Need a quick way to edit a JSON file in Ansible? Ansible has great built-in support for ini files, but a number of more modern applications are using JSON for config files.
ghetto_json lets you make some types of edits to JSON files, and remains simple enough that it's hopefully easier just to extend than to switch to a different module, and you won't feel too guilty just copy-pasting it into your codebase.
If commentjson is available, it will be used to read the file, but comments will be lost on save. This is useful for making edits to default configurations shipped with some applications; where the application itself supports comments in JSON, but they're not required.
Drop ghetto_json into your playbook's library folder,
which can be configured
but defaults to ./library inside a playbook.
Make in-place changes to simple JSON documents,
without having to resort to replace.
commentjson will be used, if available, but is explictly not required.
Python 2.7 may be required for some shlex functionality
(like working Unicode), which you probably don't care about.
The file on the target to edit.
A very simple object notation for the location of the property to edit, and its new value.
Mandatory automatic conversion will be applied. Supported conversions are listed below. Everything else will be left as a string:
- integers (
5,-17) true/falsenullunsetwill delete the keyappend:valuewill append a 'value' to an array if the value does not exist alreadyextend:value1, value2will append 'value1' and 'value2' to an array (if the values do not exist already)
For the example JSON document /foo/bar.json containing:
{ "a": 5, "b": {"c": 6, "d": "hello" }, "values": ["one", "two"] }
...you can run an invocation like:
- ghetto_json:
path=/foo/bar.json
a=7
b.c=yellow
b.d=unset
values=extend:two, three
...and the file will be left looking like:
{
"a": 7,
"b": {
"c": "yellow"
}
"values": [
"one",
"two",
"three"
]
}