I have a very simple Liquid map (extract here below):
{
{% if content.DisplayLastName__c %}
"DisplayLastName__c": "{{ content.DisplayLastName__c }}",
{% endif %}
}
The input message of the map is the following:
{
"DisplayLastName__c": "é\" r",
"FirstName": "é\""
}
I got this error due to the fact that there is a \" in one field: "After parsing a value an unexpected character was encountered".
Am I missing something?
The following worked for me, however I don't fully know why:
{% assign vDescription = system.Data.Description | Replace: '\"', '\"' %}
Related
I'm having trouble trying to concatenate variables (one with a single quote) in Jinja. My code looks like this:
{%- set my_quote = "'" -%}
{%- set invocation = invocation_id -%}
And the output I'm attempting is this:
{{ invocation ~ my_quote }}
The output from this is:
f21f9039-44e5-452f-8d7a-ee64245ada23'
Ok great! Now when I attempt to add the single quote to the beginning as well:
{{ my_quote ~ invocation ~ my_quote }}
The output is the invocation variable value without any single quotes:
f21f9039-44e5-452f-8d7a-ee64245ada23
How can I get this to output both single quotes properly?
Try the following:
"'{{ invocation_id }}'"
oh I love this question and the great reproducible example you provided. First thought, what happens if you escape the ' like this?
{%- set my_quote = "\'" -%}
Is there a reason it has to be part of the variable?
Could you do:
'{{ invocation }}'
In the HL7v2 to FHIR conversion CodeSystem templates, their default, i.e., "else" part in the if-then-else rule, is not working.
For example, for Patient.communication.language in the Resource/_Patient.liquid template, it has the following line to map a language code to a dictionary containing "code", "display", and "system".
{% include 'DataType/CWECodeableConcept' mapping: 'CodeSystem/Language', CWE: PID.15 -%}
In the CodeSystem/_Language.liquid template, if no matching code, it would output the input values.
{% else -%}
"code" : "{{ inCode }}",
"display" : "{{ inCode }}",
"system" : "",
{% endif -%}
My case has PID.15="English" and it has no matching code. So, I expect the following output.
{
"code" : "English",
"display" : "English",
"system" : "",
}
What happens is no output at all.
I also tried to add the following rule with no success.
{% elsif inCode == 'English' -%}
"code" : "{{ inCode }}",
"display" : "{{ inCode }}",
"system" : "",
I appears that CodeSystem/_Language.liquid does not take changes, nor does it handle no-matched codes with "else" condition. I tested the changes with the "FHIR Converter" extension in Visual Studio Code. What can I do to make the "else" condition in CodeSystem/_Language.liquid working?
Try this:
Make required changes in the /CodeSystem/CodeSystem.json file and that should work for you.
Also, update the VS Code extension to the latest version.
"CodeSystem/Language": {
"A": {
"code": "A's code",
"display": "A's display",
"system": "A's system"
},
"__default__": {
"code": "English",
"display": "English",
"system": "English"
}
},
Explanation:
The following statement is calling _DataType/CWECodeableConcept and passing 'CodeSystem/Language' to mapping parameter, and the value of PID.15 to the CWE parameter.
{% include 'DataType/CWECodeableConcept' mapping: 'CodeSystem/Language', CWE: PID.15 -%}
Inside the _CWECodeableConcept.liquid, since mapping is populated it executes the if blocks, and not the else blocks.
{% if mapping -%}
"code":"{{ CWE.1.Value | get_property: mapping, 'code' }}",
"display":"{{ CWE.1.Value | get_property: mapping, 'display' }}",
"system":"{{ CWE.1.Value | get_property: mapping, 'system' }}",
"version":"{{ CWE.1.Value | get_property: mapping, 'version' }}",
{% else -%}
"code":"{{ CWE.1.Value }}",
"display":"{{ CWE.2.Value }}",
"system":"{{ CWE.3.Value }}",
"version":"{{ CWE.7.Value }}",
{% endif -%}
Inside the if block, get_property is used to fetch the code, display, system, and version for the mapping CodeSystem/Language corresponding to the value in CWE.(1|4|10).Value.
get_property uses the /CodeSystem/CodeSystem.json file, and not the individual files under the CodeSystem folder.
The behavior of get_property is documented here. However, the documentation does not mention the name of the config file it uses. We are fixing the documentation in the next release.
For quicker turnaround, we recommend raising an issue on the FHIR converter github repo.
Attempting to make a decision in a template based on the last character of a variable (third level domain hostname) , but the epiphany alludes me. Make a config stanza if value else, do the other.
I set a fact in play:
- name: Set third level domain name to a variable
set_fact:
my_3rd_levelname: "{{ ansible_nodename.split('.')[0] }}"
- name: Ascertain if which server we're on
set_fact:
my_one_or_two: "{{ my_3rd_levelname[-1]|int }}"
...which appears to echo out with debug, save the casting as an int...see below.
TASK [role-test : Echo out my_one_or_two] *******************************************************************************************************************
ok: [w.x.y.42] => {
"my_one_or_two": "2"
}
Then in the template.j2...
{# If my_one_or_two is even list server1 first. If not, second. #}
{% if lookup('vars,',my_one_or_two) + my_one_or_two|int is 1 %}
[some config file stanza here]
{% else %}
[some other config file stanza instead]
I've poked and hoped until I can stand it no longer and am reaching out. I've tried just using the raw variable, e.g., {% if my_one_or_two|int == 1 %} along with many other attempts, but I'm stuck. I can't seem to overcome this error:
AnsibleError: template error while templating string: expected token 'name', got 'integer'. String: [the contents of my template]
Any input would be greatly appreciated at this juncture.
Thanks
Okay...leaving this here in case someone else doesn't realize you can use any Python method that the object supports. Here's what I did. Remember the server names end in 1 or 2 and its a String.
Created a varible in /roles/[rolename]/vars...
my_simple_hostname: "{{ ansible_nodename.split('.')[0] }}"
Then used the 'endswith' method to evaluate it....
% if my_simple_hostname.endswith('1') == true %}
[content if true]
{% else %}
[content when false]
{% endif %}
I have a variable that is an array [{'foo':1},{'bar':2}].
I want to combine it with the following hash: {'baz':3} using a set fact (?) such as my output registered variable is:
[{'foo':1, 'baz':3},{'bar':2, 'baz':3}]
I've looked into the combine filter, but it only works when I already have an hash to work with. In my case I have an array.
Is there a way to achieve that using ansible?
Actually, I have found a way. map can be used with any filters, and arguments have to be passed after a comma
- name: test
set_fact:
_test: "{{ [{'foo':1}, {'bar':2}] | map('combine', {'baz':3}) | list }}"
produces:
ok: [localhost] => {
"_test": [
{
"baz": 3,
"foo": 1
},
{
"bar": 2,
"baz": 3
}
]
}
Jinja2 doesn't have list comprehension, but I think you can use set and for loop to achieve it:
{% set outputarray = [] -%}
{% for d in inputarray -%}
{% set r = d|combine({'baz': 3}) -%}
{{ ouputarray.append(r) and '' }}
{%- endfor %}
I have a token which has hyphen in it and I am trying to read it in jinja, Here is the sample code:
{
[
{
{% for curPool in data.pool %}
"name": "{{curPool.name}}",
"my-org":"{{curPool.my-org}}"
{% endfor %}
}
]
}
The variable curPool has two fields: name and my-org ( i have no choice of defining these tokens)
When I run this code, I get an error:
Can't render, error: 'unsupported operand type(s) for -: 'Undefined' and 'Undefined''
It seems to me that it does not know how to parse the hyphen in the token name. I have tried to escape it with '\', ''', and a '%'. But nothing works. Would appreciate your help
Thanks
dict.get() can be used:
"my-org":"{{curPool.get('my-org')}}"