Code Quality
AWS CodeGuru
AWS CodeGuru is a managed code-quality service combining ML-driven static analysis (CodeGuru Reviewer) with production profiling (CodeGuru Profiler). Reviewer scans Java and Python repositories against GitHub/AWS repositories to flag defects, security, and efficiency issues on pull requests. Profiler samples runtime workloads to identify hot code lines and recommends resource optimizations. Both planes emit findings through stable APIs and CI/CD integrations.
AWS
Service information
Shortname: CodeGuru
Huawei equivalent shortnames: CodeArts Check
Keywords: code quality, static analysis, recommendations
Differences vs Huawei
CodeArts Check is Huawei's static-analysis component inside the CodeArts suite, focused on rule-based code inspection, defect and security-finding generation, and pipeline gating within CodeArts Repos/CI. Unlike CodeGuru, which pairs a reviewer with a separate runtime profiler, Check has no first-party production profiling counterpart; profiling must be sourced separately (e.g., APM/GaussDB observability) if runtime-hotspot analysis is required. Reviewer also leans on ML-trained recommenders, whereas Check is predominantly ruleset-driven with configurable inspection items.
Integration surfaces differ materially. CodeGuru Reviewer binds natively to AWS CodeCommit/GitHub/GitHub Enterprise/Bitbucket and triggers on repository associations and pull requests, with findings surfaced in CodeGuru console and reviewer APIs. CodeArts Check binds to CodeArts Repos and the CodeArts Pipeline build gates, and its findings model is oriented to CodeArts project/tenant structures. There is no documented one-click import of CodeGuru repository associations; teams must re-create code repository bindings, inspection rulesets, and quality gates inside the CodeArts project model.
Coverage and language breadth are not equivalent. CodeGuru Reviewer supports Java and Python with a curated ML rule set plus security detectors. CodeArts Check supports a broader languages list (Java, C/C++, Go, JavaScript/TypeScript, Python, etc.) but its recommendations are ruleset-organized rather than ML-inferred, so finding taxonomy, severity, and remediation prose differ. Plan parity must be validated rule-by-rule because identical CWE/coding-standard coverage is not guaranteed and performance or false-positive characteristics are not published for either.
Migration to Huawei
Start with an assessment inventory: enumerate every CodeGuru Reviewer repository association, the languages scanned, the suppressed/severity-tuned rules, the pull-request gating policy, and any Profiler deployments over production workloads. Decide whether static analysis alone satisfies the target quality gate or whether a runtime profiling surrogate must be designed. There is no Huawei-side import tool for CodeGuru configs; mappings must be reconstructed manually against CodeArts Check rule categories, so treat the migration as a ruleset re-implementation, not a config port.
For the static-analysis plane, create the equivalent CodeArts project, attach the source from CodeArts Repos (or migrate the repository), and recreate inspection tasks in Check covering the same languages. Translate CodeGuru rule semantics to the closest Check inspection items, then apply severity and gating in the CodeArts Pipeline so pull-request builds fail on previously-blocking categories. For Profiler workloads, no direct equivalent exists; profile production JVM/Python runtime separately via observability tooling and ingest findings as manual remediation items, since no automated sync path is documented.
Validate before cutover with a parallel-run period: run CodeGuru and Check against the same commit range and diff findings by category, severity, file, and false-positive rate to calibrate rulesets. Confirm that no CodeGuru-suppressed findings regress unmasked in Check and that blocking gates still enforce the intended policy. Verify pull-request decorations and pipeline gate failure behavior in the CodeArts Pipeline mirror the prior CodeCommit/GitHub experience before retiring the CodeGuru associations.
Re-model cost and quotas. CodeGuru bills per code-reviewed lines for Reviewer and per profiling-hours for Profiler, whereas CodeArts Check is entaild within selected CodeArts edition tiers and pipeline/inspection usage. Confirm current Check pricing by edition, inspection limits, and regional availability, and recompute TCO including pipeline build-in-minutes. Note that Profiler has no Huawei tariff equivalent; its replacement cost must be added separately. Do not assume feature parity or one-click re-platforming; verify each gate and finding type against current Huawei documentation before production cutover.
Official Huawei Cloud documentation
Huawei Cloud
Huawei equivalent service
Shortname: CodeArts Check
General function: Code Quality
Code quality and static analysis service.
Keywords: code quality, static analysis, security