fix: dbt unit tests feat without proper context

[devmessias]

Resolves #10539, resolves #10353 , resolves #10410

I made a very simple change to the code just to solve a similar problem 10353, but this change is obviously not the correct solution. I’ll wait for a discussion before writing any more advanced code.

Problem

absence of env_var and var in the ctx dictionary, which must be passed to the Jinja render method.

Solution

If a model has a unit test, the get_rendered method is called twice, each time with a different context. The issue arises with build_unit_test_manifest_from_test. When this method is called (

dbt-core/core/dbt/parser/unit_tests.py

Line 116 in f6088db

get_rendered(unit_test_node.raw_code, ctx, unit_test_node, capture_macros=True)

), the context doesn’t contain information about the var and environment variables (env_var).

Later, when the get_rendered method from clients/jinja.py calls render_template from dbt_common/clients/jinja.py, we encounter a compilation error due to the absence of env_var and var in the ctx dictionary, which must be passed to the Jinja render method. **The error will not appear if we don't have any kind of mixing with Nones, for example select {{ var('something,1) }} as result instead of select {{ var('something,1)*2 }} as result

Some questions:

Is this the expected behavior? Shouldn't the ctx used to render the string for the unit test include these variables?

Is there a reason why the same model is rendered more than once — once to build the model and again to render the unit test? Why is this string rendered a second time if it doesn’t seem to be used for any data manipulation?

models/_unit_tests.yml

version: 2

unit_tests:
  - name: test_unit_test
    model: test
    given: []
    expect:
      format: csv
      rows: |
        result
        10

models/test.sql

select {{ var('foo', 10) * 1 }} as result

This will only work for #10539 if ctx is constructed using env_var and var, but before coding any stuff I'll wait for the dbt-core team

Checklist

• I have read the contributing guide and understand what's expected of me.
• I have run this code in development, and it appears to resolve the stated issue.
• This PR includes tests, or tests are not required or relevant for this PR.
• This PR has no interface changes (e.g., macros, CLI, logs, JSON artifacts, config files, adapter interface, etc.) or this PR has already received feedback and approval from Product or DX.
• This PR includes type annotations for new and modified functions.

[github-actions]

Thank you for your pull request! We could not find a changelog entry for this change. For details on how to document a change, see the contributing guide.

[codecov]

Codecov Report

Attention: Patch coverage is 75.75758% with 8 lines in your changes missing coverage. Please review.

Project coverage is 89.13%. Comparing base (40c350f) to head (7def903).
Report is 6 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #10849      +/-   ##
==========================================
- Coverage   89.18%   89.13%   -0.06%     
==========================================
  Files         183      183              
  Lines       23429    23466      +37     
==========================================
+ Hits        20895    20916      +21     
- Misses       2534     2550      +16     

Flag	Coverage Δ
integration	86.41% <75.75%> (-0.08%)	⬇️
unit	62.06% <15.15%> (-0.06%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Components	Coverage Δ
Unit Tests	62.06% <15.15%> (-0.06%)	⬇️
Integration Tests	86.41% <75.75%> (-0.08%)	⬇️

[devmessias]

Sorry to bother you @gshank, I looked into the git history in those methods, and it seems you're probably the best person to ask these questions.

[devmessias]

Okay, I dove into the code and found a way to make this work for me by capturing the variable and environment variable using the same logic used for ref, source and metrics