fix(scheduling): use each candidate's own daysLeft for prediction

Different assessment targets have different end dates. Previously all
candidates used the current vehicle's daysLeft, causing wrong predictions.

Now each inventory vehicle computes its own daysLeft from its assessment
target's current_year_assessment_end_date. predictedAfterSwap uses the
candidate's own daysLeft instead of the current vehicle's.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

src/modules/scheduling/SuggestionDetail.tsx

@@ -229,7 +229,7 @@ export default function SuggestionDetail({ suggestion: s, onClose, onNotifySucce
                         <span className="text-[9px] text-slate-500 bg-slate-100 px-1.5 py-0.5 rounded flex items-center gap-0.5"><MapPin size={9} />{c.region}</span>
                         <span className="text-[9px] text-slate-400">{c.vehicleType}</span>
                         <span className="text-[9px] text-slate-300">{c.targetName || '库存'}</span>
-                        <span className="text-[9px] text-slate-400">剩余{v.daysLeft}天</span>
+                        <span className="text-[9px] text-slate-400">剩余{c.daysLeft}天</span>
                       </div>
                       {c.canQualifyAfterSwap ? (
                         <span className="text-[9px] font-bold text-emerald-600 flex items-center gap-0.5 bg-emerald-50 px-1.5 py-0.5 rounded">