tubestation/devtools/shared/gcli/source/docs/developing-gcli.md

Developing GCLI

About the code

The majority of the GCLI source is stored in the lib directory.

The docs directory contains documentation. The scripts directory contains RequireJS that GCLI uses. The build directory contains files used when creating builds. The mozilla directory contains the mercurial patch queue of patches to apply to mozilla-central. The selenium-tests directory contains selenium web-page integration tests.

The source in the lib directory is split into 4 sections:

• lib/demo contains commands used in the demo page. It is not needed except for demo purposes.
• lib/test contains a small test harness for testing GCLI.
• lib/gclitest contains tests that run in the test harness
• lib/gcli contains the actual meat

GCLI is split into a UI portion and a Model/Controller portion.

The GCLI Model

The heart of GCLI is a Requisition, which is an AST for the input. A Requisition is a command that we'd like to execute, and we're filling out all the inputs required to execute the command.

A Requisition has a Command that is to be executed. Each Command has a number of Parameters, each of which has a name and a type as detailed above.

As you type, your input is split into Arguments, which are then assigned to Parameters using Assignments. Each Assignment has a Conversion which stores the input argument along with the value that is was converted into according to the type of the parameter.

There are special assignments called CommandAssignment which the Requisition uses to link to the command to execute, and UnassignedAssignmentused to store arguments that do not have a parameter to be assigned to.

The GCLI UI

There are several components of the GCLI UI. Each can have a script portion, some template HTML and a CSS file. The template HTML is processed by domtemplate before use.

DomTemplate is fully documented in [it's own repository] (https://github.com/joewalker/domtemplate).

The components are:

• Inputter controls the input field, processing special keyboard events and making sure that it stays in sync with the Requisition.
• Completer updates a div that is located behind the input field and used to display completion advice and hint highlights. It is stored in completer.js.
• Display is responsible for containing the popup hints that are displayed above the command line. Typically Display contains a Hinter and a RequestsView although these are not both required. Display itself is optional, and isn't planned for use in the first release of GCLI in Firefox.
• Hinter Is used to display input hints. It shows either a Menu or an ArgFetch component depending on the state of the Requisition
• Menu is used initially to select the command to be executed. It can act somewhat like the Start menu on windows.
• ArgFetch Once the command to be executed has been selected, ArgFetch shows a 'dialog' allowing the user to enter the parameters to the selected command.
• RequestsView Contains a set of RequestView components, each of which displays a command that has been invoked. RequestsView is a poor name, and should better be called ReportView

ArgFetch displays a number of Fields. There are fields for most of the Types discussed earlier. See 'Writing Fields' above for more information.

Testing

GCLI contains 2 test suites:

• JS level testing is run with the test command. The tests are located in lib/gclitest and they use the test runner in lib/test. This is fairly comprehensive, however it does not do UI level testing. If writing a new test it needs to be registered in lib/gclitest/index. For an example of how to write tests, see lib/gclitest/testSplit.js. The test functions are implemented in lib/test/assert.
• Browser integration tests are included in browser_webconsole_gcli_*.js, in toolkit/components/console/hudservice/tests/browser. These are run with the rest of the Mozilla test suite.

Coding Conventions

The coding conventions for the GCLI project come from the Bespin/Skywriter and Ace projects. They are roughly [Crockford] (http://javascript.crockford.com/code.html) with a few exceptions and additions:

• var does not need to be at the top of each function, we'd like to move to let when it's generally available, and let doesn't have the same semantic twists as var.

• Strings are generally enclosed in single quotes.

• eval is to be avoided, but we don't declare it evil.

The [Google JavaScript conventions] (https://google-styleguide.googlecode.com/svn/trunk/javascriptguide.xml) are more detailed, we tend to deviate in:

• Custom exceptions: We generally just use throw new Error('message');

• Multi-level prototype hierarchies: Allowed; we don't have goog.inherits()

• else begins on a line by itself:

    if (thing) {
      doThis();
    }
    else {
      doThat();
    }
  

Startup

Internally GCLI modules have startup()/shutdown() functions which are called on module init from the top level index.js of that 'package'.

In order to initialize a package all that is needed is to require the package index (e.g. require('package/index')).

The shutdown() function was useful when GCLI was used in Bespin as part of dynamic registration/de-registration. It is not known if this feature will be useful in the future. So it has not been entirely removed, it may be at some future date.

Running the Unit Tests

Start the GCLI static server:

cd path/to/gcli
node gcli.js

Now point your browser to http://localhost:9999/localtest.html. When the page loads the tests will be automatically run outputting to the console, or you can enter the test command to run the unit tests.

Contributing Code

Please could you do the following to help minimize the amount of rework that we do:

1. Check the unit tests run correctly (see Running the Unit Tests above)

2. Check the code follows the style guide. At a minimum it should look like the code around it. For more detailed notes, see Coding Conventions above

3. Help me review your work by using good commit comments. Which means 2 things

   • Well formatted messages, i.e. 50 char summary including bug tag, followed by a blank line followed by a more in-depth message wrapped to 72 chars per line. This is basically the format used by the Linux Kernel. See the commit log for examples. The be extra helpful, please use the "shortdesc-BUGNUM: " if possible which also helps in reviews.
   • Commit your changes as a story. Make it easy for me to understand the changes that you've made.

4. Sign your work. To improve tracking of who did what, we follow the sign-off procedure used in the Linux Kernel. The sign-off is a simple line at the end of the explanation for the patch, which certifies that you wrote it or otherwise have the right to pass it on as an open-source patch. The rules are pretty simple: if you can certify the below:

    Developer's Certificate of Origin 1.1
   
    By making a contribution to this project, I certify that:
   
    (a) The contribution was created in whole or in part by me and I
        have the right to submit it under the open source license
        indicated in the file; or
   
    (b) The contribution is based upon previous work that, to the best
        of my knowledge, is covered under an appropriate open source
        license and I have the right under that license to submit that
        work with modifications, whether created in whole or in part
        by me, under the same open source license (unless I am
        permitted to submit under a different license), as indicated
        in the file; or
   
    (c) The contribution was provided directly to me by some other
        person who certified (a), (b) or (c) and I have not modified
        it.
   
    (d) I understand and agree that this project and the contribution
        are public and that a record of the contribution (including all
        personal information I submit with it, including my sign-off) is
        maintained indefinitely and may be redistributed consistent with
        this project or the open source license(s) involved.
   

   then you just add a line saying

    Signed-off-by: Random J Developer <random@developer.example.org>
   

   using your real name (sorry, no pseudonyms or anonymous contributions.)

Thanks for wanting to contribute code.