feat(plating): hide quality smart buttons at zero + order-scope NCR/CAPA counts

On the sale.order and fp.job forms, the Holds/Checks/NCRs/CAPAs/RMAs quality
smart buttons (and the SO WO button) now hide when their count is 0, so the row
shows only quality work that exists. NCR and CAPA counts are re-scoped from
customer-wide to order/job via a shared _fp_quality_ncr_ids() helper (NCRs
reached through the order's RMAs + the order/job's holds), so each badge and the
list its button opens always agree. Also aligned the job RMA button's list
domain to its (already SO-scoped) count.

Reverts the Sub-12 Phase D "always-visible (zero is OK)" choice back to the
module's documented hide-at-zero convention.

- fusion_plating_quality 19.0.8.1.0 -> 19.0.8.2.0
- fusion_plating_jobs    19.0.12.4.0 -> 19.0.12.5.0

Deployed + verified on entech (badge == helper across sampled SOs/jobs).

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
gsinghpal
2026-06-04 20:36:54 -04:00
parent 423f288507
commit c9eb61ee0c
8 changed files with 223 additions and 67 deletions

View File

@@ -0,0 +1,107 @@
# Sale Order + Job Quality Smart Buttons — hide-at-zero + order-scoped counts
**Date:** 2026-06-04
**Modules:** `fusion_plating_quality`, `fusion_plating_jobs`
**Status:** approved (brainstorming sign-off)
## Problem
The `sale.order` form shows eight Fusion Plating smart buttons: Receiving, WO,
Certificates, Holds, Checks, NCRs, CAPAs, RMAs. On a typical order only three
carry data (Receiving / WO / Certificates) yet six of the eight render
permanently — five quality buttons (Holds / Checks / NCRs / CAPAs / RMAs) and
the WO button all show even at a count of zero. The row reads as noise.
Two of those counts are also **wrong on an order**:
- **NCRs** counts the *entire customer's* NCR history (`customer_partner_id ==
partner`), not this order's. A repeat customer's brand-new order shows "NCRs 5".
- **CAPAs** counts the CAPAs under those customer-wide NCRs — same over-count.
This contradicts the module's own documented smart-button convention ("don't add
a button that always shows zero… NCRs/RMAs visible only when present"). The
Sub-12 Phase D plan that introduced these buttons chose "always visible (zero is
OK)"; this spec reverts to the convention.
The identical button row + identical over-count exists on the `fp.job` form and
is fixed in the same change.
## Decision
1. **Hide every quality button when its count is 0**, on both the `sale.order`
and `fp.job` forms. Also hide the **WO** button at 0. (Receiving and
Certificates already hide at 0 — untouched.)
2. **Re-scope the NCR and CAPA counts to the order/job**, dropping the
customer-wide union. Use only the real order-linked paths.
### Order-scoped NCR set
There is no direct `ncr.sale_order_id` or `ncr.job_id` field. The only honest
paths from an order to its NCRs (confirmed by model inspection) are:
- **RMA path:** SO → its RMAs (`rma.sale_order_id`) → NCRs (`ncr.rma_id`).
- **Hold path:** SO → its jobs (`fp.job.sale_order_id`) → holds (`hold.job_id`)
→ the NCR a hold spawned (`hold.ncr_id`).
NCR set = union of those two. CAPA set = CAPAs whose `ncr_id` is in that set.
For the `fp.job` form the same definition applies, narrowed to the one job: the
job's SO's RMAs' NCRs, plus *this job's* holds' NCRs.
Trade-off accepted: an NCR raised with neither an RMA nor a hold link (e.g. a QA
team logs one directly against the customer) no longer appears on any order or
job. That is correct — such an NCR has no order/job linkage to honestly attach
it to, and surfacing it on every one of the customer's orders was the bug.
## Changes
### `fusion_plating_quality`
- `models/fp_quality_smart_buttons.py`
- Add `sale.order._fp_quality_ncr_ids()` and `fp.job._fp_quality_ncr_ids()`
helpers returning the order/job-scoped NCR id list (single source of truth).
- Rewrite both `_compute_*` methods to set the NCR count from the helper and
the CAPA count from `[('ncr_id', 'in', <helper>)]`. Holds / Checks / RMAs
counts are already correctly scoped — leave them.
- Point `action_view_fp_ncrs`, `action_view_fp_ncrs_so`, and both
`action_view_fp_capas` at the helper-derived ids so each opened list matches
its badge.
- Narrow `fp.job.action_view_fp_rmas` from SO-OR-partner to SO-only so the RMA
list matches its (already SO-scoped) badge count (caught in review — the SO
form's RMA button already agreed).
- `views/fp_quality_smart_button_views.xml` — add `invisible="<count> == 0"` to
the five SO quality buttons.
- `__manifest__.py` — version `19.0.8.1.0` → `19.0.8.2.0`.
### `fusion_plating_jobs`
- `views/fp_job_quality_buttons.xml` — add `invisible="<count> == 0"` to the
five job-form quality buttons.
- `views/sale_order_views.xml` — add `invisible="x_fc_fp_job_count == 0"` to the
WO smart button; update the now-stale "always visible" comment.
- `__manifest__.py` — version `19.0.12.4.0` → `19.0.12.5.0`.
### Docs
- `fusion_plating/CLAUDE.md` — short note in the smart-button section recording
that the SO/job quality buttons are hide-at-zero with order-scoped NCR/CAPA
counts as of 2026-06-04 (reconciling the Sub-12 "always visible" record).
## Out of scope
- The `res.partner` "Quality History" button stays as-is. It is customer-wide
*by design*, and the user opted not to touch it.
- No schema changes, no migration — pure view attributes + non-stored compute
logic over existing fields.
## Verification
The plating modules cannot be installed on the local Community dev DB
(enterprise deps). Verification is:
1. Static: `pyflakes` on the edited `.py`; `lxml` parse on each edited `.xml`.
2. Behavioural (read-only, on an entech clone/shell): for a sample SO and job,
confirm the new helper returns only RMA-/hold-linked NCRs and that the counts
match, with `env.cr.rollback()` at the end.
3. Deploy to entech via the revert-on-failure guard only after the above pass
and with the owner's go-ahead (live client box).