#!/usr/bin/env python3 # -*- coding: utf-8 -*- """客户档案聚合 → 作战地图数据源 读 「客户档案完善」活动记录（数据/activities/*.json），按客户聚合， 与 customer_360 的 ERP 采购数据交叉，算出渗透率 / 竞对盘子 / 完善度， 输出 数据/customer_dossier.json 供客户360 页面渲染。 口径说明： · 金额一律换算为「万元」输出（customer_360 与 ERP 源为「元」） · 竞对盘子 gap_wan = 录入的 TOP3 供应商份额金额合计（我司可争取上限） · 渗透率 penetration = 我司累计采购额 / 客户最新年营收 · 完善度 score：5 维度各 20 分（决策链 / 主营业务 / 营收 / 供应商 / 我司份额） """ import glob import json import os from datetime import datetime WS = os.path.dirname(os.path.dirname(os.path.abspath(__file__))) ACT_DIR = os.path.join(WS, '数据', 'activities') C360 = os.path.join(WS, '数据', 'customer_360.json') OUT = os.path.join(WS, '数据', 'customer_dossier.json') DIM_LABEL = {'decision': '决策链', 'biz_desc': '主营业务', 'revenues': '近三年营收', 'suppliers': 'TOP3供应商', 'self_share': '我司份额'} def fnum(v): """宽容数值解析：'3,200万' / '3200' / '约3200' → 3200.0""" if v is None: return None s = str(v).replace(',', '').replace('，', '').strip() digits = '' for ch in s: if ch.isdigit() or ch == '.': digits += ch elif digits: break try: return float(digits) if digits else None except ValueError: return None def collect(): """扫描活动记录，取「客户档案完善」按客户聚合（后者覆盖前者=最新优先）""" files = sorted(glob.glob(os.path.join(ACT_DIR, 'ACT*.json'))) agg = {} for fp in files: try: r = json.load(open(fp, encoding='utf-8')) except Exception: continue if r.get('type') != '客户档案完善': continue cust = (r.get('customer') or '').strip() if not cust: continue b0 = (r.get('blocks') or [{}])[0] date = (r.get('date') or r.get('created_at') or '')[:10] cur = agg.get(cust) if cur and date and cur['last_date'] and date < cur['last_date']: continue # 已有更新的记录 # 合并：新记录为主，空字段回落到旧记录（避免分次录入互相覆盖） def pick(key, default): v = b0.get(key) if isinstance(v, list): return v if v else (cur or {}).get('_raw', {}).get(key) or default v = (v or '').strip() if isinstance(v, str) else v return v or (cur or {}).get('_raw', {}).get(key) or default raw = {k: pick(k, [] if k in ('decision', 'revenues', 'suppliers') else '') for k in ('decision', 'biz_desc', 'revenues', 'suppliers', 'self_share', 'main_rival')} agg[cust] = { 'name': cust, 'last_date': date or (cur or {}).get('last_date', ''), 'record_count': ((cur or {}).get('record_count', 0) + 1), 'sales': (r.get('sales') or (cur or {}).get('sales') or ''), '_raw': raw, } return agg def score(raw): """完善度 5×20=100""" s, missing = 0, [] dec = [x for x in (raw.get('decision') or []) if x.get('name') or x.get('title')] if dec: s += 20 else: missing.append(DIM_LABEL['decision']) if raw.get('biz_desc'): s += 20 else: missing.append(DIM_LABEL['biz_desc']) rev = [x for x in (raw.get('revenues') or []) if x.get('year') or x.get('amt')] if len(rev) >= 3: s += 20 elif rev: s += 10 missing.append('近三年营收（缺 %d 年）' % (3 - len(rev))) else: missing.append(DIM_LABEL['revenues']) sup = [x for x in (raw.get('suppliers') or []) if x.get('name') or x.get('project') or x.get('amt')] if len(sup) >= 3: s += 20 elif sup: s += 10 missing.append('TOP3供应商（缺 %d 家）' % (3 - len(sup))) else: missing.append(DIM_LABEL['suppliers']) if raw.get('self_share') or raw.get('main_rival'): s += 20 else: missing.append(DIM_LABEL['self_share']) return s, missing def main(): c360 = json.load(open(C360, encoding='utf-8')) by_name = {} for c in (c360.get('customers') or []): by_name[c.get('name')] = c agg = collect() out = {} for cust, rec in agg.items(): raw = rec.pop('_raw') sc, missing = score(raw) # —— 竞争格局 —— sup = [x for x in (raw.get('suppliers') or []) if (fnum(x.get('amt')) or fnum(x.get('pct')))] top_amts = [fnum(x.get('amt')) for x in sup if fnum(x.get('amt')) is not None] top_pcts = [fnum(x.get('pct')) for x in sup if fnum(x.get('pct')) is not None] gap_wan = round(sum(top_amts), 1) if top_amts else None top3_pct = round(sum(top_pcts), 1) if top_pcts else None # —— 营收 —— revs = [] for x in (raw.get('revenues') or []): a = fnum(x.get('amt')) if a is not None: revs.append({'year': str(x.get('year') or ''), 'amt_wan': a, 'source': x.get('source') or ''}) revs.sort(key=lambda x: x['year']) rev_latest = revs[-1] if revs else None # —— 我司采购额（ERP）—— erp = by_name.get(cust) or {} my_wan = round((erp.get('total') or 0) / 10000.0, 1) pen = round(my_wan / rev_latest['amt_wan'] * 100, 1) if (rev_latest and rev_latest['amt_wan']) else None dec = [x for x in (raw.get('decision') or []) if x.get('name') or x.get('title')] ties = [x for x in dec if x.get('role') == '决策人'] or [x for x in dec if x.get('power') == '高'] out[cust] = { 'name': cust, 'sales': rec.get('sales') or erp.get('sales') or '', 'last_date': rec.get('last_date', ''), 'record_count': rec.get('record_count', 1), 'decision': dec, 'biz_desc': raw.get('biz_desc') or '', 'revenues': revs, 'suppliers': sup, 'self_share': raw.get('self_share') or '', 'main_rival': raw.get('main_rival') or '', 'score': sc, 'grade': '完整' if sc >= 80 else ('部分' if sc >= 40 else '起步'), 'missing': missing, 'metrics': { 'my_amt_wan': my_wan, 'erp_y2026_wan': round((erp.get('y2026') or 0) / 10000.0, 1), 'revenue_latest_wan': rev_latest['amt_wan'] if rev_latest else None, 'revenue_latest_year': rev_latest['year'] if rev_latest else '', 'penetration': pen, 'top3_amt_wan': gap_wan, 'top3_pct': top3_pct, 'gap_wan': gap_wan, 'decision_count': len(dec), 'tie_names': [x.get('name') for x in ties if x.get('name')], 'in_erp': bool(erp), }, } doc = { 'updated_at': datetime.now().strftime('%Y-%m-%d %H:%M:%S'), 'count': len(out), 'customers': out, } json.dump(doc, open(OUT, 'w', encoding='utf-8'), ensure_ascii=False, separators=(',', ':')) print('DOSSIER_OK=1 count=%d out=%s (%d KB)' % (len(out), OUT, os.path.getsize(OUT) // 1024)) if out: top = sorted(out.values(), key=lambda x: -(x['metrics']['gap_wan'] or 0))[:5] print(' 竞对盘子 TOP%d：' % len(top)) for x in top: m = x['metrics'] print(' %-28s 完善度%3d ｜ 我司%8.1f万 ｜ 营收%s ｜ 份额%s%% ｜ 竞对盘子%s万' % (x['name'][:28], x['score'], m['my_amt_wan'], (m['revenue_latest_wan'] if m['revenue_latest_wan'] else '—'), x['self_share'] or '—', m['gap_wan'])) if __name__ == '__main__': main()