimeoutMillis: 60000,
idleTimeoutMillis: 30000,
max: 10 // connection pool max size
}
};
async function updateOrderCosts() {
const startTime = Date.now();
let connections;
let updatedCount = 0;
let errorCount = 0;
try {
outputProgress({
status: "running",
operation: "Order costs update",
message: "Initializing SSH tunnel..."
});
connections = await setupConnections(sshConfig);
const { prodConnection, localConnection } = connections;
// 1. Get all orders from local database that need cost updates
outputProgress({
status: "running",
operation: "Order costs update",
message: "Getting orders from local database..."
});
const [orders] = await localConnection.query(`
SELECT DISTINCT order_number, pid
FROM orders
WHERE costeach = 0 OR costeach IS NULL
ORDER BY order_number
`);
if (!orders || !orders.rows || orders.rows.length === 0) {
console.log("No orders found that need cost updates");
return { updatedCount: 0, errorCount: 0 };
}
const totalOrders = orders.rows.length;
console.log(`Found ${totalOrders} orders that need cost updates`);
// Process in batches of 1000 orders
const BATCH_SIZE = 500;
for (let i = 0; i < orders.rows.length; i += BATCH_SIZE) {
try {
// Start transaction for this batch
await localConnection.beginTransaction();
const batch = orders.rows.slice(i, i + BATCH_SIZE);
const orderNumbers = [...new Set(batch.map(o => o.order_number))];
// 2. Fetch costs from production database for these orders
outputProgress({
status: "running",
operation: "Order costs update",
message: `Fetching costs for orders ${i + 1} to ${Math.min(i + BATCH_SIZE, totalOrders)} of ${totalOrders}`,
current: i,
total: totalOrders,
elapsed: formatElapsedTime((Date.now() - startTime) / 1000)
});
const [costs] = await prodConnection.query(`
SELECT
oc.orderid as order_number,
oc.pid,
oc.costeach
FROM order_costs oc
INNER JOIN (
SELECT
orderid,
pid,
MAX(id) as max_id
FROM order_costs
WHERE orderid IN (?)
AND pending = 0
GROUP BY orderid, pid
) latest ON oc.orderid = latest.orderid AND oc.pid = latest.pid AND oc.id = latest.max_id
`, [orderNumbers]);
// Create a map of costs for easy lookup
const costMap = {};
if (costs && costs.length) {
costs.forEach(c => {
costMap[`${c.order_number}-${c.pid}`] = c.costeach || 0;
});
}
// 3. Update costs in local database by batches
// Using a more efficient update approach with a temporary table
// Create a temporary table for each batch
await localConnection.query(`
DROP TABLE IF EXISTS temp_order_costs;
CREATE TEMP TABLE temp_order_costs (
order_number VARCHAR(50) NOT NULL,
pid BIGINT NOT NULL,
costeach DECIMAL(10,3) NOT NULL,
PRIMARY KEY (order_number, pid)
);
`);
// Insert cost data into the temporary table
const costEntries = [];
for (const order of batch) {
const key = `${order.order_number}-${order.pid}`;
if (key in costMap) {
costEntries.push({
order_number: order.order_number,
pid: order.pid,
costeach: costMap[key]
});
}
}
// Insert in sub-batches of 100
const DB_BATCH_SIZE = 50;
for (let j = 0; j < costEntries.length; j += DB_BATCH_SIZE) {
const subBatch = costEntries.slice(j, j + DB_BATCH_SIZE);
if (subBatch.length === 0) continue;
const placeholders = subBatch.map((_, idx) =>
`($${idx * 3 + 1}, $${idx * 3 + 2}, $${idx * 3 + 3})`
).join(',');
const values = subBatch.flatMap(item => [
item.order_number,
item.pid,
item.costeach
]);
await localConnection.query(`
INSERT INTO temp_order_costs (order_number, pid, costeach)
VALUES ${placeholders}
`, values);
}
// Perform bulk update from the temporary table
const [updateResult] = await localConnection.query(`
UPDATE orders o
SET costeach = t.costeach
FROM temp_order_costs t
WHERE o.order_number = t.order_number AND o.pid = t.pid
RETURNING o.id
`);
const batchUpdated = updateResult.rowCount || 0;
updatedCount += batchUpdated;
// Commit transaction for this batch
await localConnection.commit();
outputProgress({
status: "running",
operation: "Order costs update",
message: `Updated ${updatedCount} orders with costs from production (batch: ${batchUpdated})`,
current: i + batch.length,
total: totalOrders,
elapsed: formatElapsedTime((Date.now() - startTime) / 1000)
});
} catch (error) {
// If a batch fails, roll back that batch's transaction and continue
try {
await localConnection.rollback();
} catch (rollbackError) {
console.error("Error during batch rollback:", rollbackError);
}
console.error(`Error processing batch ${i}-${i + BATCH_SIZE}:`, error);
errorCount++;
}
}
// 4. For orders with no matching costs, set a default based on price
outputProgress({
status: "running",
operation: "Order costs update",
message: "Setting default costs for remaining orders..."
});
// Process remaining updates in smaller batches
const DEFAULT_BATCH_SIZE = 10000;
let totalDefaultUpdated = 0;
try {
// Start with a count query to determine how many records need the default update
const [countResult] = await localConnection.query(`
SELECT COUNT(*) as count FROM orders
WHERE (costeach = 0 OR costeach IS NULL)
`);
const totalToUpdate = parseInt(countResult.rows[0]?.count || 0);
if (totalToUpdate > 0) {
console.log(`Applying default cost to ${totalToUpdate} orders`);
// Apply the default in batches with separate transactions
for (let i = 0; i < totalToUpdate; i += DEFAULT_BATCH_SIZE) {
try {
await localConnection.beginTransaction();
const [defaultUpdates] = await localConnection.query(`
WITH orders_to_update AS (
SELECT id FROM orders
WHERE (costeach = 0 OR costeach IS NULL)
LIMIT ${DEFAULT_BATCH_SIZE}
)
UPDATE orders o
SET costeach = price * 0.5
FROM orders_to_update otu
WHERE o.id = otu.id
RETURNING o.id
`);
const batchDefaultUpdated = defaultUpdates.rowCount || 0;
totalDefaultUpdated += batchDefaultUpdated;
await localConnection.commit();
outputProgress({
status: "running",
operation: "Order costs update",
message: `Applied default costs to ${totalDefaultUpdated} of ${totalToUpdate} orders`,
current: totalDefaultUpdated,
total: totalToUpdate,
elapsed: formatElapsedTime((Date.now() - startTime) / 1000)
});
} catch (error) {
try {
await localConnection.rollback();
} catch (rollbackError) {
console.error("Error during default update rollback:", rollbackError);
}
console.error(`Error applying default costs batch ${i}-${i + DEFAULT_BATCH_SIZE}:`, error);
errorCount++;
}
}
}
} catch (error) {
console.error("Error counting or updating remaining orders:", error);
errorCount++;
}
updatedCount += totalDefaultUpdated;
const endTime = Date.now();
const totalSeconds = (endTime - startTime) / 1000;
outputProgress({
status: "complete",
operation: "Order costs update",
message: `Updated ${updatedCount} orders (${totalDefaultUpdated} with default values) in ${formatElapsedTime(totalSeconds)}`,
elapsed: formatElapsedTime(totalSeconds)
});
return {
status: "complete",
updatedCount,
errorCount
};
} catch (error) {
console.error("Error during order costs update:", error);
return {
status: "error",
error: error.message,
updatedCount,
errorCount
};
} finally {
if (connections) {
await closeConnections(connections).catch(err => {
console.error("Error closing connections:", err);
});
}
}
}
// Run the script only if this is the main module
if (require.main === module) {
updateOrderCosts().then((results) => {
console.log('Cost update completed:', results);
// Force exit after a small delay to ensure all logs are written
setTimeout(() => process.exit(0), 500);
}).catch((error) => {
console.error("Unhandled error:", error);
// Force exit with error code after a small delay
setTimeout(() => process.exit(1), 500);
});
}
// Export the function for use in other scripts
module.exports = updateOrderCosts;