HRD

HRD is the abbreviation for Human Readable Data.

We use it as the file format of all configuration files of JumpScale.

The system wide configuration files are in /optvar/hrd/.

Example of an HRD file

bootstrap.ip=localhost
bootstrap.login=root
bootstrap.passwd=rooter
bootstrap.type=ssh

HRD schema

An HRD schema defines the structure of an HRD file, based on which an HRD file can be generated.

Example:

email = descr:'comma separated list of email addresses' type:email alias:'mail,mailaddr' @ask list
mobile = descr:'comma separate list of mobile phone nrs' type:tel alias:'tel,landline' @ask list
expire = descr:'format $day:$month:$year' type:date alias:till @ask
test = type:int default:1
testf = type:float default:1.1
testb = type:bool default:False

Properties of an HRD schema:

• descr describes the field

• type for specifying the type of data for the field

  • Can be any of the following values: str, email, int, float, bool, multiline, tel, ipaddr, date
  • Date = epoch (int)

• default specifies the default value for the field

• regex for validating the entry against a regex

• minval/maxval for specifying minimum and maximum values for a field

  • Only relevant for fields of type int

• multichoice for specifying a list of items people can choose from, e.g. 'red,blue,orange'

• singlechoice for specifyng a single selection

• alias for setting an alias name or multiple alias names for a field

• @ask is a tag for specifying that the value needs to be provided for by the user

  • If this is not mentioned then the default value will be used

• list to specify that the field is a list

  • Can be a list of integers, strings, ...

• id a tag for specifying that the field is the identifier

  • If not specified name = $(instance) will be autoadded

• consume to specify the dependencies to other services

  • Format $role:$minamount:$maxamount,$role2:$min$max, ...
  • $minamount-$maxamount is optional
  • $role is role of other AYS service, e.g. node (consume service from a node)
  • Example: node:1:1,redis:1:3
  • The min-max is important because they define the dependency requirements, e.g. node:1:1 means I need 1 node to be in good shape and if node is not there I cannot function myself.

• parent for specifying the role

  • $role is role of other AYS server, e.g. node (consume service from a node)
  • Acts like consume $role:1:1 but has special (operational) meaning
  • When parent then the service instance will be subdir of parent in ays repo

• parentauto

  • is tag to parent
  • means will automatically create the parent if it does not exist yet

Consume example:

node = type:str list descr:'node on which we are installed' consume:node:1:1
etcd = type:str list consume:etcd:3:3
mongodb = type:str list consume:mongodb:1:3
nameserver = type:str list consume:ns

Get HRD from HRD schema

TODO:

Usage As template engine

Getting application instance HRD's

TODO: needs to be reworked
hrd=j.application.getAppInstanceHRD(name, instance, domain='jumpscale')
#then e.g. use
j.application.config.applyOnDir
j.application.config.applyOnFile

Getting system wide HRD's

they are all mapped under j.application.config you can e.g. use following 2 functions to apply your templates to dirs or files

TODO: needs to be reworked
j.application.config.applyOnDir
j.application.config.applyOnFile

to look at the HRD just go in ipshell & print the config

The templating function will look for template params $(hrdkey) and replace them

you can replace additional arguments e.g:

j.application.config.applyOnDir(adir,additionalArgs={"whoami","kds"})

would replace $(whoami) with kds additional to what found in hrd's