admin/Odoo-Modules
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

chore(plating): de-dash shipped code + intake-neutral customer emails

Browse Source 
Replace em-dashes and en-dashes with hyphens across 789 shipped source
files (py/xml/js/scss) so the delivered module reads as human-written;
em-dashes had become a recognizable AI-generated tell. Internal .md dev
notes are excluded. The WO-sticker mojibake strippers keep their dash
search targets (now written — / –). No logic changes: comments
and display strings only; validated with py_compile + lxml parse.

Rewrite the 7 customer notification emails to be intake-neutral
(ship-in / drop-off / pickup) and repair-aware, and fix the Shipped
email documents line (packing slip vs bill of lading; certificate only
when issued). Subjects use a hyphen separator.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
gsinghpal
2026-06-05 00:16:19 -04:00
parent c9eb61ee0c
commit 8c76a16366
789 changed files with 4692 additions and 4692 deletions
Download Patch File Download Diff File Expand all files Collapse all files

62
fusion_plating/fusion_plating_jobs/wizards/fp_cert_issue_wizard.py
View File

@@ -15,7 +15,7 @@ readings before confirming. On confirm:
and the parsed readings are written as fp.thickness.reading rows.
- cert.action_issue() is called for each cert.
The wizard is a convenience layer — it does NOT replace the per-cert
The wizard is a convenience layer - it does NOT replace the per-cert
Issue button on the cert form, which stays as the fallback path.
"""
import base64
@@ -53,7 +53,7 @@ _FISCHER_READING_RE = re.compile(
)
# Capture every {\pict ... \wmetafile8 ...hex...} group in an RTF, in
# document order. The hex blob can be interspersed with whitespace
# (RTF wraps to 80 cols) — the consumer strips it.
# (RTF wraps to 80 cols) - the consumer strips it.
_RTF_PICT_WMF_RE = re.compile(
r'\{\\pict'
r'(?:\\[a-zA-Z]+-?\d*\s?)*?'
@@ -71,7 +71,7 @@ def _fp_extract_rtf_images(raw_bytes):
XDAL 600 RTF exports embed each picture as a WMF metafile wrapping
the actual raster. ImageMagick on Debian Bookworm doesn't carry a
WMF delegate, so we shell out to `wmf2svg` (from libwmf-bin) — it
WMF delegate, so we shell out to `wmf2svg` (from libwmf-bin) - it
writes a thin SVG and a side-file `*-N.png` per raster block. We
keep the PNGs, drop the SVG/WMF temp files.
@@ -108,7 +108,7 @@ def _fp_extract_rtf_images(raw_bytes):
)
except (FileNotFoundError, subprocess.TimeoutExpired) as e:
_logger.warning(
'wmf2svg unavailable or timed out (%s) — skipping '
'wmf2svg unavailable or timed out (%s) - skipping '
'RTF image extraction.', e,
)
return []
@@ -125,7 +125,7 @@ def _fp_extract_rtf_images(raw_bytes):
def _fp_pick_microscope_image(png_bytes_list):
"""Pick the largest-area PNG (by pixel count, not file size) from
the list — that's almost always the microscope photo. Header
the list - that's almost always the microscope photo. Header
banners are wide-but-thin so their pixel area falls below the
threshold. Returns (png_bytes, width, height) or (None, 0, 0)
when no PNG meets the threshold.
@@ -153,7 +153,7 @@ _FISCHER_CALIB_RE = re.compile(r'Calibr\.\s*Std\.\s*Set\s+(.+?)(?:\s{2,}|$)',
_FISCHER_OPERATOR_RE = re.compile(r'Operator:\s*(\S+)', re.IGNORECASE)
_FISCHER_DATE_RE = re.compile(r'Date:\s*([\d/]+)', re.IGNORECASE)
_FISCHER_TIME_RE = re.compile(r'Time:\s*([\d:]+\s*[APMapm]*)')
# XDAL 600 header lines — only present on full RTF reports (not on
# XDAL 600 header lines - only present on full RTF reports (not on
# the .docx body the upstream parser already handled).
_FISCHER_PRODUCT_RE = re.compile(r'Product:\s*([^\r\n]+?)(?:\s{2,}|$)', re.IGNORECASE)
_FISCHER_DIRECTORY_RE = re.compile(r'Directory:\s*([^\r\n]+?)(?:\s{2,}|$)', re.IGNORECASE)
@@ -168,14 +168,14 @@ def _fp_strip_rtf(raw_bytes):
all of those plus the hex-encoded image data so the Fischerscope
reading regex hits clean text.
Not a full parser — meant for the narrow case of XRF/XDAL reports
Not a full parser - meant for the narrow case of XRF/XDAL reports
that have a simple body wrapped around an embedded WMF image.
"""
if not raw_bytes:
return ''
# RTF is ASCII-safe; latin-1 round-trips every byte.
text = raw_bytes.decode('latin-1', errors='replace')
# Drop destination groups entirely — these are the image data,
# Drop destination groups entirely - these are the image data,
# font tables, color tables, etc. The pattern `{\* ...}` and other
# nested destinations carry binary-ish hex strings we never want.
text = re.sub(r'\{\\\*[^{}]*\}', ' ', text)
@@ -185,7 +185,7 @@ def _fp_strip_rtf(raw_bytes):
# part between `\pict...goal\d+` and the closing brace of the group.
# Easier: nuke anything matching the picture marker through the
# next closing brace at the same depth (single-level approximation
# — works for FedEx/XRF docs that have one image per pict block).
# - works for FedEx/XRF docs that have one image per pict block).
text = re.sub(r'\{\\pict[^{}]*\}', ' ', text)
# Remove control words like \rtf1, \ansicpg1252, \par, \tab,
# \tx2840, etc. (`\` + letters + optional digits + optional space)
@@ -204,7 +204,7 @@ def _fp_strip_rtf(raw_bytes):
def _fp_parse_fischerscope_rtf(raw_bytes):
"""Fischerscope XDAL 600 RTF export → same dict shape as the
.docx parser. RTF detection is by magic bytes (`{\\rtf`) — the
.docx parser. RTF detection is by magic bytes (`{\\rtf`) - the
XRF software names the file `.doc` for legacy reasons, but the
contents are RTF.
"""
@@ -267,7 +267,7 @@ def _fp_parse_fischerscope_docx(raw_bytes):
}
Soft-fails to an empty dict-like result when python-docx isn't
installed or the bytes don't parse — the wizard still works, the
installed or the bytes don't parse - the wizard still works, the
operator just has to type readings manually.
"""
empty = {
@@ -280,7 +280,7 @@ def _fp_parse_fischerscope_docx(raw_bytes):
import docx # python-docx
except ImportError:
_logger.info(
'python-docx not installed — Fischerscope auto-parse '
'python-docx not installed - Fischerscope auto-parse '
'skipped. Operator will enter readings manually.'
)
return empty
@@ -335,7 +335,7 @@ def _fp_parse_fischerscope_docx(raw_bytes):
class FpCertIssueWizard(models.TransientModel):
_name = 'fp.cert.issue.wizard'
_description = 'Fusion Plating — Issue Certs Wizard'
_description = 'Fusion Plating - Issue Certs Wizard'
job_id = fields.Many2one(
'fp.job', string='Job', required=True, readonly=True,
@@ -357,7 +357,7 @@ class FpCertIssueWizard(models.TransientModel):
@api.model
def open_for_job(self, job):
"""Factory — create a wizard pre-populated with one line per
"""Factory - create a wizard pre-populated with one line per
draft cert on the job. Returns an action dict that opens the
wizard form."""
Cert = self.env['fp.certificate'].sudo()
@@ -375,7 +375,7 @@ class FpCertIssueWizard(models.TransientModel):
})
return {
'type': 'ir.actions.act_window',
'name': _('Issue Certs — %s') % job.name,
'name': _('Issue Certs - %s') % job.name,
'res_model': self._name,
'res_id': wiz.id,
'view_mode': 'form',
@@ -416,7 +416,7 @@ class FpCertIssueWizard(models.TransientModel):
class FpCertIssueWizardLine(models.TransientModel):
_name = 'fp.cert.issue.wizard.line'
_description = 'Fusion Plating — Issue Certs Wizard Line'
_description = 'Fusion Plating - Issue Certs Wizard Line'
wizard_id = fields.Many2one(
'fp.cert.issue.wizard', required=True, ondelete='cascade',
@@ -438,7 +438,7 @@ class FpCertIssueWizardLine(models.TransientModel):
fischer_filename = fields.Char(string='Filename')
# Optional: microscope/coupon image exported separately from the
# XDAL 600. The RTF carries an embedded WMF that the entech host
# can't rasterize (no imagemagick/libwmf — see CLAUDE.md "entech
# can't rasterize (no imagemagick/libwmf - see CLAUDE.md "entech
# apt is in a broken-deps state"), so the operator exports a PNG
# from the XDAL software and uploads it here. Rendered inline on
# the CoC's thickness section when present.
@@ -464,8 +464,8 @@ class FpCertIssueWizardLine(models.TransientModel):
'cert_id.partner_id.x_fc_strict_thickness_required',
'cert_id.x_fc_job_id.recipe_id.requires_thickness_report')
def _compute_needs_thickness(self):
# Delegate to fp.certificate._fp_needs_thickness_data — the single
# source of truth shared with the action_issue hard gate — so the
# Delegate to fp.certificate._fp_needs_thickness_data - the single
# source of truth shared with the action_issue hard gate - so the
# wizard's readiness hint and the gate can never drift. Honours
# recipe-level thickness suppression (passivation = no thickness
# even if the customer asked).
@@ -493,7 +493,7 @@ class FpCertIssueWizardLine(models.TransientModel):
@api.onchange('fischer_file', 'fischer_filename')
def _onchange_fischer_file(self):
"""Parse .docx OR RTF on upload (XDAL 600 names RTF files
`.doc` — detected by magic bytes; see CLAUDE.md "Fischerscope
`.doc` - detected by magic bytes; see CLAUDE.md "Fischerscope
XDAL 600 `.doc` files are actually RTF"). Prefill the readings
+ summary so the operator can verify before issuing."""
if not self.fischer_file:
@@ -511,7 +511,7 @@ class FpCertIssueWizardLine(models.TransientModel):
parsed = _fp_parse_fischerscope_docx(raw)
else:
self.parsed_summary = _(
'Non-parseable upload (%s) — file will be attached as '
'Non-parseable upload (%s) - file will be attached as '
'evidence. Type readings manually below if needed.'
) % (self.fischer_filename or 'unnamed')
return
@@ -531,9 +531,9 @@ class FpCertIssueWizardLine(models.TransientModel):
'Operator: %(o)s · Date: %(d)s %(t)s'
) % {
'n': len(readings),
'c': parsed.get('calibration') or '—',
'o': parsed.get('operator') or '—',
'd': parsed.get('date_str') or '—',
'c': parsed.get('calibration') or '-',
'o': parsed.get('operator') or '-',
'd': parsed.get('date_str') or '-',
't': parsed.get('time_str') or '',
}
@@ -559,7 +559,7 @@ class FpCertIssueWizardLine(models.TransientModel):
for fname, fval in field_map:
if fname in cert._fields and fval:
vals[fname] = fval
# Combine the gauge's date+time and parse to Datetime — try a
# Combine the gauge's date+time and parse to Datetime - try a
# few formats since XDAL exports vary (12h vs 24h, with/without
# seconds). Best-effort: leave the field blank if no format
# matches rather than crashing the cert issue.
@@ -589,7 +589,7 @@ class FpCertIssueWizardLine(models.TransientModel):
Order matters: operator-uploaded PNG must run LAST so it wins
over any image the RTF auto-extraction picked. Reverse order
(PNG first, then RTF) lets the WMF blow away the explicit
operator choice — exactly the bug we just hit.
operator choice - exactly the bug we just hit.
"""
self.ensure_one()
cert = self.cert_id.sudo()
@@ -602,13 +602,13 @@ class FpCertIssueWizardLine(models.TransientModel):
name = (self.fischer_filename or 'fischerscope').lower()
calibration = '' # backfilled below if the parser hits
if name.endswith('.pdf'):
# Drop the PDF into the cert-local field — merges into page 2.
# Drop the PDF into the cert-local field - merges into page 2.
cert.write({
'x_fc_local_thickness_pdf': self.fischer_file,
'x_fc_local_thickness_pdf_filename': self.fischer_filename,
})
else:
# .doc / .docx / anything else — attach as evidence AND
# .doc / .docx / anything else - attach as evidence AND
# link the attachment to the cert's evidence slot so the
# thickness-required gate recognises it. Without the link,
# the gate would still raise (it checks specific fields,
@@ -675,7 +675,7 @@ class FpCertIssueWizardLine(models.TransientModel):
'Fischerscope file <b>%s</b> attached via Issue wizard.'
)) % (self.fischer_filename or 'unnamed'))
self._push_readings_to_cert(calibration=calibration)
# Operator's PNG upload wins over auto-extracted WMF — runs
# Operator's PNG upload wins over auto-extracted WMF - runs
# last so it overwrites x_fc_thickness_image_id if both paths
# supplied an image.
self._apply_image_to_cert(cert)
@@ -704,7 +704,7 @@ class FpCertIssueWizardLine(models.TransientModel):
def _push_readings_to_cert(self, calibration=''):
"""Create fp.thickness.reading rows on the cert from wizard rows.
Skips when no rows. Does not deduplicate against existing
readings — the manager has just told us this is the new data.
readings - the manager has just told us this is the new data.
Per-reading calibration_std_ref is stamped from the optional
`calibration` arg so the printed CoC's calibration line stays
accurate even when readings are re-pushed from a fresh upload.
@@ -729,7 +729,7 @@ class FpCertIssueWizardLine(models.TransientModel):
class FpCertIssueWizardReading(models.TransientModel):
_name = 'fp.cert.issue.wizard.reading'
_description = 'Fusion Plating — Issue Certs Wizard Reading Row'
_description = 'Fusion Plating - Issue Certs Wizard Reading Row'
_order = 'sequence, id'
line_id = fields.Many2one(

4
fusion_plating/fusion_plating_jobs/wizards/fp_cert_issue_wizard_views.xml
View File

@@ -15,7 +15,7 @@
<sheet>
<div class="oe_title">
<h2>
Issue Certs —
Issue Certs -
<field name="job_id" readonly="1" nolabel="1"/>
</h2>
</div>
@@ -30,7 +30,7 @@
</div>
<!-- 2026-05-20: surface the file upload INLINE in the
list instead of behind a row-click into a sub-form.
Operators kept missing the upload affordance — the
Operators kept missing the upload affordance - the
list looked like a status display, not an action
surface. Adding the binary field as a column lets
them drop the Fischerscope file right where they

26
fusion_plating/fusion_plating_jobs/wizards/fp_job_step_input_wizard.py
View File

@@ -18,7 +18,7 @@ the job form.
Captured values land on a synthetic `fp.job.step.move` row with
transfer_type='step' (an in-place move, no destination change) so the
existing CoC chronological QWeb template renders them in the same
format as the tablet-captured values — single source of truth for
format as the tablet-captured values - single source of truth for
report rendering.
"""
@@ -51,7 +51,7 @@ _FP_INPUT_TYPE_SELECTION = [
class FpJobStepInputWizard(models.TransientModel):
_name = 'fp.job.step.input.wizard'
_description = 'Fusion Plating — Step Input Recording (Backend)'
_description = 'Fusion Plating - Step Input Recording (Backend)'
step_id = fields.Many2one(
'fp.job.step', string='Step', required=True, readonly=True,
@@ -76,11 +76,11 @@ class FpJobStepInputWizard(models.TransientModel):
return defaults
defaults['step_id'] = step.id
node = step.recipe_node_id
# Sub 12d — master switch — when off, return no input rows.
# Sub 12d - master switch - when off, return no input rows.
if hasattr(node, 'collect_measurements') and not node.collect_measurements:
defaults['line_ids'] = []
return defaults
# Filter to step_input prompts only — transition inputs go on the
# Filter to step_input prompts only - transition inputs go on the
# Move wizard, not here. Also filter to collect=True (per-recipe
# opt-out, default True).
inputs = node.input_ids
@@ -106,7 +106,7 @@ class FpJobStepInputWizard(models.TransientModel):
'Click "Add a line" in the table above to enter an '
'ad-hoc measurement.'
))
# Ad-hoc rows must have a prompt name — otherwise we can't tell
# Ad-hoc rows must have a prompt name - otherwise we can't tell
# what was being measured on the audit trail.
unnamed = self.line_ids.filtered(
lambda l: not l.node_input_id and not (l.name or '').strip()
@@ -143,7 +143,7 @@ class FpJobStepInputWizard(models.TransientModel):
'value_boolean': line.value_boolean,
'value_date': line.value_date or False,
}
# Sub 12d — composite + photo input types serialise differently.
# Sub 12d - composite + photo input types serialise differently.
if line.is_photo_type and line.photo_value:
att = Attachment.create({
'name': line.photo_filename or 'photo.jpg',
@@ -206,12 +206,12 @@ class FpJobStepInputWizard(models.TransientModel):
class FpJobStepInputWizardLine(models.TransientModel):
_name = 'fp.job.step.input.wizard.line'
_description = 'Fusion Plating — Step Input Wizard Line'
_description = 'Fusion Plating - Step Input Wizard Line'
wizard_id = fields.Many2one(
'fp.job.step.input.wizard', required=True, ondelete='cascade',
)
# 2026-04-28 fix — node_input_id is optional now so operators can
# 2026-04-28 fix - node_input_id is optional now so operators can
# record ad-hoc measurements when the recipe has no authored prompts
# (the screenshot case: a step with zero step_input definitions
# rendered an empty wizard with no way to add anything). Authored
@@ -221,7 +221,7 @@ class FpJobStepInputWizardLine(models.TransientModel):
'fusion.plating.process.node.input', ondelete='set null',
)
name = fields.Char(string='Prompt')
# 2026-04-28 — convert input_type + target_unit from Char → Selection
# 2026-04-28 - convert input_type + target_unit from Char → Selection
# so operators pick from the curated dropdown. Free-text led to "kg"
# vs "kgs" vs "kilo" inconsistencies on the audit trail.
input_type = fields.Selection(
@@ -233,7 +233,7 @@ class FpJobStepInputWizardLine(models.TransientModel):
target_unit = fields.Selection(
FP_UOM_SELECTION,
string='Unit',
help='Pick from the curated list — keeps every step\'s readings '
help='Pick from the curated list - keeps every step\'s readings '
'in the same vocabulary across the shop.',
)
@@ -242,7 +242,7 @@ class FpJobStepInputWizardLine(models.TransientModel):
value_boolean = fields.Boolean(string='Yes/No')
value_date = fields.Datetime(string='Date / Time')
# Sub 12d — composite + photo input types
# Sub 12d - composite + photo input types
photo_value = fields.Binary(string='Photo', attachment=True)
photo_filename = fields.Char(string='Photo Filename')
point_1 = fields.Float(string='R1')
@@ -274,7 +274,7 @@ class FpJobStepInputWizardLine(models.TransientModel):
is_authored = fields.Boolean(
compute='_compute_is_authored',
help='True when this row originated from an authored recipe input. '
'Drives field readonly state — authored prompts are locked, '
'Drives field readonly state - authored prompts are locked, '
'ad-hoc rows are fully editable.',
)
@@ -285,7 +285,7 @@ class FpJobStepInputWizardLine(models.TransientModel):
# ---- Single-column value editor -----------------------------------------
# The previous wizard exposed FOUR value columns (text / number /
# yes-no / date) — operators saw 9 columns wide and got lost. We
# yes-no / date) - operators saw 9 columns wide and got lost. We
# collapse them into one "Value" column whose widget routes to the
# right typed field based on input_type. Booleans and dates get
# their own dedicated field (still per-row) so the widget behaves

14
fusion_plating/fusion_plating_jobs/wizards/fp_job_step_input_wizard_views.xml
View File

@@ -94,7 +94,7 @@
</record>
<!-- ================================================================== -->
<!-- v3 — Card-based stacked layout (light + dark mode aware) -->
<!-- v3 - Card-based stacked layout (light + dark mode aware) -->
<!-- -->
<!-- Replaces v2's wide editable table. Each measurement renders as a -->
<!-- card with: prompt name (header), type/unit pills (meta), and ONLY -->
@@ -133,7 +133,7 @@
<field name="line_ids" class="o_fp_input_card_list" nolabel="1">
<list editable="bottom" create="true" delete="true">
<!-- Hidden flag fields — drive value-cell visibility -->
<!-- Hidden flag fields - drive value-cell visibility -->
<field name="is_authored" column_invisible="1"/>
<field name="is_boolean_type" column_invisible="1"/>
<field name="is_date_type" column_invisible="1"/>
@@ -144,14 +144,14 @@
<field name="point_avg" column_invisible="1"/>
<field name="photo_filename" column_invisible="1"/>
<!-- Card header — prompt name (large, bold via SCSS) -->
<!-- Card header - prompt name (large, bold via SCSS) -->
<field name="name"
string="Measurement"
readonly="is_authored"
placeholder="e.g. Oven Temp, Bath Reading, Operator Initials"
class="o_fp_iw_prompt"/>
<!-- Meta — type + unit rendered as pills (top-right).
<!-- Meta - type + unit rendered as pills (top-right).
Distinct classes so each pill lands in its own
grid column (otherwise they stack on top of
each other and the labels overlap). -->
@@ -165,12 +165,12 @@
class="o_fp_iw_meta o_fp_iw_meta_unit"
optional="show"/>
<!-- Hidden by default — operator can opt in via the cog menu
<!-- Hidden by default - operator can opt in via the cog menu
if they want to see/edit target ranges per row -->
<field name="target_min" optional="hide"/>
<field name="target_max" optional="hide"/>
<!-- Mutually exclusive value widgets — only the one
<!-- Mutually exclusive value widgets - only the one
matching the row's input_type renders -->
<field name="value_number"
string="Value"
@@ -196,7 +196,7 @@
invisible="not is_photo_type"
class="o_fp_iw_value"/>
<!-- Composite type 1: Multi-Point Thickness — 5 readings -->
<!-- Composite type 1: Multi-Point Thickness - 5 readings -->
<field name="point_1" string="R1"
invisible="not is_multi_point_type"
class="o_fp_iw_extra" optional="show"/>

30
fusion_plating/fusion_plating_jobs/wizards/fp_job_step_move_wizard.py
View File

@@ -8,7 +8,7 @@
Mirrors the tablet's Move Parts dialog (`fusion_plating_shopfloor`'s
`move_parts_dialog.js`) so a manager running the whole job from the
backend form on a low-staffing day captures the same chain-of-custody
record the operator would create from the tablet — same `fp.job.step.move`
record the operator would create from the tablet - same `fp.job.step.move`
row + same `transition_input_value_ids` snapshot, same chatter trail,
same downstream report rendering.
@@ -40,7 +40,7 @@ _FP_INPUT_TYPE_SELECTION = [
class FpJobStepMoveWizard(models.TransientModel):
_name = 'fp.job.step.move.wizard'
_description = 'Fusion Plating — Move Step Wizard (Backend)'
_description = 'Fusion Plating - Move Step Wizard (Backend)'
job_id = fields.Many2one('fp.job', string='Job', required=True, readonly=True)
from_step_id = fields.Many2one(
@@ -97,7 +97,7 @@ class FpJobStepMoveWizard(models.TransientModel):
'wizard_id',
string='Compliance Prompts',
help='Authored transition inputs from the to-step\'s recipe node. '
'Capture the operator\'s answers — they snapshot to '
'Capture the operator\'s answers - they snapshot to '
'fp.job.step.move.input.value when the wizard commits.',
)
@@ -108,7 +108,7 @@ class FpJobStepMoveWizard(models.TransientModel):
ctx = self.env.context
from_step_id = ctx.get('default_from_step_id') or ctx.get('active_id')
if from_step_id and self.env.context.get('active_model') != 'fp.job.step':
# Came from job form button — active_id is the job, not the step
# Came from job form button - active_id is the job, not the step
from_step_id = ctx.get('default_from_step_id')
if from_step_id:
from_step = self.env['fp.job.step'].browse(from_step_id)
@@ -133,7 +133,7 @@ class FpJobStepMoveWizard(models.TransientModel):
defaults['to_step_id'] = next_step.id
# Pre-seed input_value_ids from authored prompts on
# both ends of the move so programmatic creators
# (script tests, RPC clients) get them too —
# (script tests, RPC clients) get them too -
# @api.onchange only fires in interactive UI.
seen = set()
rows = []
@@ -173,14 +173,14 @@ class FpJobStepMoveWizard(models.TransientModel):
def _onchange_to_step_seed_inputs(self):
"""Seed prompt rows from BOTH
* the to-step's recipe node `transition_input` prompts —
* the to-step's recipe node `transition_input` prompts -
authored compliance fields fired on move-in.
* the from-step's recipe node `step_input` prompts —
* the from-step's recipe node `step_input` prompts -
measurements that should be captured BEFORE leaving the
from-step (operator answers "what did you actually run?"
while the data is fresh).
2026-04-28 fix — previously only transition_input was pulled,
2026-04-28 fix - previously only transition_input was pulled,
which left the section empty for steps where the author only
defined step_input prompts. Operators tried to record inputs
at move time and got an unfillable form.
@@ -189,7 +189,7 @@ class FpJobStepMoveWizard(models.TransientModel):
wiz.input_value_ids = [(5, 0, 0)]
seen = set()
rows = []
# 1. From-step's step_input prompts — measurements captured
# 1. From-step's step_input prompts - measurements captured
# while finalising the step.
if wiz.from_step_id and wiz.from_step_id.recipe_node_id:
from_node = wiz.from_step_id.recipe_node_id
@@ -205,7 +205,7 @@ class FpJobStepMoveWizard(models.TransientModel):
'name': '%s (Step Input)' % inp.name,
'input_type': inp.input_type,
}))
# 2. To-step's transition_input prompts — compliance gates
# 2. To-step's transition_input prompts - compliance gates
# fired on entry to the next step.
if wiz.to_step_id and wiz.to_step_id.recipe_node_id:
to_node = wiz.to_step_id.recipe_node_id
@@ -243,7 +243,7 @@ class FpJobStepMoveWizard(models.TransientModel):
qty_here = int(self.from_step_id.qty_at_step or 0)
if qty_here > 0 and self.qty_moved > qty_here:
raise UserError(_(
'Cannot move %(req)s parts — only %(here)s currently '
'Cannot move %(req)s parts - only %(here)s currently '
'parked at "%(step)s". Adjust Qty Moved or split '
'across multiple moves.'
) % {
@@ -277,7 +277,7 @@ class FpJobStepMoveWizard(models.TransientModel):
'value_boolean': line.value_boolean,
'value_date': line.value_date or False,
}
# Ad-hoc rows (no node_input_id) — preserve the operator's typed
# Ad-hoc rows (no node_input_id) - preserve the operator's typed
# prompt label in value_text so the chronological CoC report
# still shows what was measured.
if not line.node_input_id and line.name:
@@ -325,13 +325,13 @@ class FpJobStepMoveWizardInput(models.TransientModel):
a transient mirror means the wizard form can be filled, cancelled,
and reopened without polluting the chain-of-custody audit log.
2026-04-28 — `node_input_id` is now optional so operators can add
2026-04-28 - `node_input_id` is now optional so operators can add
ad-hoc input rows directly from the Move dialog (operator types
the prompt label + value). Authored prompts still pre-fill
name + type as readonly; ad-hoc rows are fully editable. Same
pattern as the standalone Record Inputs wizard."""
_name = 'fp.job.step.move.wizard.input'
_description = 'Fusion Plating — Move Wizard Input Row'
_description = 'Fusion Plating - Move Wizard Input Row'
wizard_id = fields.Many2one(
'fp.job.step.move.wizard',
@@ -355,7 +355,7 @@ class FpJobStepMoveWizardInput(models.TransientModel):
is_authored = fields.Boolean(
compute='_compute_is_authored',
help='True when this row originated from an authored recipe input. '
'Drives field readonly state — authored prompts are locked, '
'Drives field readonly state - authored prompts are locked, '
'ad-hoc rows are fully editable.',
)

2
fusion_plating/fusion_plating_jobs/wizards/fp_job_step_move_wizard_views.xml
View File

@@ -48,7 +48,7 @@
</field>
<separator string="Notes"/>
<field name="notes" nolabel="1"
placeholder="Optional context — why this move, what to watch for next..."/>
placeholder="Optional context - why this move, what to watch for next..."/>
</sheet>
<footer>
<button name="action_commit" type="object"