Bug 1924952

Description of problem:
Following the instructions in https://access.redhat.com/documentation/en-us/red_hat_openstack_platform/16.1/html/advanced_overcloud_customization/chap-configuration_hooks#sect-Customizing_Hieradata_for_Individual_Nodes
leads to a json decoding error.


Steps to Reproduce:
1.
Create a template with the following:
resource_registry:
  OS::TripleO::ComputeExtraConfigPre: /usr/share/openstack-tripleo-heat-templates/puppet/extraconfig/pre_deploy/per_node.yaml

parameter_defaults:
  NodeDataLookup:
    {"C06C60C2-EC19-4237-B7EB-507F9E303EFD": {"nova::compute::cpu_dedicated_set": [ "1-4" ]}}

2.
Run a stack update

Actual results:
(heat-config) [INFO] b\\'\\'\\n[2021-02-03 23:55:58,781] (heat-config) [DEBUG] b\\'Traceback (most recent call last):\\\\n  File \"<string>\", line 5, in <module>\\\\n  File \"/usr/
lib64/python3.6/json/__init__.py\", line 354, in loads\\\\n    return _default_decoder.decode(s)\\\\n  File \"/usr/lib64/python3.6/json/decoder.py\", line 339, in decode\\\\n    obj, end = self.raw_decode(s, idx=_w(s, 0).end())\\\\n  File
\"/usr/lib64/python3.6/json/decoder.py\", line 355, in raw_decode\\\\n    obj, end = self.scan_once(s, idx)\\\\njson.decoder.JSONDecodeError: Expecting property name enclosed in double quotes: line 1 column 2 (char 1)\\\\nTraceback (most r
ecent call last):\\\\n  File \"<string>\", line 5, in <module>\\\\n  File \"/usr/lib64/python3.6/json/__init__.py\", line 354, in loads\\\\n    return _default_decoder.decode(s)\\\\n  File \"/usr/lib64/python3.6/json/decoder.py\", line 339
, in decode\\\\n    obj, end = self.raw_decode(s, idx=_w(s, 0).end())\\\\n  File \"/usr/lib64/python3.6/json/decoder.py\", line 355, in raw_decode\\\\n    obj, end = self.scan_once(s, idx)\\\\njson.decoder.JSONDecodeError: Expecting proper
ty name enclosed in double quotes: line 1 column 2 (char 1)\\\\n\\'\\n

Expected results:
A file located at /etc/puppet/hieradata/<uuid>.json is created with the following content:
{"nova::compute::cpu_dedicated_set": [ "1-4" ]}}

Additional info:
It seems like the input to the script is coming in as a map/dictionary (using single quotes) rather than a proper json string, which causes json.loads to fail with the aforementioned decoding error.

  inputs:
    - name: node_lookup
      description:
      type: String
      value: |-
        {'C06C60C2-EC19-4237-B7EB-507F9E303EFD': {'nova::compute::cpu_dedicated_set': ['1-4']}}

Workaround:
Replacing json.loads(input) with ast.literal_eval(input) looks like it will properly create a dictionary object in python whether it's a json string or the single-quoted mapping style that heat is using.

Here's a patch demonstrating the changes:
--- a/usr/share/openstack-tripleo-heat-templates/puppet/extraconfig/pre_deploy/per_node.yaml    2020-08-21 20:32:02.000000000 -0400
+++ b/usr/share/openstack-tripleo-heat-templates/puppet/extraconfig/pre_deploy/per_node.yaml    2021-02-03 19:28:20.090743051 -0500
@@ -47,16 +47,18 @@
         echo $node_lookup | $(get_python) -c "
         import json
         import sys
+        import ast
         input = sys.stdin.readline() or '{}'
-        cnt = json.loads(input)
+        cnt = ast.literal_eval(input)
         print(json.dumps(cnt.get('${node_id}', {})))
         " > /etc/puppet/hieradata/${node_id}.json
         # handle upper case node id LP#1816652
         echo $node_lookup | $(get_python) -c "
         import json
         import sys
+        import ast
         input = sys.stdin.readline() or '{}'
-        cnt = json.loads(input)
+        cnt = ast.literal_eval(input)
         print(json.dumps(cnt.get('${node_id_upper}', {})))
         " > /etc/puppet/hieradata/${node_id_upper}.json