CODEWITHSUNDEEP
linuxyaml
Piotr
Piotr

Reputation: 827

Linux: updating values in yaml file

I have a configuration file in YAML (Cassandra) and I would like to update a few values inside it. Do you have any suggestions for a script that would allow me to do this? Perhaps some existing examples?

INPUT:

cluster_name: 'CassandraCluster0'
initial_token:
seed_provider:
    - class_name: org.apache.cassandra.locator.SimpleSeedProvider
      parameters:
          - seeds: "127.0.0.1"

OUTPUT:

cluster_name: 'CassandraCluster0'
initial_token: 582023494802482234
seed_provider:
    - class_name: org.apache.cassandra.locator.SimpleSeedProvider
      parameters:
          - seeds: "10.12.3.4, 1.3.4.3"

Upvotes: 1

Views: 2379

Answers (1)

Anthon
Anthon

Reputation: 76578

You can of course update text in a YAML file with sed/awk/perl, but as with updating CSV, INI, XML, HTML using those tools, they horribly fail on more complex examples where values "all of a sudden" span multiple lines, or are otherwise semantically the same as what worked, but are no longer consisting of the same matched strings.

It is better to just use a programming language and a real parser (which is also the recommendation for the other formats). Here is how you can do so with Python and ruamel.yaml, of which I am the author. So if your input is in input.yml, this:

import ruamel.yaml as yaml

data = yaml.load(open('input.yml'), Loader=yaml.RoundTripLoader)
data['initial_token'] = 582023494802482234
data['seed_provider'][0]['parameters'][0]['seeds'] = "10.12.3.4, 1.3.4.3"
print yaml.dump(data, Dumper=yaml.RoundTripDumper)

will get you:

cluster_name: CassandraCluster0
initial_token: 582023494802482234
seed_provider:
- class_name: org.apache.cassandra.locator.SimpleSeedProvider
  parameters:
  - seeds: 10.12.3.4, 1.3.4.3

Please note that there are not quotes around the IP addresses, as they are not necessary. It is possible, but not very easy to fine control the style to only put (double) quotes around values. And even more difficult to do so around specific values, so I hope that is not an issue.

If you have to update a set of input files for these two values often you can make script that takes the filename, token and seed provider from the commandline:

update_yml input.yml 582023494802482234 "10.12.3.4, 1.3.4.3"

and as script update_yml:

#! /usr/bin/env python

import sys
import ruamel.yaml as yaml

data = yaml.load(open(sys.argv[1]), Loader=yaml.RoundTripLoader)
data['initial_token'] = sys.argv[2]
data['seed_provider'][0]['parameters'][0]['seeds'] = sys.argv[3]
print yaml.dump(data, Dumper=yaml.RoundTripDumper)

Upvotes: 1

Related Questions