From f22402e1554826d6563536e560e987458ac5ef68 Mon Sep 17 00:00:00 2001 From: xcs Date: Sat, 29 Aug 2026 13:44:32 +0800 Subject: [PATCH] =?UTF-8?q?docs(spec):=20=E6=8B=86=E6=8E=89=20=C2=A74.3=20?= =?UTF-8?q?=E4=BC=AA=E7=A0=81=E9=87=8C=E7=9A=84=E5=8F=8D=E6=A8=A1=E5=BC=8F?= =?UTF-8?q?=20=E2=80=94=E2=80=94=20=E3=80=8C=3D=20=E6=9C=AC=E7=B1=BB?= =?UTF-8?q?=E5=9E=8B=E5=AD=97=E6=AE=B5=E6=95=B0=E3=80=8D=E4=B8=8D=E6=98=AF?= =?UTF-8?q?=E5=9F=BA=E5=BA=A7=E6=95=B0=E5=BE=97=E5=87=BA=E6=9D=A5=E7=9A=84?= =?UTF-8?q?=E6=95=B0?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Task 3 的评审指出来的:parts_type_param_rows 数的是本单草稿写下的字段行(人新建类型时加的那几个), 不是整本字典。拿它当分母就是「编一个数不出来的数」—— 本仓栽过六次。 Task 3 的实现者避开了,但下游照抄伪码就会把它带回来。行数由 prune + each 字典结构保证。 Co-Authored-By: Claude Opus 5 (1M context) Claude-Session: https://claude.ai/code/session_01LKMZDwMbyKzjw6Hembwvet --- .../specs/2026-08-29-act-work-05-rewrite-design.md | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/docs/superpowers/specs/2026-08-29-act-work-05-rewrite-design.md b/docs/superpowers/specs/2026-08-29-act-work-05-rewrite-design.md index 7eb0162..44f9a07 100644 --- a/docs/superpowers/specs/2026-08-29-act-work-05-rewrite-design.md +++ b/docs/superpowers/specs/2026-08-29-act-work-05-rewrite-design.md @@ -285,11 +285,17 @@ doneWhen sub_eval_part_exists_status_found = 2 ⑤ say -doneWhen difference_item_rows = 本类型字段数 +doneWhen substitution_difference_item_rows > 0 + and substitution_difference_item_missing = 0 and substitution_relation_rows = 1 and validation_record_rows = 1 ``` +> **为什么不写 `= 本类型字段数`**(这份 spec 早先的伪码里写过,是错的): +> 「本类型有几个字段」**不是基座数得出来的数** —— `parts_type_param_rows` 数的是**本单草稿写下的** +> 字段行(人新建类型时加的那几个),不是整本字典。拿它当分母就是「编一个数不出来的数」, +> 本仓在这上面栽过六次。行数由 `prune` + `each 字典` 结构保证;判据只查「写了」与「不缺」。 + **这一步一张卡都不发,人不出现。** 三处刻意的克制: - **AI 决定不了「有几行」。** 明细的行由类型字典 `each` 出来,AI 的出参只是拿 `param_name` 去