feat: Add configuration option to scope application listing / getting to particular namespace

[grahamalama]

In #854, AIU made changes so that it scanned applications across all namespaces. This introduced a few regressions for some use cases, namely #902.

In this PR, I introduce a config option so that users can optionally scope the K8s client to a specific namespace (as was the case before #854). Since searching across all namespaces was recently introduced as the default, I opted to keep that the default.

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 80.95238% with 4 lines in your changes missing coverage. Please review.

Project coverage is 63.60%. Comparing base (042785f) to head (3acfb8d).

Files with missing lines	Patch %	Lines
pkg/argocd/argocd.go	84.21%	2 Missing and 1 partial ⚠️
cmd/run.go	50.00%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #1112      +/-   ##
==========================================
+ Coverage   63.49%   63.60%   +0.10%     
==========================================
  Files          15       15              
  Lines        2326     2341      +15     
==========================================
+ Hits         1477     1489      +12     
- Misses        758      760       +2     
- Partials       91       92       +1     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[grahamalama]

I was unsure how to add tests for this change. I saw these tests, but I'm not sure how to set up something similar to assert that "if the client is configured with an appNamespace, it will only search for that namespace". Maybe asserting the fake client was called with certain parameters? Otherwise I think we'd end up testing the mock.

[chengfang]

@grahamalama thanks for the contribution. Can you please fix the DCO check failure? Is this PR also intended to address issue #1111 ?

[grahamalama]

@chengfang just pushed the DCO fix.

Is this PR also intended to address issue #1111?

It could be I suppose! For me, my goal was to fix #902 specifically. But I think it'd addresses both.

[chengfang]

I was unsure how to add tests for this change. I saw these tests, but I'm not sure how to set up something similar to assert that "if the client is configured with an appNamespace, it will only search for that namespace". Maybe asserting the fake client was called with certain parameters? Otherwise I think we'd end up testing the mock.

can we add tests with non-nil appNamespace, just for some basic sanity check?

[chengfang]

@grahamalama can you also update docs/install/reference.md to document the new flag? (https://argocd-image-updater.readthedocs.io/en/latest/install/reference/

[chengfang]

@grahamalama can you also take a look at the comment on the related issue #1111 (comment). Although it covers a different issue (still related to this one), I think the comment about using listApp vs getApp is also revelant here. Some namespace may contain large number of apps and list operation in this namespace could still be a performance bottleneck. If we know image-updater is configured to 1 namespace, we should be able to just use getApp.

[grahamalama]

@grahamalama can you also take a look at the comment on the related issue #1111 (comment). Although it covers a different issue (still related to this one), I think the comment about using listApp vs getApp is also revelant here. Some namespace may contain large number of apps and list operation in this namespace could still be a performance bottleneck. If we know image-updater is configured to 1 namespace, we should be able to just use getApp.

Commented in #1111, and depending on the outcome of that conversation I'll update this PR accordingly. I will say though if the decision is to add some sort of caching option, I think that would be better suited for a separate PR.

[chengfang]

@grahamalama from the discussion here and in #1111, we are in agreement to improve this PR by using getApp instead of listApp when single namespace is configured in image-updater.

[grahamalama]

@chengfang, made the changes we discussed in #1111 and added the requested documentation