Getting ready

To run this recipe, we will need a script that we believe has a tendency to fail.

The script in this recipe is available in the downloadable code repository as recipe04.js under chapter02. If we run the provided example script, we must change to the root directory for the book's sample code.

How to do it…

Consider the following script:

phantom.onError = function(message, trace) {
  console.error('[PHANTOMJS ERROR] ' + message);
  trace.forEach(function(t) {
    console.error('  >> [' + t.line + '] ' +
      (t.function ? '[' + t.function + '] ' : '') +
      t.file || t.sourceURL);
  });
  phantom.exit(1);
};

function doSomeErrorProneStuff() {
  throw new Error('Gremlins fed after midnight.');
}

doSomeErrorProneStuff();

console.log('Exiting cleanly.');
phantom.exit(0);

Given the preceding script, enter the following at the command line:

phantomjs chapter02/recipe04.js

Our output should look like the following:

[PHANTOMJS ERROR] Error: Gremlins fed after midnight.
 >> [13] [doSomeErrorProneStuff] chapter02/recipe04.js
 >> [16] chapter02/recipe04.js

How it works…

Our preceding example script performs the following actions:

Note that once we enter the onError handler, we are not returned to our previous execution context. Depending on the specifications of our script, we will need to consider how to proceed—is it sufficient just to console out the error message and stack trace? Or do we need to reattempt an operation (for example, rerun a request with different arguments or a longer timeout, and so on)? By applying a function to phantom.onError, we create a global error handler that will catch all otherwise unhandled exceptions.

There's more…

As previously mentioned, when assigned, phantom.onError effectively creates a global error handler in the PhantomJS runtime. It is worth noting here that simply setting up a function on phantom.onError is not a substitute for safe code. Our scripts should still perform the appropriate checks (for example, for types and non-null values, and so on) and use the if or try-catch statements for flow control, wherever they make sense. However, there will be occasions where it makes sense to set up error handlers with onError; for example, when we cannot know all the places where an error might occur and we must ensure that our script exits, even if it exits with an error code.

See also