Ansible get a subset of items from a list-CodePudding

I have a list of output from an MQ command that shows the status of a channel, the stdout_lines of my registered variable is;

'stdout_lines': [u'CHANNEL(SVRCONN.CHL) CHLTYPE(SVRCONN) CONNAME(1.2.3.4) STATUS(RUNNING) SUBSTATE(RECEIVE)']}

This out worked OK for what I was intending on doing as I need to display the output of the CHANNEL and STATUS only and could find the values using a combinations of .split(')')[1].split('(')[1] and split(')')[4].split('(')[1] to get values SVRCONN.CHL and RUNNING

I have also been using this task for CLUSTER channel but found it was failing, the reason for this was that there is was an extra field,

'stdout_lines': [u'CHANNEL(CLUSTER_CHL) CHLTYPE(CLUSRCVR) CONNAME(2.3.4.5) RQMNAME(CLUS00) STATUS(RUNNING) SUBSTATE(RECEIVE)']}

now i could do a check to see if RQMNAME is in the output and then use a different split, but is there a way that this can be done using a string to find the location in output of STATUS?

CodePudding user response：

The best way is to convert it to real structured data, and you can then access any value that you need. The following is a somewhat hacky way to achieve that, but works:

- hosts: localhost
  tasks:
    - debug:
        msg: "{{ cooked.CHANNEL }} / {{ cooked.STATUS }}"
      vars:
        cooked: "{{ dict(item | split | map('regex_replace', '\\((. )\\)$', ' \\1') | map('split')) }}"
      loop:
        - 'CHANNEL(SVRCONN.CHL) CHLTYPE(SVRCONN) CONNAME(1.2.3.4) STATUS(RUNNING) SUBSTATE(RECEIVE)'
        - 'CHANNEL(CLUSTER_CHL) CHLTYPE(CLUSRCVR) CONNAME(2.3.4.5) RQMNAME(CLUS00) STATUS(RUNNING) SUBSTATE(RECEIVE)'

To parse this into a dictionary we split on whitespace, convert each key(value) into key value, split on whitespace again to separate keys from values, then pass that into dict().

PLAY [localhost] ****************************************************************

TASK [debug] ********************************************************************
ok: [localhost] => (item=CHANNEL(SVRCONN.CHL) CHLTYPE(SVRCONN) CONNAME(1.2.3.4) STATUS(RUNNING) SUBSTATE(RECEIVE)) => {
    "msg": "SVRCONN.CHL / RUNNING"
}
ok: [localhost] => (item=CHANNEL(CLUSTER_CHL) CHLTYPE(CLUSRCVR) CONNAME(2.3.4.5) RQMNAME(CLUS00) STATUS(RUNNING) SUBSTATE(RECEIVE)) => {
    "msg": "CLUSTER_CHL / RUNNING"
}

CodePudding user response：

Given the list

lines:
  - CHANNEL(SVRCONN.CHL) CHLTYPE(SVRCONN) CONNAME(1.2.3.4) STATUS(RUNNING) SUBSTATE(RECEIVE)
  - CHANNEL(CLUSTER_CHL) CHLTYPE(CLUSRCVR) CONNAME(2.3.4.5) RQMNAME(CLUS00) STATUS(RUNNING) SUBSTATE(RECEIVE)

You can create the list of dictionaries without iteration. For example,

cooked: "{{ lines|map('split')|
                  map('map', 'regex_replace', '\\((. )\\)$', ' \\1')|
                  map('map', 'split')|
                  map('community.general.dict')|
                  list }}"

gives

cooked:
  - CHANNEL: SVRCONN.CHL
    CHLTYPE: SVRCONN
    CONNAME: 1.2.3.4
    STATUS: RUNNING
    SUBSTATE: RECEIVE
  - CHANNEL: CLUSTER_CHL
    CHLTYPE: CLUSRCVR
    CONNAME: 2.3.4.5
    RQMNAME: CLUS00
    STATUS: RUNNING
    SUBSTATE: RECEIVE

^{Example of a complete playbook for testing
- hosts: localhost

vars:

lines:
- 'CHANNEL(SVRCONN.CHL) CHLTYPE(SVRCONN) CONNAME(1.2.3.4) STATUS(RUNNING) SUBSTATE(RECEIVE)'
- 'CHANNEL(CLUSTER_CHL) CHLTYPE(CLUSRCVR) CONNAME(2.3.4.5) RQMNAME(CLUS00) STATUS(RUNNING) SUBSTATE(RECEIVE)'

cooked: "{{ lines|map('split')|
map('map', 'regex_replace', '\$(. )\$$', ' \\1')|
map('map', 'split')|
map('community.general.dict')|
list }}"

tasks:

- debug:
var: cooked
- debug:
msg: "{{ item.CHANNEL }} / {{ item.STATUS }}"
loop: "{{ cooked }}"}