Default Action Jobs

This page describes the jobs that are automatically available to all skeleton-based applications (particularly, applications with config.json files that include the framework's application.json config file). Mainly this is just a reference list with short descriptions of what the jobs do. But in some cases, there is comprehensive documentation about the interface of this job and how it can be parametrized (This would usually require changing your config.json configuration file).

These jobs can be invoked with the generator, e.g. as <jobname>.


Create api doc for the current library. Use the following macros to tailor the scope of classes that are going to show up in the customized apiviewer application:

"API_INCLUDE" = ["<class_patt1>", "<class_patt2>", ...]
"API_EXCLUDE" = ["<class_patt1>", "<class_patt2>", ...]

The syntax for the class pattern is like those for the include config key.

Classes, which are not covered by API_INCLUDE, are nevertheless included in the api data with their full class description if they are required for inheritance relationships (e.g. a class that is included derives from a class which is not). If such a required class is explicitly excluded with API_EXCLUDE, a stub entry for it will be included in the api data to just show the inheritance relationship.


Create the api data for the current library. This is included in the api job, but allows you to re-generate the api data .json files for the classes without re-generating the Apiviewer application as well. Moreover, you can supply class names as command line arguments to only re-generate the api data for those:

sh> api-data my.own.ClassA ...

Beware though that in such a case the tree information provided to the Apiviewer (i.e. what you see in the Apiviewer's tree view on the left) is also restricted to those classes (augmented by stubs for their ancestors for hierarchy resolution). But this should be fine for developing API documentation for specific classes.


Create build version of current application.


This job is available in the qx.Website skeleton. Create minified build version of current application.


This job is available in the qx.Website skeleton. Instead of building a single all-in-one qx.Website script file (see build and build-min) this will build all modules separately (qx.Website splitted in n files).


This job is available in the qx.Website skeleton. Instead of building a single all-in-one qx.Website script file (see build and build-min) this will build all modules separately (qx.Website splitted in n files) and minified.


Remove local cache and generated .js files (source/build).


This job is available in the qx.Mobile skeleton. The compile-scss job compiles SCSS files to CSS (See the article about mobile theming). The *.scss files usually reside in your application's source/theme/<name_space>/mobile/scss folder, and will be compiled into the css sibling folder. This job is run automatically during each source and build run. See watch-scss if you want automatic compilations when changing the SCSS files.


Create dependency information for the current library which is stored as a Json file (under source/script/dependencies.json). If this file exists and is current, the Generator will use its information when following dependencies of the classes of the library.

This is particularly interesting for libraries that are used in other applications. It allows you to speed up cold-cache builds for the other application. (Mind that it doesn't make much sense for the application itself, as a clean or distclean will also wipe the dependencies Json file).


Remove the cache and all generated artefacts of this library (source, build, ...).


Normalize whitespace in .js files of the current library (tabs, eol, ...).


Running this job will print out various information about your setup on the console. Information includes your qooxdoo and Python version, whether source and/or build version of your app has been built, stats on the cache, asf.


Check the source code of the .js files of the current library.


Migrate the .js files of the current library to the current qooxdoo version.

Running the migration job

Here is a sample run of the migration job:

./ migration
NOTE:    To apply only the necessary changes to your project, we
         need to know the qooxdoo version it currently works with.

Please enter your current qooxdoo version [1.0] :

Enter your qooxdoo version or just hit return if you are using the version given in square brackets.


Current qooxdoo version:   1.0
Upgrade path:              1.0.1 -> 1.1 -> 1.2

Affected Classes:

NOTE:    It is advised to do a ' distclean' before migrating any files.
         If you choose 'yes', a subprocess will be invoked to run distclean,
         and after completion you will be prompted if you want to
         continue with the migration. If you choose 'no', the distclean
         step will be skipped (which might result in potentially unnecessary
         files being migrated).

Do you want to run 'distclean' now? [yes] :

Enter "yes".

WARNING: The migration process will update the files in place. Please make
         sure, you have a backup of your project. The complete output of the
         migration process will be logged to 'migration.log'.

Do you want to start the migration now? [no] :

Enter "yes".

Check migration.log for messages that contain foo.js has been modified. Storing modifications ... to verify changes to class code.


Pretty-formatting of the source code of the current library.


Create a source version of the application, using the original file path for each class.

The source version of an application is tailored towards development activities. It makes it easy to write code, run the application, test, debug and inspect the application code, fix issues, add enhancements, and repeat.

With the source job all the classes of the application are in their original source form, and their files are directly loaded from their original paths on the file system. If you inspect your application in a JavaScript debugger like Firebug or Chrome Developer Tools, you can identify each file individually, read its code and comments, set breakpoints, inspect variables and so forth.

If you find yourself in a situation where you want to inspect more than your current application's class files in the debugger (e.g. because you are debugging another library along the way), this job is preferable.

You have to re-run this job when you introduce new dependencies, e.g. by instantiating a class you haven't used before. This changes the set of necessary classes for your application, and the generator has to re-create the corresponding loader.

There are two variants of the source job available which you might find interesting. One is called source-all and will include all available classes of all involved libraries, the other is source-hybrid which improves loading speed by concatenating some of the class code. See their respective entries.


Create a source version of the application, with all classes.

source-all will include all known classes, be they part of your application, the qooxdoo framework, or any other qooxdoo library or contribution you might be using. All those classes are included in the build, whether they are currently required or not. This allows you develop your code more freely as you don't have to re-generate the application when introducing new dependencies to existing classes. All classes are already there. You only have to re-run this job when you add an entirely new class that you want to use.

The downside of this job is that due to the number of classes your application is larger and loads slower in the browser, so it is a trade-off between development speed and loading speed.


Create a source version of the application, concatenating some of the class code.

The source-hybrid job concatenates the contents of the classes that make up the application into a few files, only leaving your own application classes separate. Having the other class files (framework, libraries, contribs) chunked together you get the loading speed of nearly the build version, while at the same time retaining the accessibility of your own application files for debugging. This makes this job ideal for fast and focused development of the application-specific classes.

Only the classes that are actually needed for the application are included, so you have to re-run this job when you introduce new dependencies.

To review the three different source jobs, if you are just getting started with qooxdoo development, use the source-all version, which is the most convenient if you are not too impatient. If you are concerned about loading speed during development, but don't mind hitting the up and return keys in your shell window once in a while, go with the default source-hybrid job. If your emphasis on the other hand is on inspection, and you want to see exactly which class files get loaded into your application and which code they provide, the source version will be your choice.



Run a mini web server that serves the source version of an application. The web server will export as document root a root path common to all libraries used by the source version. This overcomes e.g. restrictions by modern browsers that do not allow XHR requests over the file:// protocol by default.

By default the server will randomly pick a free port on the local machine to run at. You can assign it a fixed port by setting the SOURCE_SERVER_PORT macro, e.g. like source-server -m SOURCE_SERVER_PORT:44161.



Same as source-server, but adds an automatic reload feature. The web server watches the loader file of the exported source version (usually source/script/<application>.js), and triggers an automatic reload of the application in the browser if this changes. You usually want to use this job together with the watch job (running separately) which automatically re-generates the loader when the application classes change. This way, both jobs work hand in hand to reload the most up-to-date version of the application in the browser whenever the source files change. If the generation fails, e.g. due to a syntax error, the loader is not updated and hence the browser not reloaded.

Like with source-server the server prints at startup the URL to the application's index.html. If you want to load the application through the reload server it is important that you use exactly this URL in your browser (including the trailing ".../index.html"). On requesting this URL the reload server will instrument the file with information necessary for the reload behavior.

The reload feature can also be used when running the main application from the file system (with the file:// protocol) or over a separate web server like Apache. In this case you just have to manually add the URL of the reload client script in the app's index.html, e.g. adding

<script type="text/javascript" src="http://localhost:44161/_active_reload/active_reload.js"/>

to the header section, assuming that 44161 is the port where the source server runs on. /_active_reload/active_reload.js is the URL path to the reload client script. You then load the application over your standard web server. Just the reload notification is handled over the source server.



This job is similar in intent to the source-server job. But instead of starting a dedicated web server, it will create a small web server configuration to be used with an already existing web server on your machine. Various popular web servers are supported (Apache, lighttpd, nginx) and it is usually straight-forward to include the generated configuration file into the main server/virtual host configuration. The file contains hints how to achieve that for the given server implementation. You can tweak most of the settings involved in the job (server type, server URL, ...), the config key behind it is web-server-config.

In this way the source version of your application is integrated with an existing web server environment which comes in handy if you e.g. want to interact with backend services that are already hosted on the same web server.

The generated configuration is actually template-driven so you can add your own templates if your web server is not yet supported.


Create a test runner app for unit tests of the current library.

  • Use the following macro to tailor the scope of classes in which unit test classes are searched for:

    "TEST_INCLUDE" = ["<class_patt1>", "<class_patt2>", ...]

    The syntax for the class pattern is like those for the include config key.

  • The libraries from the libraries job will be included when building the test application (the application containing your unit tests is a separate application which is loaded into the runner application).

  • If you want to break out from the reliance on the libraries job altogether, or have very specific settings that must be applied to the test application, you can provide a custom includer job common-tests which may contain a custom library key and other keys. But then you have to make sure it contains the Testrunner library as well.

    "common-tests" :
      "extend"    : [ "libraries" ],
      "let" :      { "LOCALES" : ["de", "de_DE", "fr", "fr_FR" ] },
      "library" :
        { "manifest" : "${QOOXDOO_PATH}/framework/Manifest.json" },
        { "manifest" : "${TESTRUNNER_ROOT}/Manifest.json" }
      "include" : ["testrunner.TestLoader", "${TEST_INCLUDE}", "${QXTHEME}"],
      "environment" :
        "qx.theme" : "${QXTHEME}",
        "qx.globalErrorHandling" : true
      "cache" :
        "compile" : "${CACHE}"

    This allows you to tailor most of the parameters that influence the creation of the test application.


Create a test runner app for unit tests (source version) of the current library.

The same customization interface applies as for the default test job.


Create .po files for the current library.


Validates the Config (config.json) - and recursively all included Configs - against a schema. If a jobname argument is given only this job map (within the root Config) is checked.

This job helps especially with nested config keys, where a misconfiguration might be silently disregarded by the Generator which eventually leads to unexpected behaviour.


Validates the Manifest.json against a schema.

Some entries in Manifest files are informational and therefore optional, others are required to successfully use the current library with the Generator. The job is especially helpful for developers of contributions, as those require some extra keys.


The watch job watches the source/class path of your application for changed JavaScript files, and automatically runs the default Generator job (usually "source-hybrid") in case of a change. The config key behind it is watch-files.

When you run the job the process will starting telling you the path it is watching, and will continue until you terminate it with Ctrl-C. On *ix like systems you can put the job in the shell's background with &, in order to get your shell prompt back. The job will continue running, and only produce some console output when its associate command is being run. In order to terminate it you have to bring it to the foreground again and then press Ctrl-C (Or you can use a process manager to kill it).

The implementation uses a simple polling mechanism to detect file changes, the check interval is configurable. There are technological alternatives that hook into OS kernel events, but these approaches come with a certain overhead and are more difficult to maintain cross-platform.


This job is available in the qx.Mobile skeleton. The watch-scss job watches SCSS files, and compiles them to CSS once they change (See the article about mobile theming). The *.scss files usually reside in your application's source/theme/<name_space>/scss folder, and will be compiled into source/resource<name_space>/css.