Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/apla/conf-fu
configuration for everything
https://github.com/apla/conf-fu
Last synced: about 2 months ago
JSON representation
configuration for everything
- Host: GitHub
- URL: https://github.com/apla/conf-fu
- Owner: apla
- Created: 2014-05-22T12:07:55.000Z (over 10 years ago)
- Default Branch: master
- Last Pushed: 2016-04-26T18:32:26.000Z (over 8 years ago)
- Last Synced: 2024-11-07T17:52:22.931Z (2 months ago)
- Language: JavaScript
- Homepage:
- Size: 117 KB
- Stars: 2
- Watchers: 2
- Forks: 1
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
conf-fu
==========
[](https://travis-ci.org/apla/conf-fu)
[](https://www.npmjs.org/package/conf-fu)
[](https://codecov.io/github/apla/conf-fu?branch=master)
Simple configuration module to handle complex problems
Install
-------
npm install -g conf-fu
Synopsis
--------
```javascript
var ConfFu = require ('conf-fu');
// init from files
var config = new ConfFu ({
configFile: path.join (configDir, 'index.json'),
fixupFile: path.join (configDir, 'fixup.json')
});
config.verbose = globalVerbose || false;
config.on ('ready', function () {
done();
});
config.on ('error', function (eOrigin, eType, eData, eFile) {
assert (false, 'wrong config');
});
// init from structures
var config = new ConfFu ({
config: {a: {b: {c: "<#who knows>"}}, x: true}
});
```
Goals
----------
1. Easy to use
Common formats, you can set up a location of configuration files.
If you already have one, you do not need to change files location for each project,
you can just point this location (set up environment) for all needed projects.
2. Scalable
Hierarchical, any section can live in a separate file, easy access to frequently changing sections.
3. Easy deployment
Configuration is actually is splitted to the project configuration and the instance fixup.
You can have as many fixups as you want, not only development and production.
4. Persistent
Current configuration snapshot is always on disk and in vcs. No code allowed within configuration.
Tools never change project config, only fixup allowed to change.
Those is not acceptable:
1. `require ('config.json')` — 1, 2, 3;
2. nconf - 4 (require code to load includes)
3. node-convict - 2 (no access to frequently changing sections), 3 (no separation beween project config and instance fixup)
4. dotenv - 2, 4 (recommended not to commit config)
5. node-config - 1 (cannot change location), 4 (instance name not stored on disk)
Configuration file
---------------
Configuration format when parsed must be presented with tree or key-value structure.
Each string can be basic or enchanted. `conf-fu` supports those enchantments:
1. variables
examples:
* `<$db.mongo.collection>` — linked to the other value in configuration
* `<$bool(on|off):db_unix_sock=on>` — variable with type and default value
* `<$http_domain>:<$http_port>` — string concatenation
2. placeholders
examples:
* `<#placeholder>` — just a placeholder, need to be fullfilled
* `<#optional:placeholder>` — you can omit this value in fixup
* `<#default:127.0.0.1>` — you can omit this value in fixup and default value will be used
3. includes
examples:
* `<>` — parsed content from filename need to be inserted into that node
Usage
---------------
You can use `conf-fu` standalone, integrate within project or use it in browser.
####node.js in standalone mode
conf-fu configuration is located within `.conf-fu` directory in project root.
Main configuration file named `project.json`, fixup directories located at same level as the `project.json` file.
1. `project.json` file is loaded and parsed. `json`, `ini` and `yaml` formats supported, new formats can be added easily;
2. `project.json` contents is scanned for includes in form `""`;
3. when all includes is loaded, config tree is scanned for variables (`"<$config.path.variable>"`) and placeholders (`"<#please fix me>"`, `"<#optional:please fix me>"`, `"<#default:127.0.0.1>"`);
4. `fixup.json` file loaded and checked, whether all variables and placeholders fulfilled;
5. if resulting config is fulfilled, `conf-fu` emits `ready` event; otherwise, `error` event emitted.
#####Environment variables to drive config
required:
* `CONF_FU` core file path, parent directory is assumed as config location
* `CONF_FU_FIXUP` fixup file path
* `INSTANCE` or `CONF_FU_INSTANCE` instance name
optional:
* `CONF_FU_PROJECT` (relative to project root, dir name to search for `project.json` file)
#### Integration to the existing project
You have two options to integrate conf-fu to existing projects:
1. Use as configuration management script
For example, one part of your project written on `php` and you want to add `conf-fu`.
You'll need to install `conf-fu` from npm, then just use it from command line to edit configuration,
validate it, fullfill variables and placeholders. Then, just ask `conf-fu` to store
configuration artifact in place where you can read it with `php`.
2. Use within code to get a configuration.
TODO
Links
---------------
* http://thejeffchao.com/blog/2013/09/26/an-alternative-node-dot-js-configuration-management-pattern/
* http://metaduck.com/03-konphyg.html
* https://github.com/mozilla/node-convict
* https://github.com/scottmotte/dotenv