fix: address review NITs — readability, test dedup, consistent SHA var
PR Ready Gate / clear-labels (pull_request) Successful in 2s
CI / test (pull_request) Successful in 30s
CI / review (anthropic--claude-4.6-sonnet, sonnet, SONNET_REVIEW_TOKEN) (pull_request) Successful in 22s
CI / review (gpt-5, security, ., rodin/security-patterns, SECURITY_REVIEW.md, SECURITY_REVIEW_TOKEN) (pull_request) Successful in 34s
CI / review (gpt-5, gpt, GPT_REVIEW_TOKEN) (pull_request) Successful in 1m6s

- vcs/types.go: restore blank line between CommitID and Comments fields
  for visual grouping (scalar vs slice fields)
- gitea/adapter_test.go: merge duplicate TestAdapter_PostReview_CommitID
  tests into one (Threading was a superset)
- cmd/review-bot/main.go: use evaluatedSHA instead of pr.Head.SHA for
  inline comment CommitID for consistency with review-level usage
This commit is contained in:
claw
2026-05-13 16:31:11 -07:00
parent be68e51898
commit 9a6298cc4f
3 changed files with 3 additions and 36 deletions
+1
View File
@@ -98,5 +98,6 @@ type ReviewRequest struct {
// If empty, the platform defaults to the current PR head.
// Adapters use this as the primary commit anchor for the review submission.
CommitID string `json:"commit_id,omitempty"`
Comments []ReviewComment `json:"comments,omitempty"`
}