Track ocl-content source loads: ICD-10-CM FY27, WHO-ATC and SNOMED GPS uploads blocked by open decisions
@filiperochalopes 已经在做这个了。
开始于 2026年10月14日。
评估
这个 Issue 还没有评估数据。
描述
User story
As the CIEL content owner, I want every source upload, creation and update tracked in one place, so each load has an owner, its open decisions and an audit gate before anything is written to OCL.
Use case
The ocl-content pipelines build OCL bulk-import bundles for several terminologies. The current drafts for three sources are not ready to load: each audit file has upload_ready: false and the reason is recorded in the source's own open decisions or in a required fresh rehearsal.
Workstreams
| Source | Load mode | Current state | Blocked by |
|---|---|---|---|
| US-NCHS / ICD-10-CM FY27 | full-source refresh | Draft bundle, upload_ready: false |
Six open decisions in TBD.md: instruction-note model (EXCLUDES1/2, includes, seventh-character notes); hierarchy rule (direct XML containment vs derived ranges); legacy UMLS enrichment (CUI, semantic types) keep or remove; non-official T### semantic-type concepts and B10-B10-style range artifacts (retirement); FY27 release identifier and artifact names |
| WHO / WHOATC | full-source load and retirement update | Draft bundle and retirement audit, both upload_ready: false |
Authority decision (WHOCC ATC/DDD Index 2026 vs BioPortal); cut-off policy for 2027 decisions; UMLS licensing and provenance; fresh OCL snapshot and rehearsal before retirements |
| SNOMED International / SNOMED GPS | full-source refresh to the January 2026 scope | Draft bundle, upload_ready: false |
Fresh OCL snapshot and local rehearsal (the importer replaces collection fields) |
Not assessed in this issue: LOINC and the WHO ICD-10 pipeline. Add them when reviewed.
Open questions for the next meeting
- Which decision on ICD-10-CM instruction notes (structured property, typed description, or mixed model)? Owner: Joe.
- Is the hierarchy the direct XML tree or derived range containment? Owner: Joe.
- Are the UMLS CUI and semantic-type values on ICD-10-CM intentional published enrichment to keep, or legacy data to remove? Owner: Joe.
- Can the non-official
T###concepts and the range artifacts be retired? Owner: Joe. - Approve the WHOCC ATC/DDD Index 2026 as the authority for WHO-ATC, with BioPortal/UMLS for enrichment only? Owner: product team.
- Confirm the FY27 version identifier and the WHO-ATC cut-off policy for 2027. Owner: product team.
Acceptance criteria
- Each workstream has an owner and a status: decided, blocked, or on hold.
- Each open decision above is answered, or explicitly deferred with a reason recorded here.
- Before any manual import: a fresh OCL snapshot is taken, a local rehearsal passes, and the audit is reviewed.
-
upload_readyis set to true only after the review above. - No automated write to production. Each load is a manual step with its audit attached.
- Source versions are created only after the release identifier is confirmed.
Technical constraints
- Production writes are manual; the pipelines do not upload.
- OCL replaces supplied
extrasand collection fields wholesale, so a partial refresh can delete data. - The WHO-ATC retirement audit used a paginated listing because the HEAD export endpoint returned HTTP 204. Candidate discovery must be re-run before any import.
- Pipeline work is on the
ocl-contentrepository, branchfeature/snomed-gps-icd-10-cm-who-atc(not merged to main).
Related
- #2725 (Mapper vectorization after a source version release)
- #2555 (CIEL Lab production migration: formal source canonicals)
- 主要语言
- 没有语言数据
- 星标
- 4
- 派生
- 2
- 平均合并
- 2 小时 12 分钟
- 30 天内合并 PR
- 1
环境准备
- 没有 Dockerfile 或 Docker Compose 文件
- 没有 Pull Request 模板
- 阅读贡献指南
从这里开始
- 先读完整个 Issue，再读项目的贡献指南。
- 在 Issue 下留言说明你要接手 —— 这能避免两个人做同样的事。
- Fork 仓库，在一个分支上完成修改。
- 提交 Pull Request，并在描述里引用这个 Issue 编号。
OpenConceptLab/ocl_issues 的其他 Issue
-
component/fhir type/bug
难度 2/5 1-3 小时 新手友好度 82/100
OpenConceptLab/ocl_issues#3026 ·
-
component/api decision/human-review signal/needs-access signal/quick-win signal/small-scope signal/well-specified stage/triaged type/bug
难度 2/5 1-3 小时 新手友好度 72/100
OpenConceptLab/ocl_issues#3012 · 2 条评论 ·
-
component/design-system decision/ai-ready signal/low-risk signal/quick-win signal/small-scope signal/well-specified stage/triaged type/docs
难度 2/5 半天 新手友好度 70/100
OpenConceptLab/ocl_issues#2983 · 1 条评论 ·
-
component/mapper decision/ai-ready signal/low-risk signal/quick-win signal/small-scope signal/well-specified stage/triaged type/bug
难度 2/5 1-3 小时 新手友好度 78/100
OpenConceptLab/ocl_issues#2966 · 2 条评论 ·
-
component/mapper component/web decision/human-review signal/low-risk signal/needs-human signal/quick-win signal/small-scope signal/well-specified stage/triaged type/bug
难度 2/5 1-3 小时 新手友好度 76/100
OpenConceptLab/ocl_issues#2920 · 1 条评论 ·