OpenJS Architect is an Infrastructure as Code (IaC) solution. The critical insight of Infastructure as Code is determinism. Infrastructure resources are defined in a declarative manifest file with the code that depends on them. This ensures deployment artifacts alway have the exact runtime resources expected for every version of the code.
Architect looks in the following places for the primary definition/configuration manifest file:
The .arc
format is unique to Architect with many readability advantages; but is not required
.arc
.arc
is a text format for storing structured configuration data; it is not for serializing or transporting data
The .arc
format:
- Comments follow
#
symbols - Top level keys start with
@
(example: @pragma
) - Pragmas contain: scalar values or complex values
- Scalar values are:
string
, number
and boolean
- Complex values are:
array
, vector
and map
- Whitespace is significant
Example
Consider a file some-arc-file.txt
with the following contents:
# this is a comment
@section-one
simple-string-value # String
another-value
4.2 # Number
true # Boolean
@section-of-arrays
vector of values
vector tuple
@vectors-section
named
vector
of
values
@this-section-has-a-map
hello-world
name some-value
Parsing the file with the following code:
#!/usr/bin/env node
const parser = require('@architect/parser')
const fs = require('fs')
const text = fs.readFileSync('./some-arc-file.txt').toString()
const result = parse(text)
console.log(result)
Prints the following plain object to the console:
{
"section-one": [
"simple-string-value",
"another-value",
4.2,
true
],
"section-of-arrays": [
["vector", "of", "values"],
["vector", "tuple"]
],
"vectors-section": [
{named: ["vector", "of", "values"]},
],
"this-section-has-a-map": [{
"hello-world": {
"name": "some-value"
}
}]
}
Things to notice:
array
values are space seperated scalar values on a single linevector
is a named array
with scalar values indented two spaces on newlinesmap
is a named value followed by keys and values indented two spaces