The Divvun and Giellatekno teams build language technology aimed at minority and indigenous languages
For information on Phonegap, see phonegap/README.md
The frontend client uses Node.js’s environment, and specifically Brunch.io for compilation and project structure management. In order to prepare the development environment, first install Node.js and npm (Node Package Manager), then:
1.) In ~/main/apps/aajege/src/sma-client/ run npm install. 2.) As a convenience, add node_modules/.bin/ to your $PATH variable
Familiarize yourself a little with Brunch.io, but generally speaking you’ll be most interested in:
brunch watch --server
Buildin for release on the other hand will require:
brunch build --production
This will minify everything to prepare it for web or inclusion in apps.
This is just a short overview to the most important files and structure. For details, look at any source file For details, look at any source file
config.coffee: handles brunch configuration, file concatenation order overrides, minimization, and managing source directories.
package.json: build dependencies and installation configuration for npm.
app/: Source!
application.coffee: the main application file, handles initialization of all the basic stuff, routers, client-side database models and server/client synchronization.
routers/routers.coffee: URL routing and view processing. New views need to be set up here.
views/: The views directory should include one folder per view, each containing its own templates/ and styles/ folder. The build process automatically finds where to include new templates and styles, and there is no need to include them anywhere. Any general view documentation should be included in the main view file in each directory.
models/: one file per model or collection. Models should be well documented, if they aren’t, they need to be.
The database structure data is fetched from the server on app initialization and stored locally. (This means, word information, word relations to media files, etc.). Some user data is more or less always live, and data on user activity is synced automatically when a connection is available.
Backbone.js handles data storage, and generates models and collections for searching htesse.
General note on Collections and data fetching
TODO: @fetch method; @server.offline_media vs. @server.path
Concept is a general term for learning information. It may be a word, an image or an audio file, but the data is all heavily cross-linked so that it is easy to find a word for an image, or a related sound file.
Maintains the main screen category list, as well as is used for an organizational tool for question construction.
Several defaults are provided which happen to line up with the progression in Gïelese, however question types are possible to be defined on the server side. These are then fetched by the client for gameplay.