-
Notifications
You must be signed in to change notification settings - Fork 292
ci-operator: add automatic dockerfile inputs detection #4851
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
base: main
Are you sure you want to change the base?
Conversation
|
Pipeline controller notification For optional jobs, comment This repository is configured in: LGTM mode |
WalkthroughAdds Dockerfile-based base-image auto-detection: new pkg/dockerfile parsing and detection, integration into pkg/defaults build-step generation to append detected InputImageTag steps, unit tests for parsing/processing, and end-to-end tests with Dockerfile fixtures and a test config. Changes
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~20–30 minutes
✨ Finishing touches
🧪 Generate unit tests (beta)
Comment |
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: Prucek The full list of commands accepted by this bot can be found here. The pull request process is described here DetailsNeeds approval from an approver in each of these files:
Approvers can indicate their approval by writing |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Actionable comments posted: 0
🧹 Nitpick comments (1)
pkg/defaults/defaults_test.go (1)
2445-2457: Prefer standard librarystrings.Contains.These helper functions reimplement functionality available in the standard library. Using
strings.Containsdirectly would simplify the code and improve maintainability.Apply this diff:
-func containsString(s, substr string) bool { - return len(s) >= len(substr) && (s == substr || len(substr) == 0 || - (len(s) > 0 && len(substr) > 0 && contains(s, substr))) -} - -func contains(s, substr string) bool { - for i := 0; i <= len(s)-len(substr); i++ { - if s[i:i+len(substr)] == substr { - return true - } - } - return false -}Then update line 2237 to use
strings.Containsdirectly:+import "strings" + if tc.expectedErrContain != "" && err != nil { - if !containsString(err.Error(), tc.expectedErrContain) { + if !strings.Contains(err.Error(), tc.expectedErrContain) { t.Errorf("expected error to contain %q, got: %v", tc.expectedErrContain, err) } }
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro
Cache: Disabled due to data retention organization setting
Knowledge base: Disabled due to Reviews -> Disable Knowledge Base setting
📒 Files selected for processing (12)
pkg/defaults/defaults.go(2 hunks)pkg/defaults/defaults_test.go(3 hunks)pkg/dockerfile/inputs.go(1 hunks)pkg/dockerfile/inputs_test.go(1 hunks)test/e2e/dockerfile-inputs/config.yaml(1 hunks)test/e2e/dockerfile-inputs/e2e_test.go(1 hunks)test/e2e/dockerfile-inputs/testdata/Dockerfile.copy-from(1 hunks)test/e2e/dockerfile-inputs/testdata/Dockerfile.manual(1 hunks)test/e2e/dockerfile-inputs/testdata/Dockerfile.multiple(1 hunks)test/e2e/dockerfile-inputs/testdata/Dockerfile.no-refs(1 hunks)test/e2e/dockerfile-inputs/testdata/Dockerfile.quay-proxy(1 hunks)test/e2e/dockerfile-inputs/testdata/Dockerfile.single(1 hunks)
🧰 Additional context used
📓 Path-based instructions (1)
**
⚙️ CodeRabbit configuration file
-Focus on major issues impacting performance, readability, maintainability and security. Avoid nitpicks and avoid verbosity.
Files:
test/e2e/dockerfile-inputs/config.yamltest/e2e/dockerfile-inputs/testdata/Dockerfile.quay-proxytest/e2e/dockerfile-inputs/testdata/Dockerfile.manualtest/e2e/dockerfile-inputs/testdata/Dockerfile.multipletest/e2e/dockerfile-inputs/testdata/Dockerfile.copy-fromtest/e2e/dockerfile-inputs/testdata/Dockerfile.no-refspkg/dockerfile/inputs_test.gopkg/dockerfile/inputs.gopkg/defaults/defaults_test.gopkg/defaults/defaults.gotest/e2e/dockerfile-inputs/e2e_test.gotest/e2e/dockerfile-inputs/testdata/Dockerfile.single
🧬 Code graph analysis (4)
pkg/dockerfile/inputs_test.go (2)
pkg/api/types.go (2)
ImageBuildInputs(2748-2758)ImageStreamTagReference(494-501)pkg/dockerfile/inputs.go (1)
DetectInputsFromDockerfile(33-90)
pkg/dockerfile/inputs.go (1)
pkg/api/types.go (2)
ImageBuildInputs(2748-2758)ImageStreamTagReference(494-501)
pkg/defaults/defaults.go (3)
pkg/api/types.go (9)
ReleaseBuildConfiguration(37-135)StepConfiguration(618-632)ProjectDirectoryImageBuildStepConfiguration(2661-2684)ImageBuildInputs(2748-2758)ImageStreamTagReference(494-501)InputImageTagStepConfiguration(638-641)InputImage(680-687)PipelineImageStreamTagReference(2493-2493)ImageStreamSource(699-702)pkg/dockerfile/inputs.go (1)
DetectInputsFromDockerfile(33-90)cmd/repo-init/api.go (1)
BaseImages(69-69)
test/e2e/dockerfile-inputs/e2e_test.go (1)
pkg/testhelper/accessory.go (1)
T(51-57)
🔇 Additional comments (15)
test/e2e/dockerfile-inputs/testdata/Dockerfile.no-refs (1)
1-2: LGTM!Appropriate test fixture for verifying the "no registry references" detection path.
test/e2e/dockerfile-inputs/testdata/Dockerfile.manual (1)
1-2: LGTM!Appropriate fixture for testing manual inputs precedence over auto-detection.
test/e2e/dockerfile-inputs/testdata/Dockerfile.quay-proxy (1)
1-2: LGTM!Appropriate test fixture for verifying quay-proxy reference detection and parsing.
test/e2e/dockerfile-inputs/testdata/Dockerfile.single (1)
1-2: LGTM!Appropriate fixture for the basic single-reference auto-detection scenario.
test/e2e/dockerfile-inputs/testdata/Dockerfile.multiple (1)
1-6: LGTM!Appropriate multi-stage fixture for testing detection of multiple base image references.
test/e2e/dockerfile-inputs/testdata/Dockerfile.copy-from (1)
1-6: LGTM!Appropriate fixture for testing COPY --from directive handling in multi-stage builds.
pkg/defaults/defaults_test.go (2)
7-7: LGTM!The new imports are appropriately used in the test functions.
Also applies to: 20-20
2138-2443: Comprehensive test coverage.The test functions provide thorough coverage of Dockerfile reading and base image detection scenarios, including edge cases for custom paths, literals, manual inputs, and multi-stage builds.
test/e2e/dockerfile-inputs/config.yaml (1)
1-30: LGTM!Well-structured e2e test configuration that covers the key scenarios: single/multiple auto-detection, quay-proxy references, manual inputs precedence, COPY --from handling, and no-registry-references cases.
pkg/defaults/defaults.go (1)
1029-1137: LGTM: Clean integration of Dockerfile input detection.The implementation follows good practices:
- Soft-fail approach (logs and continues) for individual image failures is appropriate, as one malformed Dockerfile shouldn't block the entire build
- Functions are well-factored and focused on single responsibilities
- Error handling is consistent with the rest of the codebase
- Mutations to
config.BaseImagesandimage.Inputsfollow existing patterns in this filetest/e2e/dockerfile-inputs/e2e_test.go (1)
12-90: LGTM: Comprehensive e2e test coverage.The test cases cover the key scenarios:
- Auto-detection of single and multiple registry references
- Quay-proxy reference handling
- Manual input override behavior
- COPY --from references
- No registry references case
Test structure is clean and maintainable.
pkg/dockerfile/inputs_test.go (1)
11-293: LGTM: Thorough unit test coverage.The test suite covers:
- Edge cases (empty dockerfile, duplicates, mixed registry/non-registry refs)
- Multiple registry formats (registry.ci, registry.svc.ci, quay-proxy)
- Error conditions (malformed references)
- Manual input override behavior
Good use of table-driven tests and clear assertions.
pkg/dockerfile/inputs.go (3)
18-19: Regex pattern is permissive but acceptable.The pattern
\S+matches any non-whitespace characters after the registry domain, which is quite permissive. However, this is acceptable because:
orgRepoTagFromPullString()validates the format downstream (Lines 138-162)- Invalid formats are caught and returned as errors
- The simple approach aligns with the stated goal of parity with registry-replacer
33-90: LGTM: Well-structured detection logic.The function correctly:
- Returns early for empty input or when manual inputs are defined
- Deduplicates registry references
- Handles parsing errors appropriately
- Logs useful diagnostic information at appropriate levels
The approach of honoring manual
inputs.as[]configuration while auto-detecting when not present is the right design choice.
181-220: Incorrect review comment. The functionextractReplacementCandidatesFromDockerfileatpkg/dockerfile/inputs.go:183IS actively called in the codebase atcmd/registry-replacer/main.go:305. It is not dead code or unused. The function serves its intended purpose within the codebase.Likely an incorrect or invalid review comment.
|
/test e2e |
|
|
||
| // readDockerfileForImage reads the Dockerfile content for a given image configuration | ||
| // Returns content, a description of the source, and any error | ||
| func readDockerfileForImage(image *api.ProjectDirectoryImageBuildStepConfiguration, readFile readFile) ([]byte, string, error) { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The nesting levels of this body can be decreased by doing so:
func readDockerfileForImage(image *api.ProjectDirectoryImageBuildStepConfiguration, readFile readFile) ([]byte, string, error) {
if image.DockerfileLiteral != nil {
return []byte(*image.DockerfileLiteral), "dockerfile_literal"
}
// Everything else here, no `else`
}|
/test images |
fdaf457 to
f2792f4
Compare
|
/test e2e |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Actionable comments posted: 2
🧹 Nitpick comments (1)
pkg/dockerfile/inputs.go (1)
93-101: Clarify or remove currently-unused per-reference manual override helper
hasManualInputsshort-circuits auto-detection when anyinputs.asentry is present, which matches the intended “manual config wins” semantics. Given that, the per-reference helperhasManualReplacementForis currently unreachable in production: with the existing call pattern it can only run in cases where there are no manualinputs.asentries.That’s not harmful but slightly confusing for future readers. Two options:
- If you plan to support a “mixed” mode (some refs auto-detected, some overridden manually), add a comment on
DetectInputsFromDockerfileexplaining the intended future use ofhasManualReplacementFor, or adjust the short-circuit logic accordingly.- If not, consider dropping
hasManualReplacementForfor now and reintroducing it when such a mode is actually needed.This keeps the detection path easier to reason about and avoids dead-code-like helpers.
Also applies to: 129-137
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro
Cache: Disabled due to data retention organization setting
Knowledge base: Disabled due to Reviews -> Disable Knowledge Base setting
📒 Files selected for processing (12)
pkg/defaults/defaults.go(2 hunks)pkg/defaults/defaults_test.go(3 hunks)pkg/dockerfile/inputs.go(1 hunks)pkg/dockerfile/inputs_test.go(1 hunks)test/e2e/dockerfile-inputs/config.yaml(1 hunks)test/e2e/dockerfile-inputs/e2e_test.go(1 hunks)test/e2e/dockerfile-inputs/testdata/Dockerfile.copy-from(1 hunks)test/e2e/dockerfile-inputs/testdata/Dockerfile.manual(1 hunks)test/e2e/dockerfile-inputs/testdata/Dockerfile.multiple(1 hunks)test/e2e/dockerfile-inputs/testdata/Dockerfile.no-refs(1 hunks)test/e2e/dockerfile-inputs/testdata/Dockerfile.quay-proxy(1 hunks)test/e2e/dockerfile-inputs/testdata/Dockerfile.single(1 hunks)
🚧 Files skipped from review as they are similar to previous changes (7)
- test/e2e/dockerfile-inputs/testdata/Dockerfile.quay-proxy
- test/e2e/dockerfile-inputs/testdata/Dockerfile.manual
- test/e2e/dockerfile-inputs/testdata/Dockerfile.single
- pkg/defaults/defaults_test.go
- test/e2e/dockerfile-inputs/testdata/Dockerfile.copy-from
- test/e2e/dockerfile-inputs/testdata/Dockerfile.multiple
- test/e2e/dockerfile-inputs/config.yaml
🧰 Additional context used
📓 Path-based instructions (1)
**
⚙️ CodeRabbit configuration file
-Focus on major issues impacting performance, readability, maintainability and security. Avoid nitpicks and avoid verbosity.
Files:
test/e2e/dockerfile-inputs/testdata/Dockerfile.no-refstest/e2e/dockerfile-inputs/e2e_test.gopkg/defaults/defaults.gopkg/dockerfile/inputs.gopkg/dockerfile/inputs_test.go
🧬 Code graph analysis (2)
test/e2e/dockerfile-inputs/e2e_test.go (1)
pkg/testhelper/accessory.go (1)
T(51-57)
pkg/dockerfile/inputs.go (1)
pkg/api/types.go (2)
ImageBuildInputs(2748-2758)ImageStreamTagReference(494-501)
🔇 Additional comments (3)
test/e2e/dockerfile-inputs/testdata/Dockerfile.no-refs (1)
1-2: No-registry fixture correctly exercises negative pathThis Dockerfile intentionally contains no
registry.ci.openshift.org/quay-proxyrefs and is appropriate for the "no auto-detect" scenario.test/e2e/dockerfile-inputs/e2e_test.go (1)
1-90: E2E coverage for Dockerfile inputs is comprehensive and wired correctlyThe table-driven
TestDockerfileInputsexercises all key scenarios (single/multiple/quay-proxy/manual/COPY-from/no-refs) via the standard e2e framework, passing pull secrets andJOB_SPECappropriately. Assertions on success and log output look sound.pkg/dockerfile/inputs_test.go (1)
1-293: Unit tests thoroughly exercise detection and parsing helpers
TestDetectInputsFromDockerfile,TestOrgRepoTagFromPullString, andTestExtractReplacementCandidatesFromDockerfilecollectively cover the important happy-path and edge cases (multiple registries, quay-proxy, legacyregistry.svc, manual inputs, mixed/duplicate refs, and COPY/AS stages). This gives good confidence in the new parsing logic.
| // Detect Dockerfile inputs for project images and add InputImageTagStepConfiguration entries | ||
| dockerfileInputSteps, err := detectDockerfileInputs(config, readFile) | ||
| if err != nil { | ||
| return nil, fmt.Errorf("failed to detect Dockerfile inputs: %w", err) | ||
| } | ||
| buildSteps = append(buildSteps, dockerfileInputSteps...) | ||
|
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Make Dockerfile-input detection best-effort instead of failing the build graph
Right now any error from detectDockerfileInputs (e.g., an unexpected registry reference that DetectInputsFromDockerfile cannot parse) will cause runtimeStepConfigsForBuild to return an error and abort graph construction. That’s a behavior change compared to today, where unusual but syntactically valid Dockerfiles still run, just without auto-detected inputs.
Given this feature is an additive convenience, it’s safer to degrade gracefully on detection errors and continue without auto-generated InputImageTagStepConfigurations. Otherwise, a single odd-looking reference to registry.ci.openshift.org/quay-proxy.ci.openshift.org can break the entire job.
Consider handling errors like this instead:
- // Detect Dockerfile inputs for project images and add InputImageTagStepConfiguration entries
- dockerfileInputSteps, err := detectDockerfileInputs(config, readFile)
- if err != nil {
- return nil, fmt.Errorf("failed to detect Dockerfile inputs: %w", err)
- }
- buildSteps = append(buildSteps, dockerfileInputSteps...)
+ // Detect Dockerfile inputs for project images and add InputImageTagStepConfiguration entries.
+ // This is best-effort and should not break existing configurations.
+ dockerfileInputSteps, err := detectDockerfileInputs(config, readFile)
+ if err != nil {
+ logrus.WithError(err).Warn("Dockerfile inputs detection failed; continuing without auto-detected base images")
+ } else {
+ buildSteps = append(buildSteps, dockerfileInputSteps...)
+ }This keeps the new functionality while avoiding regressions for configs with unusual but valid image references.
Also applies to: 1042-1131
🤖 Prompt for AI Agents
In pkg/defaults/defaults.go around lines 1029 to 1035 (and similarly for
1042–1131), the call to detectDockerfileInputs currently returns an error that
aborts graph construction; make detection best‑effort by catching and logging
the error instead of returning it. Change the code to call
detectDockerfileInputs, and if it returns an error, write a warning-level log or
comment (including the error message) and continue without appending
dockerfileInputSteps; only return on fatal/unexpected errors, not on
parse/registry detection failures. Ensure the same non-fatal behavior is applied
to the other block(s) noted (1042–1131).
| func DetectInputsFromDockerfile( | ||
| dockerfile []byte, | ||
| existingInputs map[string]api.ImageBuildInputs, | ||
| ) (map[string]api.ImageStreamTagReference, error) { | ||
| if len(dockerfile) == 0 { | ||
| return nil, nil | ||
| } |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Avoid failing on unparseable registry refs; treat them as best-effort skips
DetectInputsFromDockerfile currently returns an error if orgRepoTagFromPullString cannot parse a matched registry reference (e.g., if the path doesn’t look like exactly /<org>/<repo>:<tag>). Combined with the caller in runtimeStepConfigsForBuild, this means:
- Any Dockerfile line that happens to match
registryRegexbut uses a slightly different yet syntactically valid shape (extra path segment, missing explicit tag, future quay-proxy variant, etc.) can cause the entire job to fail, even though auto-detection is an additive feature. - The repo already has explicit unit tests for error cases in
orgRepoTagFromPullString; those shouldn’t necessarily translate into hard failures for ci-operator.
To keep the feature robust while still surfacing problems, I’d recommend treating parse failures as “skip this ref with a warning” instead of aborting:
for _, ref := range registryRefs {
- // Skip if there's already a manual input for this reference
+ // Skip if there's already a manual input for this reference
if hasManualReplacementFor(existingInputs, ref) {
logrus.WithField("reference", ref).Debug("Skipping auto-detection: manual replacement exists")
continue
}
- // Parse the registry reference into org/repo/tag
- orgRepoTag, err := orgRepoTagFromPullString(ref)
- if err != nil {
- return nil, fmt.Errorf("failed to parse registry reference %s: %w", ref, err)
- }
+ // Parse the registry reference into org/repo/tag. If the shape is unexpected,
+ // log and skip instead of failing the whole job.
+ orgRepoTag, err := orgRepoTagFromPullString(ref)
+ if err != nil {
+ logrus.WithError(err).Warnf("Dockerfile-inputs: skipping unrecognized registry reference %q", ref)
+ continue
+ }With this change (and making the caller in runtimeStepConfigsForBuild best-effort as suggested earlier), detection cannot regress existing jobs simply because of a new or slightly different registry path pattern, while still emitting useful diagnostics when something looks off.
Also applies to: 56-67, 139-163
🤖 Prompt for AI Agents
In pkg/dockerfile/inputs.go around lines 34-40 (and similarly at 56-67 and
139-163), the function currently treats failures from orgRepoTagFromPullString
as fatal errors; change that behavior to be best-effort: when
orgRepoTagFromPullString returns an error, emit a warning (or debug log)
describing the unparseable ref and continue without adding that ref to the
result map instead of returning the error, so DetectInputsFromDockerfile skips
problematic registry refs but keeps processing the rest of the Dockerfile.
|
/test images |
f2792f4 to
eede1fe
Compare
|
/test e2e |
eede1fe to
b1d622c
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Actionable comments posted: 1
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro
Cache: Disabled due to data retention organization setting
Knowledge base: Disabled due to Reviews -> Disable Knowledge Base setting
📒 Files selected for processing (8)
test/e2e/dockerfile-inputs/config.yaml(1 hunks)test/e2e/dockerfile-inputs/e2e_test.go(1 hunks)test/e2e/dockerfile-inputs/testdata/Dockerfile.copy-from(1 hunks)test/e2e/dockerfile-inputs/testdata/Dockerfile.manual(1 hunks)test/e2e/dockerfile-inputs/testdata/Dockerfile.multiple(1 hunks)test/e2e/dockerfile-inputs/testdata/Dockerfile.no-refs(1 hunks)test/e2e/dockerfile-inputs/testdata/Dockerfile.quay-proxy(1 hunks)test/e2e/dockerfile-inputs/testdata/Dockerfile.single(1 hunks)
✅ Files skipped from review due to trivial changes (1)
- test/e2e/dockerfile-inputs/testdata/Dockerfile.manual
🚧 Files skipped from review as they are similar to previous changes (5)
- test/e2e/dockerfile-inputs/testdata/Dockerfile.quay-proxy
- test/e2e/dockerfile-inputs/config.yaml
- test/e2e/dockerfile-inputs/testdata/Dockerfile.single
- test/e2e/dockerfile-inputs/testdata/Dockerfile.multiple
- test/e2e/dockerfile-inputs/testdata/Dockerfile.copy-from
🧰 Additional context used
📓 Path-based instructions (1)
**
⚙️ CodeRabbit configuration file
-Focus on major issues impacting performance, readability, maintainability and security. Avoid nitpicks and avoid verbosity.
Files:
test/e2e/dockerfile-inputs/testdata/Dockerfile.no-refstest/e2e/dockerfile-inputs/e2e_test.go
🧬 Code graph analysis (1)
test/e2e/dockerfile-inputs/e2e_test.go (1)
pkg/testhelper/accessory.go (1)
T(51-57)
🔇 Additional comments (3)
test/e2e/dockerfile-inputs/testdata/Dockerfile.no-refs (1)
1-2: LGTM!Valid test fixture for the no-registry-references case. The Dockerfile correctly uses a public base image and explicitly documents the test intent.
test/e2e/dockerfile-inputs/e2e_test.go (2)
1-13: LGTM!Standard e2e test structure with appropriate build tags and minimal imports. The defaultJobSpec is lengthy but acceptable as test fixture data.
15-73: LGTM!Comprehensive test coverage with clear table-driven structure. The test cases appropriately cover various scenarios including single/multiple references, quay-proxy, manual overrides, COPY --from, and the no-references case.
| if testCase.success != (err == nil) { | ||
| t.Fatalf("%s: didn't expect an error from ci-operator: %v; output:\n%v", testCase.name, err, string(output)) | ||
| } |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Clarify the error message for both failure modes.
The error message "didn't expect an error from ci-operator" is misleading when testCase.success is false but the command succeeds. The condition catches both cases (unexpected success and unexpected failure), but the message only describes one.
Consider this diff to make the message clearer:
-if testCase.success != (err == nil) {
- t.Fatalf("%s: didn't expect an error from ci-operator: %v; output:\n%v", testCase.name, err, string(output))
-}
+if testCase.success && err != nil {
+ t.Fatalf("%s: expected success but got error: %v; output:\n%v", testCase.name, err, string(output))
+} else if !testCase.success && err == nil {
+ t.Fatalf("%s: expected failure but command succeeded; output:\n%v", testCase.name, string(output))
+}📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| if testCase.success != (err == nil) { | |
| t.Fatalf("%s: didn't expect an error from ci-operator: %v; output:\n%v", testCase.name, err, string(output)) | |
| } | |
| if testCase.success && err != nil { | |
| t.Fatalf("%s: expected success but got error: %v; output:\n%v", testCase.name, err, string(output)) | |
| } else if !testCase.success && err == nil { | |
| t.Fatalf("%s: expected failure but command succeeded; output:\n%v", testCase.name, string(output)) | |
| } |
🤖 Prompt for AI Agents
In test/e2e/dockerfile-inputs/e2e_test.go around lines 82-84, the t.Fatalf
message only describes an unexpected error but the condition covers both
unexpected success and unexpected failure; update the logic to distinguish the
two cases and produce clear messages: when testCase.success is true and err !=
nil log "expected success but got error" with err and output, and when
testCase.success is false and err == nil log "expected failure but command
succeeded" with output; keep t.Fatalf usage and include both err and output
where relevant for debugging.
|
/test e2e |
|
@Prucek: The following tests failed, say
Full PR test history. Your PR dashboard. DetailsInstructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository. I understand the commands that are listed here. |
Added
pkg/dockerfile/inputs.go- Core detection logic:registry.ci.openshift.organdquay-proxy.ci.openshift.orgreferences in Dockerfilesorg/repo/tagcomponentsbase_imagesandinputs.as[]configuration on-the-flyIntegrated into
pkg/defaults/defaults.go:detectDockerfileInputs()function reads Dockerfiles from source checkoutDetectInputsFromDockerfile()during build graph constructioninputs.as[]takes precedence)