ConfigHound makes it easy to load configuration data.
ConfigHound.load
supports config in JSON, YAML or TOML formats, and returns raw Ruby data.
config = ConfigHound.load("config.yml") # or "config.{json,toml}"
As well as local files, you can load from any URI supported by OpenURI
, e.g.
# load over HTTP
ConfigHound.load("http://config-source/app-config.yml")
# load from S3
require "open-uri-s3"
config = ConfigHound.load("s3://config-bucket/app-config.json")
If you specify a list of config sources, ConfigHound will load them all, and deep-merge the data. Files specified earlier in the list take precedence.
ConfigHound.load(["config.yml", "defaults.yml"])
You can include raw data (Hashes) in the list, too, which is handy if you have defaults or overrides already in Ruby format.
overrides = { ... }
ConfigHound.load([overrides, "config.yml"])
You can also "include" other file (or URIs) from within a config file.
Just list the paths under the key _include
.
For example, in config.yml
:
pool:
size: 10
log:
file: "app.log"
_include:
- defaults.yml
then in defaults.yml
log:
level: INFO
pool:
size: 1
Values in the original config file take precedence over those from included files. Multiple levels of inclusion are possible.
If the placeholder "_include
" doesn't suit, you can specify
another, e.g.
config = ConfigHound.load("config.yml", :include_key => "defaults")
ConfigHound can expand references of the form <(X.Y.Z)>
in config values, which can help DRY up configuration, e.g.
name: myapp
aws:
region: us-west-1
log:
stream: <(name)>-logs-<(aws.region)>
Enable reference expansion with the :expand_refs
option.
ConfigHound.load(config_files, :expand_refs => true)
Reference expansion is performed after all config is loaded and merged, so you can reference config specified in other files.
It's on GitHub; you know the drill.
ConfigHound works well with: