A helper application for deploying DocumentDb databases, collections and server scripts (stored procedures, triggers and user defined functions). Designed to handle both the creation and updating on deploy. A work in progress; issues are welcome.
Designed to be installed globally and then initialized in your project directory.
Tested with Node 6.9.2
npm install -g azdocdb
In your project directory run:
azdocdb init
This will create azdocdb.json file that controls all configuration such as documentdb connection details.
{
"globs": ["**/*.js"], //Global walk-sync minimatch.Minimatch glob of files and directories to include
"ignore": ["azdocdb.json"], //ignore these globs
"env": [
{
"id": "dev", //A unique
"endpointUri": "<your endpoint URI>", //From your documentdb setup
"primaryKey": "<endpoint primary key>", //From your documentdb setup
"database": "<database name>", //Database to create and connect to
"collections": [
"<collection name>"
], //Array of collections to validate
"deploy": true, //Whether this environment should be deployed
"autoDeploy": false //Not yet used
"globs": [], //Environment specific include glob
"ignore": [] //ignore these globs
},
]
}
Use the create command to set server side scripts that contain all of your business logic.
azdocdb create <type> <filename>
e.g. azdocdb create sp my-first-sp.js
This will create a template file in the current directory, using the filename as a basis for the server scripts id and function name.
- its OK to change these, just ensure they meet documentDb's rules. Now just write your script in the serverScript.
module.exports = {
id: 'myFirstSp', //A unique ID for each object
type: 'sp', //Either sp or udf
collections:['activity', 'device'], //Array of collections in which object is to be written, empty for all
serverScript: function myFirstSp() {
var context = getContext();
var response = context.getResponse();
response.setBody("Your logic goes here in the body");
},
};
Run the deploy command. Optionally you can include and env ID and only that environment will be deployed. Otherwise every environment with deploy: true will be processed. Processing validates the database and each collection, creating them if they dont exist. Then it Processes each file included in the walk and upserts the server side script.
azdocdb deploy [env]
- [camelcase] (https://github.com/sindresorhus/camelcase) - Generating unique Ids from filenames
- [commander] (https://github.com/tj/commander.js) - Parsing your commands
- [documentdb-q-promises] (https://github.com/Azure/azure-documentdb-node-q) - Promising docuemntdb
- [elegant-status] (https://github.com/inikulin/elegant-status) - Your feedback loop
- [jsonfile] (https://github.com/jprichardson/node-jsonfile) - Lovely JSON writing
- [lodash] (https://lodash.com) - Because _!
- [q] (https://github.com/kriskowal/q) - Promises, promises!
- [walk-sync] (https://github.com/joliss/node-walk-sync) - Gathering all those server scripts for us
Contributions are welcome, send me a pull request or issue
This project is licensed under the MIT License - see the LICENSE file for details