Express plugin for Debugr

This plugin provides Debugr middleware for the Express web server. With this middleware a new task will automatically be started for each HTTP request that Express handles; additionally, both the HTTP request and the HTTP response or any error which reaches Express error handling will be automatically logged. If the HTTP response code is >= 500 (or >= 400 if the e4xx option is set) the response will be logged with the configured error level.

Installation

npm install --save @debugr/express

Usage

import * as express from 'express';
import { Logger, LogLevel } from '@debugr/core';
import { ExpressCollector } from '@debugr/express';
import { ConsoleHandler } from '@debugr/console';

const globalContext = {
  applicationName: 'example',
};

const logger = new Logger(globalContext, [
  new ConsoleHandler({
    threshold: LogLevel.INFO,
  }),
  new ExpressCollector(),
]);

const app = express();

// as the very first middleware:
app.use(logger.getPlugin('express').createRequestHandler());

// apply your other middlewares like body parser and your routes
app.post('/my-api', function(req, res) {
  // ...
});

// and then as the very last middleware - this will log
// any uncaught errors thrown in middlewares and request handlers:
app.use(logger.getPlugin('express').createErrorHandler());

app.listen(8000);

Options

The ExpressCollector constructor accepts an optional options object with the following keys as the first argument:

| Option | Type | Default | Description | |---------------------------|------------|-------------------------------|---------------------------------------------------------------------------------------------------------| | level | LogLevel | Logger.INFO | The level at which the request and response will be logged | | errorLevel | LogLevel | Logger.ERROR | The level at which error responses will be logged | | uncaughtLevel | LogLevel | options.errorLevel | The level at which uncaught errors which reach the collector's error handler middleware will be logged | | e4xx | boolean | false | Consider HTTP 4xx status code as an error response and log appropriately | | captureBody | | | See below; global setting for both request and response | | excludeHeaders | string[] | | Redact the contents of the specified headers when logging; global setting for both request and response | | request | object | | Request-specific logging options | | request.captureBody | | | See below; overrides global setting | | request.excludeHeaders | string[] | ['Authorization', 'Cookie'] | Redact the contents of the specified headers when logging; overrides global setting | | response | object | | Response-specific logging options | | response.captureBody | | | See below; overrides global setting | | response.excludeHeaders | string[] | ['Set-Cookie'] | Redact the contents of the specified headers when logging; overrides global setting |

captureBody

The captureBody option controls whether the request or response body will be captured in the debug log. It can be set in many ways, so it bears explaining in more detail:

• A boolean simply means what boolean usually means - captureBody: true will capture the body always. This is probably slightly overkill. Typically, you'll use this to disable capturing the raw request body if you're sure you don't ever care about it - e.g. if you log the decoded body by another plugin.
• A number means "max size in bytes" - the body will be captured if its size is at most this value.
• A (comma-separated) string or string[] means "content type". The body will be captured if it matches the given content type. You can use * as a simple placeholder for one or more characters, so you can use e.g. text/*.
• An object with string keys and number values combines content type and content length conditions, so you can use e.g. { 'text/*': 2e6 } to allow capturing any text body up to 2MB in size. Note that you can still use comma-separated types in the keys.

The default value for the captureBody option, for both request and response, is { 'text/*, application/json': 2e6 }. What this means for capturing a given request or response body is left as an exercise for the reader.