Repository navigation
/import: common spreadsheet cell shapes fail whole rows (dotted / compact dates, option value case, one error per row) #25493
Description
Ruled: 6377844879 · batch #9 item 3a · letter A · 2026-10-24T00:43:45Z (proposals 1, 2, 4; not 3)
Context
Trial import of a simulated messy supplier ledger (hotsrm#10, the kind of sheet a purchasing department keeps in Excel for years). These are the cell shapes that failed rows which a person reads without hesitation. Each failure rejects the whole row, including rows where the bad cell is an optional field.
| Cell | Field type | Result | Expected |
|---|---|---|---|
2011.07.01 |
date | invalid_date |
2011-07-01 (YYYY.MM.DD is the most common Chinese written date form) |
20090615 |
date | invalid_date |
2009-06-15 (8-digit compact date, common in ERP exports) |
2005/3/8 |
date | accepted ✓ | — |
CN |
select, option value cn |
invalid_option |
match option values case-insensitively (labels already are) |
中国大陆 / 物流 |
select (labels 中国 / 物流商) |
invalid_option |
fine to refuse, but there is no way to declare synonyms on an option or picklist, so every legacy sheet needs a hand clean-up |
And one reporting problem: when a row has several bad cells, only the first coercion error is reported (coerceRow collects all errors, the runner keeps errors[0]). The operator fixes one cell, re-runs, finds the next one. Two to three rounds per row in our trial.
Proposal
- Accept
YYYY.MM.DDandYYYYMMDDinparseDateCell(unambiguous, year first). - Case-insensitive match on option values.
Optional(not in letter A — a new permanent metadata key waits for a second app; ruling 6377844879)aliases: string[]on select options / picklist options, used by import coercion only.- Report all cell errors of a row (e.g.
errors: [{ field, code, message }]next to the existing first-error keys).
Temporary measure in hotsrm
The import guide lists a cleaning checklist (date format, option labels, ISO country labels) to run on the sheet before importing.
Activity
objectstack-fleet commentedon Oct 23, 2026
Triage: decision-shaped (filing-gate ②: a feature / contract-shape proposal from a real app). Lands in the import cell coercion in the server lane (
parseDateCell, option matching,coerceRow/ the runner'serrors[0]); option aliases would also touchpackages/spec. Markedneeds-user-decision; not dispatchable until ruled.Path: records · 把客户的存量数据搬进来 | records-forms.import-transform-matrix (P2) | step ⑥
Type: Feature
维护者速读
客户多年维护的 Excel 台账里,
2011.07.01、20090615这样的日期,CN这样大小写不同的选项值,都会让整行导入失败;而且一行有多个错只报第一个,要改一个跑一遍。YYYY.MM.DD和YYYYMMDD;选项值大小写不敏感匹配;一行的所有单元格错误都报出来。选项别名不做。aliases(同义词,只给导入用)。推荐 A:三项都是无歧义的放宽和更完整的报错,不新增元数据键;别名是新契约面,等第二个应用要时再议。问:A / B / C?
os-decision-facets
aliases;C 零新增。Governing text: the import coercion contract in
packages/spec(ImportRequestSchema) and the runner's row report.Prior rulings read: unresolved (not searched with check-prior-rulings.mjs).
Batch note: one of five import / automation proposals from hotsrm's trial import filed on 2026-10-23 (#25475, #25476, #25477, #25481, #25493); the earlier hotsrm platform evaluation is the decision card #25496. They can be ruled as one batch.
Session:
session_01AkmasAfXt3cu83xGzkuKhe(objectstack triage, run on the maintainer's request in the project thread)Generated by Claude Code
os-elon-musk commentedon Oct 24, 2026
Ruling: batch #9 item 3a · letter A · maintainer 「#9 同意」 2026-10-24T00:43:45Z
A =
parseDateCell接受YYYY.MM.DD与YYYYMMDD（年在前、无歧义，不放宽其它形式）；选项值大小写不敏感匹配（标签已是）；一行所有单元格错误一次报全（报告加errors: [{ field, code, message }]，原有第一错字段保留）。不做aliases（B），等第二个应用要。工作量 server S，零元数据键。执行席先复现「一行只报第一个错」的那一行代码再改（总监席未定位到）。标签
needs-user-decision→pm:queue，分级按协议变更口径由分诊定；裁不派。简报：https://claude.ai/artifact/XKBVwFLq2V5nBTNy3enkbz（卡 3）。Generated by Claude Code
objectstack-fleet commentedon Oct 24, 2026
Closed 2026-10-24T01:40Z. Superseded by #25582 — the same card ("common spreadsheet cell shapes fail whole /import rows"), re-filed as
objectstack-fleet[bot]with this body and these labels. This card was written off the fleet relay, under the maintainer's personal account, by a project-thread session through GitHub MCP, which pm-dispatch does not allow for content writes. The thread above stays here for reference; the queue continues on #25582.Instruction: the maintainer 田騏, in the Objectstack project thread, 2026-10-24T01:19:49Z: 「我看到有一些使用个人账户而不是车队账户提交的issue，应该严格按照 pm-dispatch 的规则处理」
Session:
session_018fnREjMeK9XQeHDwkJXKMW