Add scraperLogger.js with credential redaction (#6)

Co-authored-by: Stephen Minakian <stephenminakian@gmail.com>
Co-committed-by: Stephen Minakian <stephenminakian@gmail.com>

services/scraperLogger.js

@@ -0,0 +1,164 @@
+/**
+ * Scraper-specific structured JSON logger
+ * Produces one JSON object per line to stdout
+ */
+
+const LEVELS = {
+  info: 'info',
+  warn: 'warn',
+  error: 'error'
+};
+
+/**
+ * Create a logger instance scoped to a job ID
+ * @param {string} jobId - Job identifier for correlation
+ * @returns {Object} Logger object with info, warn, error methods
+ */
+function createLogger(jobId) {
+  /**
+   * Internal log function
+   * @param {string} level - Log level
+   * @param {string} message - Log message
+   * @param {Object} context - Additional context data
+   */
+  const log = (level, message, context = {}) => {
+    const entry = {
+      timestamp: new Date().toISOString(),
+      level,
+      message: truncateMessage(message),
+      jobId,
+      context: sanitizeContext(context)
+    };
+
+    // Output as single-line JSON
+    console.log(JSON.stringify(entry));
+  };
+
+  return {
+    info: (message, context) => log(LEVELS.info, message, context),
+    warn: (message, context) => log(LEVELS.warn, message, context),
+    error: (message, context) => log(LEVELS.error, message, context)
+  };
+}
+
+/**
+ * Truncate message to prevent log bloat
+ * @param {string} message - Message to truncate
+ * @param {number} maxLength - Maximum length (default 1000)
+ * @returns {string} Truncated message
+ */
+function truncateMessage(message, maxLength = 1000) {
+  if (message === null || message === undefined) {
+    return '';
+  }
+  const str = String(message);
+  if (str.length <= maxLength) {
+    return str;
+  }
+  return str.substring(0, maxLength) + '... [truncated]';
+}
+
+/**
+ * Recursively process an object to handle Buffers, circular references, and sensitive data
+ * @param {*} obj - Object to process
+ * @param {WeakSet} seen - Set of seen objects for circular reference detection
+ * @returns {*} Processed value
+ */
+function processValue(obj, seen = new WeakSet()) {
+  // Handle null/undefined
+  if (obj === null) {
+    return null;
+  }
+  if (obj === undefined) {
+    return null;
+  }
+
+  // Handle Buffer BEFORE checking for object (Buffer is an object)
+  if (Buffer.isBuffer(obj)) {
+    return `[Buffer: ${obj.length} bytes]`;
+  }
+
+  // Handle strings - check for MongoDB connection strings
+  if (typeof obj === 'string') {
+    if (/mongodb(\+srv)?:\/\//.test(obj)) {
+      return redactConnectionString(obj);
+    }
+    return obj;
+  }
+
+  // Handle primitives
+  if (typeof obj !== 'object') {
+    return obj;
+  }
+
+  // Handle circular references
+  if (seen.has(obj)) {
+    return '[Circular]';
+  }
+  seen.add(obj);
+
+  // Handle arrays
+  if (Array.isArray(obj)) {
+    return obj.map(item => processValue(item, seen));
+  }
+
+  // Handle plain objects
+  const result = {};
+  const sensitiveKeys = ['password', 'secret', 'token', 'apikey', 'authorization'];
+
+  for (const key of Object.keys(obj)) {
+    // Check for sensitive keys
+    if (sensitiveKeys.some(k => key.toLowerCase().includes(k))) {
+      result[key] = '[REDACTED]';
+    } else {
+      result[key] = processValue(obj[key], seen);
+    }
+  }
+
+  return result;
+}
+
+/**
+ * Sanitize context object for safe logging
+ * - Remove circular references
+ * - Redact sensitive data
+ * - Handle special types (Buffer, undefined)
+ * @param {Object} context - Context object
+ * @returns {Object} Sanitized context
+ */
+function sanitizeContext(context) {
+  if (!context || typeof context !== 'object') {
+    return {};
+  }
+
+  try {
+    return processValue(context);
+  } catch (error) {
+    // If sanitization fails, return empty context
+    return { sanitizationError: 'Failed to sanitize context' };
+  }
+}
+
+/**
+ * Redact credentials from MongoDB connection string
+ * @param {string} uri - Connection string
+ * @returns {string} Redacted string
+ */
+function redactConnectionString(uri) {
+  try {
+    // Match mongodb://user:pass@host or mongodb+srv://user:pass@host
+    return uri.replace(
+      /mongodb(\+srv)?:\/\/([^:]+):([^@]+)@/,
+      'mongodb$1://[user]:[REDACTED]@'
+    );
+  } catch {
+    return '[REDACTED CONNECTION STRING]';
+  }
+}
+
+module.exports = {
+  createLogger,
+  LEVELS,
+  // Export internal functions for testing
+  redactConnectionString
+};