analysis_report_node.py 33.5 KB
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945
"""
分析报告生成节点

在补货建议工作流的最后一个节点执行,生成结构化分析报告。
包含四大板块的统计计算函数:库存概览、销量分析、库存健康度、补货建议。
"""

import logging
from decimal import Decimal, ROUND_HALF_UP

logger = logging.getLogger(__name__)


def _to_decimal(value) -> Decimal:
    """安全转换为 Decimal"""
    if value is None:
        return Decimal("0")
    return Decimal(str(value))


def calculate_inventory_overview(part_ratios: list[dict]) -> dict:
    """
    计算库存总体概览统计数据

    有效库存 = in_stock_unlocked_cnt + on_the_way_cnt + has_plan_cnt
    资金占用 = in_stock_unlocked_cnt + on_the_way_cnt(仅计算实际占用资金的库存)

    Args:
        part_ratios: PartRatio 字典列表

    Returns:
        库存概览统计字典
    """
    total_in_stock_unlocked_cnt = Decimal("0")
    total_in_stock_unlocked_amount = Decimal("0")
    total_on_the_way_cnt = Decimal("0")
    total_on_the_way_amount = Decimal("0")
    total_has_plan_cnt = Decimal("0")
    total_has_plan_amount = Decimal("0")
    total_avg_sales_cnt = Decimal("0")
    # 资金占用合计 = (在库未锁 + 在途) * 成本价
    total_capital_occupation = Decimal("0")

    for p in part_ratios:
        cost_price = _to_decimal(p.get("cost_price", 0))

        in_stock = _to_decimal(p.get("in_stock_unlocked_cnt", 0))
        on_way = _to_decimal(p.get("on_the_way_cnt", 0))
        has_plan = _to_decimal(p.get("has_plan_cnt", 0))

        total_in_stock_unlocked_cnt += in_stock
        total_in_stock_unlocked_amount += in_stock * cost_price
        total_on_the_way_cnt += on_way
        total_on_the_way_amount += on_way * cost_price
        total_has_plan_cnt += has_plan
        total_has_plan_amount += has_plan * cost_price
        
        # 资金占用 = 在库未锁 + 在途
        total_capital_occupation += (in_stock + on_way) * cost_price

        # 月均销量
        out_stock = _to_decimal(p.get("out_stock_cnt", 0))
        locked = _to_decimal(p.get("storage_locked_cnt", 0))
        ongoing = _to_decimal(p.get("out_stock_ongoing_cnt", 0))
        buy = _to_decimal(p.get("buy_cnt", 0))
        avg_sales = (out_stock + locked + ongoing + buy) / Decimal("3")
        total_avg_sales_cnt += avg_sales

    total_valid_storage_cnt = (
        total_in_stock_unlocked_cnt
        + total_on_the_way_cnt
        + total_has_plan_cnt
    )
    total_valid_storage_amount = (
        total_in_stock_unlocked_amount
        + total_on_the_way_amount
        + total_has_plan_amount
    )

    # 库销比:月均销量为零时标记为特殊值
    if total_avg_sales_cnt > 0:
        overall_ratio = total_valid_storage_cnt / total_avg_sales_cnt
    else:
        overall_ratio = Decimal("999")

    return {
        "total_valid_storage_cnt": total_valid_storage_cnt,
        "total_valid_storage_amount": total_valid_storage_amount,
        "total_capital_occupation": total_capital_occupation,
        "total_in_stock_unlocked_cnt": total_in_stock_unlocked_cnt,
        "total_in_stock_unlocked_amount": total_in_stock_unlocked_amount,
        "total_on_the_way_cnt": total_on_the_way_cnt,
        "total_on_the_way_amount": total_on_the_way_amount,
        "total_has_plan_cnt": total_has_plan_cnt,
        "total_has_plan_amount": total_has_plan_amount,
        "total_avg_sales_cnt": total_avg_sales_cnt,
        "overall_ratio": overall_ratio,
        "part_count": len(part_ratios),
    }


def calculate_sales_analysis(part_ratios: list[dict]) -> dict:
    """
    计算销量分析统计数据

    月均销量 = (out_stock_cnt + storage_locked_cnt + out_stock_ongoing_cnt + buy_cnt) / 3

    Args:
        part_ratios: PartRatio 字典列表

    Returns:
        销量分析统计字典
    """
    total_out_stock_cnt = Decimal("0")
    total_storage_locked_cnt = Decimal("0")
    total_out_stock_ongoing_cnt = Decimal("0")
    total_buy_cnt = Decimal("0")
    total_avg_sales_amount = Decimal("0")
    has_sales_part_count = 0
    no_sales_part_count = 0

    for p in part_ratios:
        cost_price = _to_decimal(p.get("cost_price", 0))

        out_stock = _to_decimal(p.get("out_stock_cnt", 0))
        locked = _to_decimal(p.get("storage_locked_cnt", 0))
        ongoing = _to_decimal(p.get("out_stock_ongoing_cnt", 0))
        buy = _to_decimal(p.get("buy_cnt", 0))

        total_out_stock_cnt += out_stock
        total_storage_locked_cnt += locked
        total_out_stock_ongoing_cnt += ongoing
        total_buy_cnt += buy

        avg_sales = (out_stock + locked + ongoing + buy) / Decimal("3")
        total_avg_sales_amount += avg_sales * cost_price

        if avg_sales > 0:
            has_sales_part_count += 1
        else:
            no_sales_part_count += 1

    total_avg_sales_cnt = (
        total_out_stock_cnt + total_storage_locked_cnt + total_out_stock_ongoing_cnt + total_buy_cnt
    ) / Decimal("3")

    return {
        "total_avg_sales_cnt": total_avg_sales_cnt,
        "total_avg_sales_amount": total_avg_sales_amount,
        "total_out_stock_cnt": total_out_stock_cnt,
        "total_storage_locked_cnt": total_storage_locked_cnt,
        "total_out_stock_ongoing_cnt": total_out_stock_ongoing_cnt,
        "total_buy_cnt": total_buy_cnt,
        "has_sales_part_count": has_sales_part_count,
        "no_sales_part_count": no_sales_part_count,
    }


def _classify_part(p: dict) -> str:
    """
    将配件分类为缺货/呆滞/低频/正常

    分类规则(按优先级顺序判断):
    - 缺货件: 有效库存 = 0 且 月均销量 >= 1
    - 呆滞件: 有效库存 > 0 且 90天出库数 = 0
    - 低频件: 月均销量 < 1 或 出库次数 < 3 或 出库间隔 >= 30天
    - 正常件: 不属于以上三类
    """
    in_stock = _to_decimal(p.get("in_stock_unlocked_cnt", 0))
    on_way = _to_decimal(p.get("on_the_way_cnt", 0))
    has_plan = _to_decimal(p.get("has_plan_cnt", 0))
    valid_storage = in_stock + on_way + has_plan

    out_stock = _to_decimal(p.get("out_stock_cnt", 0))
    locked = _to_decimal(p.get("storage_locked_cnt", 0))
    ongoing = _to_decimal(p.get("out_stock_ongoing_cnt", 0))
    buy = _to_decimal(p.get("buy_cnt", 0))
    avg_sales = (out_stock + locked + ongoing + buy) / Decimal("3")

    out_times = int(p.get("out_times", 0) or 0)
    out_duration = int(p.get("out_duration", 0) or 0)

    # 缺货件
    if valid_storage == 0 and avg_sales >= 1:
        return "shortage"

    # 呆滞件
    if valid_storage > 0 and out_stock == 0:
        return "stagnant"

    # 低频件
    if avg_sales < 1 or out_times < 3 or out_duration >= 30:
        return "low_freq"

    return "normal"


def calculate_inventory_health(part_ratios: list[dict]) -> dict:
    """
    计算库存构成健康度统计数据

    将每个配件归类为缺货件/呆滞件/低频件/正常件,统计各类型数量/金额/百分比,
    并生成 chart_data 供前端图表使用。

    Args:
        part_ratios: PartRatio 字典列表

    Returns:
        健康度统计字典(含 chart_data)
    """
    categories = {
        "shortage": {"count": 0, "amount": Decimal("0")},
        "stagnant": {"count": 0, "amount": Decimal("0")},
        "low_freq": {"count": 0, "amount": Decimal("0")},
        "normal": {"count": 0, "amount": Decimal("0")},
    }

    for p in part_ratios:
        cat = _classify_part(p)
        cost_price = _to_decimal(p.get("cost_price", 0))

        # 有效库存金额
        in_stock = _to_decimal(p.get("in_stock_unlocked_cnt", 0))
        on_way = _to_decimal(p.get("on_the_way_cnt", 0))
        has_plan = _to_decimal(p.get("has_plan_cnt", 0))
        valid_storage = in_stock + on_way + has_plan
        amount = valid_storage * cost_price

        categories[cat]["count"] += 1
        categories[cat]["amount"] += amount

    total_count = len(part_ratios)
    total_amount = sum(c["amount"] for c in categories.values())

    # 计算百分比
    result = {}
    for cat_name, data in categories.items():
        count_pct = (data["count"] / total_count * 100) if total_count > 0 else 0.0
        amount_pct = (float(data["amount"]) / float(total_amount) * 100) if total_amount > 0 else 0.0
        result[cat_name] = {
            "count": data["count"],
            "amount": data["amount"],
            "count_pct": round(count_pct, 2),
            "amount_pct": round(amount_pct, 2),
        }

    result["total_count"] = total_count
    result["total_amount"] = total_amount

    # chart_data 供前端 Chart.js 使用
    labels = ["缺货件", "呆滞件", "低频件", "正常件"]
    cat_keys = ["shortage", "stagnant", "low_freq", "normal"]
    result["chart_data"] = {
        "labels": labels,
        "count_values": [categories[k]["count"] for k in cat_keys],
        "amount_values": [float(categories[k]["amount"]) for k in cat_keys],
    }

    return result


def calculate_replenishment_summary(part_results: list) -> dict:
    """
    计算补货建议生成情况统计数据

    按优先级分类统计:
    - priority=1: 急需补货
    - priority=2: 建议补货
    - priority=3: 可选补货

    Args:
        part_results: 配件汇总结果列表(字典或 ReplenishmentPartSummary 对象)

    Returns:
        补货建议统计字典
    """
    urgent = {"count": 0, "amount": Decimal("0")}
    suggested = {"count": 0, "amount": Decimal("0")}
    optional = {"count": 0, "amount": Decimal("0")}

    for item in part_results:
        # 兼容字典和对象两种形式
        if isinstance(item, dict):
            priority = int(item.get("priority", 0))
            amount = _to_decimal(item.get("total_suggest_amount", 0))
        else:
            priority = getattr(item, "priority", 0)
            amount = _to_decimal(getattr(item, "total_suggest_amount", 0))

        if priority == 1:
            urgent["count"] += 1
            urgent["amount"] += amount
        elif priority == 2:
            suggested["count"] += 1
            suggested["amount"] += amount
        elif priority == 3:
            optional["count"] += 1
            optional["amount"] += amount

    total_count = urgent["count"] + suggested["count"] + optional["count"]
    total_amount = urgent["amount"] + suggested["amount"] + optional["amount"]

    return {
        "urgent": urgent,
        "suggested": suggested,
        "optional": optional,
        "total_count": total_count,
        "total_amount": total_amount,
    }


# ============================================================
# LLM 分析函数
# ============================================================

import os
import json
import time
from langchain_core.messages import SystemMessage, HumanMessage


def _load_prompt(filename: str) -> str:
    """从 prompts 目录加载提示词文件"""
    prompt_path = os.path.join(
        os.path.dirname(os.path.dirname(os.path.dirname(os.path.dirname(__file__)))),
        "prompts",
        filename,
    )
    with open(prompt_path, "r", encoding="utf-8") as f:
        return f.read()


def _format_decimal(value) -> str:
    """将 Decimal 格式化为字符串,用于填充提示词"""
    if value is None:
        return "0"
    return str(round(float(value), 2))


def _get_season_from_date(date_str: str) -> str:
    """
    根据日期字符串获取季节

    Args:
        date_str: 日期字符串,格式如 "2024-01-15" 或 "20240115"

    Returns:
        季节名称:春季/夏季/秋季/冬季
    """
    from datetime import datetime

    try:
        # 尝试解析不同格式的日期
        if "-" in date_str:
            dt = datetime.strptime(date_str[:10], "%Y-%m-%d")
        else:
            dt = datetime.strptime(date_str[:8], "%Y%m%d")
        month = dt.month
    except (ValueError, TypeError):
        # 解析失败时使用当前月份
        month = datetime.now().month

    if month in (3, 4, 5):
        return "春季(3-5月)"
    elif month in (6, 7, 8):
        return "夏季(6-8月)"
    elif month in (9, 10, 11):
        return "秋季(9-11月)"
    else:
        return "冬季(12-2月)"


def _parse_llm_json(content: str) -> dict:
    """
    解析 LLM 返回的 JSON 内容

    尝试直接解析,如果失败则尝试提取 ```json 代码块中的内容。
    """
    text = content.strip()

    # 尝试直接解析
    try:
        return json.loads(text)
    except json.JSONDecodeError:
        pass

    # 尝试提取 ```json ... ``` 代码块
    import re
    match = re.search(r"```json\s*(.*?)\s*```", text, re.DOTALL)
    if match:
        try:
            return json.loads(match.group(1))
        except json.JSONDecodeError:
            pass

    # 尝试提取 { ... } 块
    start = text.find("{")
    end = text.rfind("}")
    if start != -1 and end != -1 and end > start:
        try:
            return json.loads(text[start : end + 1])
        except json.JSONDecodeError:
            pass

    # 解析失败
    raise json.JSONDecodeError("无法从 LLM 响应中解析 JSON", text, 0)


def llm_analyze_inventory_overview(stats: dict, statistics_date: str = "", llm_client=None) -> tuple[dict, dict]:
    """
    LLM 分析库存概览

    Args:
        stats: calculate_inventory_overview 的输出
        statistics_date: 统计日期
        llm_client: LLM 客户端实例,为 None 时自动获取

    Returns:
        (llm_analysis_dict, usage_dict)
    """
    from ..llm import get_llm_client

    if llm_client is None:
        llm_client = get_llm_client()

    current_season = _get_season_from_date(statistics_date)

    prompt_template = _load_prompt("report_inventory_overview.md")
    prompt = prompt_template.format(
        part_count=stats.get("part_count", 0),
        total_valid_storage_cnt=_format_decimal(stats.get("total_valid_storage_cnt")),
        total_valid_storage_amount=_format_decimal(stats.get("total_valid_storage_amount")),
        total_avg_sales_cnt=_format_decimal(stats.get("total_avg_sales_cnt")),
        overall_ratio=_format_decimal(stats.get("overall_ratio")),
        total_in_stock_unlocked_cnt=_format_decimal(stats.get("total_in_stock_unlocked_cnt")),
        total_in_stock_unlocked_amount=_format_decimal(stats.get("total_in_stock_unlocked_amount")),
        total_on_the_way_cnt=_format_decimal(stats.get("total_on_the_way_cnt")),
        total_on_the_way_amount=_format_decimal(stats.get("total_on_the_way_amount")),
        total_has_plan_cnt=_format_decimal(stats.get("total_has_plan_cnt")),
        total_has_plan_amount=_format_decimal(stats.get("total_has_plan_amount")),
        current_season=current_season,
        statistics_date=statistics_date or "未知",
    )

    messages = [HumanMessage(content=prompt)]
    response = llm_client.invoke(messages)

    try:
        analysis = _parse_llm_json(response.content)
    except json.JSONDecodeError:
        logger.warning(f"库存概览 LLM JSON 解析失败,原始响应: {response.content[:200]}")
        analysis = {"error": "JSON解析失败", "raw": response.content[:200]}

    usage = {
        "provider": response.usage.provider,
        "model": response.usage.model,
        "prompt_tokens": response.usage.prompt_tokens,
        "completion_tokens": response.usage.completion_tokens,
    }
    return analysis, usage


def llm_analyze_sales(stats: dict, statistics_date: str = "", llm_client=None) -> tuple[dict, dict]:
    """
    LLM 分析销量

    Args:
        stats: calculate_sales_analysis 的输出
        statistics_date: 统计日期
        llm_client: LLM 客户端实例

    Returns:
        (llm_analysis_dict, usage_dict)
    """
    from ..llm import get_llm_client

    if llm_client is None:
        llm_client = get_llm_client()

    current_season = _get_season_from_date(statistics_date)

    prompt_template = _load_prompt("report_sales_analysis.md")
    prompt = prompt_template.format(
        total_avg_sales_cnt=_format_decimal(stats.get("total_avg_sales_cnt")),
        total_avg_sales_amount=_format_decimal(stats.get("total_avg_sales_amount")),
        has_sales_part_count=stats.get("has_sales_part_count", 0),
        no_sales_part_count=stats.get("no_sales_part_count", 0),
        total_out_stock_cnt=_format_decimal(stats.get("total_out_stock_cnt")),
        total_storage_locked_cnt=_format_decimal(stats.get("total_storage_locked_cnt")),
        total_out_stock_ongoing_cnt=_format_decimal(stats.get("total_out_stock_ongoing_cnt")),
        total_buy_cnt=_format_decimal(stats.get("total_buy_cnt")),
        current_season=current_season,
        statistics_date=statistics_date or "未知",
    )

    messages = [HumanMessage(content=prompt)]
    response = llm_client.invoke(messages)

    try:
        analysis = _parse_llm_json(response.content)
    except json.JSONDecodeError:
        logger.warning(f"销量分析 LLM JSON 解析失败,原始响应: {response.content[:200]}")
        analysis = {"error": "JSON解析失败", "raw": response.content[:200]}

    usage = {
        "provider": response.usage.provider,
        "model": response.usage.model,
        "prompt_tokens": response.usage.prompt_tokens,
        "completion_tokens": response.usage.completion_tokens,
    }
    return analysis, usage


def llm_analyze_inventory_health(stats: dict, statistics_date: str = "", llm_client=None) -> tuple[dict, dict]:
    """
    LLM 分析库存健康度

    Args:
        stats: calculate_inventory_health 的输出
        statistics_date: 统计日期
        llm_client: LLM 客户端实例

    Returns:
        (llm_analysis_dict, usage_dict)
    """
    from ..llm import get_llm_client

    if llm_client is None:
        llm_client = get_llm_client()

    current_season = _get_season_from_date(statistics_date)

    prompt_template = _load_prompt("report_inventory_health.md")
    prompt = prompt_template.format(
        total_count=stats.get("total_count", 0),
        total_amount=_format_decimal(stats.get("total_amount")),
        shortage_count=stats.get("shortage", {}).get("count", 0),
        shortage_count_pct=stats.get("shortage", {}).get("count_pct", 0),
        shortage_amount=_format_decimal(stats.get("shortage", {}).get("amount")),
        shortage_amount_pct=stats.get("shortage", {}).get("amount_pct", 0),
        stagnant_count=stats.get("stagnant", {}).get("count", 0),
        stagnant_count_pct=stats.get("stagnant", {}).get("count_pct", 0),
        stagnant_amount=_format_decimal(stats.get("stagnant", {}).get("amount")),
        stagnant_amount_pct=stats.get("stagnant", {}).get("amount_pct", 0),
        low_freq_count=stats.get("low_freq", {}).get("count", 0),
        low_freq_count_pct=stats.get("low_freq", {}).get("count_pct", 0),
        low_freq_amount=_format_decimal(stats.get("low_freq", {}).get("amount")),
        low_freq_amount_pct=stats.get("low_freq", {}).get("amount_pct", 0),
        normal_count=stats.get("normal", {}).get("count", 0),
        normal_count_pct=stats.get("normal", {}).get("count_pct", 0),
        normal_amount=_format_decimal(stats.get("normal", {}).get("amount")),
        normal_amount_pct=stats.get("normal", {}).get("amount_pct", 0),
        current_season=current_season,
        statistics_date=statistics_date or "未知",
    )

    messages = [HumanMessage(content=prompt)]
    response = llm_client.invoke(messages)

    try:
        analysis = _parse_llm_json(response.content)
    except json.JSONDecodeError:
        logger.warning(f"健康度 LLM JSON 解析失败,原始响应: {response.content[:200]}")
        analysis = {"error": "JSON解析失败", "raw": response.content[:200]}

    usage = {
        "provider": response.usage.provider,
        "model": response.usage.model,
        "prompt_tokens": response.usage.prompt_tokens,
        "completion_tokens": response.usage.completion_tokens,
    }
    return analysis, usage


def llm_analyze_replenishment_summary(stats: dict, statistics_date: str = "", llm_client=None) -> tuple[dict, dict]:
    """
    LLM 分析补货建议

    Args:
        stats: calculate_replenishment_summary 的输出
        statistics_date: 统计日期
        llm_client: LLM 客户端实例

    Returns:
        (llm_analysis_dict, usage_dict)
    """
    from ..llm import get_llm_client

    if llm_client is None:
        llm_client = get_llm_client()

    current_season = _get_season_from_date(statistics_date)

    prompt_template = _load_prompt("report_replenishment_summary.md")
    prompt = prompt_template.format(
        total_count=stats.get("total_count", 0),
        total_amount=_format_decimal(stats.get("total_amount")),
        urgent_count=stats.get("urgent", {}).get("count", 0),
        urgent_amount=_format_decimal(stats.get("urgent", {}).get("amount")),
        suggested_count=stats.get("suggested", {}).get("count", 0),
        suggested_amount=_format_decimal(stats.get("suggested", {}).get("amount")),
        optional_count=stats.get("optional", {}).get("count", 0),
        optional_amount=_format_decimal(stats.get("optional", {}).get("amount")),
        current_season=current_season,
        statistics_date=statistics_date or "未知",
    )

    messages = [HumanMessage(content=prompt)]
    response = llm_client.invoke(messages)

    try:
        analysis = _parse_llm_json(response.content)
    except json.JSONDecodeError:
        logger.warning(f"补货建议 LLM JSON 解析失败,原始响应: {response.content[:200]}")
        analysis = {"error": "JSON解析失败", "raw": response.content[:200]}

    usage = {
        "provider": response.usage.provider,
        "model": response.usage.model,
        "prompt_tokens": response.usage.prompt_tokens,
        "completion_tokens": response.usage.completion_tokens,
    }
    return analysis, usage


# ============================================================
# LangGraph 并发子图
# ============================================================

from typing import TypedDict, Optional, Any, Annotated, Dict

from langgraph.graph import StateGraph, START, END


def _merge_dict(left: Optional[dict], right: Optional[dict]) -> Optional[dict]:
    """合并字典,保留非 None 的值"""
    if right is not None:
        return right
    return left


def _sum_int(left: int, right: int) -> int:
    """累加整数"""
    return (left or 0) + (right or 0)


def _merge_str(left: Optional[str], right: Optional[str]) -> Optional[str]:
    """合并字符串,保留非 None 的值"""
    if right is not None:
        return right
    return left


class ReportLLMState(TypedDict, total=False):
    """并发 LLM 分析子图的状态"""

    # 输入:四大板块的统计数据(只读,由主函数写入)
    inventory_overview_stats: Annotated[Optional[dict], _merge_dict]
    sales_analysis_stats: Annotated[Optional[dict], _merge_dict]
    inventory_health_stats: Annotated[Optional[dict], _merge_dict]
    replenishment_summary_stats: Annotated[Optional[dict], _merge_dict]

    # 输入:统计日期(用于季节判断)
    statistics_date: Annotated[Optional[str], _merge_str]

    # 输出:四大板块的 LLM 分析结果(各节点独立写入)
    inventory_overview_analysis: Annotated[Optional[dict], _merge_dict]
    sales_analysis_analysis: Annotated[Optional[dict], _merge_dict]
    inventory_health_analysis: Annotated[Optional[dict], _merge_dict]
    replenishment_summary_analysis: Annotated[Optional[dict], _merge_dict]

    # LLM 使用量(累加)
    total_prompt_tokens: Annotated[int, _sum_int]
    total_completion_tokens: Annotated[int, _sum_int]
    llm_provider: Annotated[Optional[str], _merge_dict]
    llm_model: Annotated[Optional[str], _merge_dict]


def _node_inventory_overview(state: ReportLLMState) -> ReportLLMState:
    """并发节点:库存概览 LLM 分析"""
    stats = state.get("inventory_overview_stats")
    statistics_date = state.get("statistics_date", "")
    if not stats:
        return {"inventory_overview_analysis": {"error": "无统计数据"}}

    try:
        analysis, usage = llm_analyze_inventory_overview(stats, statistics_date)
        return {
            "inventory_overview_analysis": analysis,
            "total_prompt_tokens": usage.get("prompt_tokens", 0),
            "total_completion_tokens": usage.get("completion_tokens", 0),
            "llm_provider": usage.get("provider", ""),
            "llm_model": usage.get("model", ""),
        }
    except Exception as e:
        logger.error(f"库存概览 LLM 分析失败: {e}")
        return {"inventory_overview_analysis": {"error": str(e)}}


def _node_sales_analysis(state: ReportLLMState) -> ReportLLMState:
    """并发节点:销量分析 LLM 分析"""
    stats = state.get("sales_analysis_stats")
    statistics_date = state.get("statistics_date", "")
    if not stats:
        return {"sales_analysis_analysis": {"error": "无统计数据"}}

    try:
        analysis, usage = llm_analyze_sales(stats, statistics_date)
        return {
            "sales_analysis_analysis": analysis,
            "total_prompt_tokens": usage.get("prompt_tokens", 0),
            "total_completion_tokens": usage.get("completion_tokens", 0),
            "llm_provider": usage.get("provider", ""),
            "llm_model": usage.get("model", ""),
        }
    except Exception as e:
        logger.error(f"销量分析 LLM 分析失败: {e}")
        return {"sales_analysis_analysis": {"error": str(e)}}


def _node_inventory_health(state: ReportLLMState) -> ReportLLMState:
    """并发节点:健康度 LLM 分析"""
    stats = state.get("inventory_health_stats")
    statistics_date = state.get("statistics_date", "")
    if not stats:
        return {"inventory_health_analysis": {"error": "无统计数据"}}

    try:
        analysis, usage = llm_analyze_inventory_health(stats, statistics_date)
        return {
            "inventory_health_analysis": analysis,
            "total_prompt_tokens": usage.get("prompt_tokens", 0),
            "total_completion_tokens": usage.get("completion_tokens", 0),
            "llm_provider": usage.get("provider", ""),
            "llm_model": usage.get("model", ""),
        }
    except Exception as e:
        logger.error(f"健康度 LLM 分析失败: {e}")
        return {"inventory_health_analysis": {"error": str(e)}}


def _node_replenishment_summary(state: ReportLLMState) -> ReportLLMState:
    """并发节点:补货建议 LLM 分析"""
    stats = state.get("replenishment_summary_stats")
    statistics_date = state.get("statistics_date", "")
    if not stats:
        return {"replenishment_summary_analysis": {"error": "无统计数据"}}

    try:
        analysis, usage = llm_analyze_replenishment_summary(stats, statistics_date)
        return {
            "replenishment_summary_analysis": analysis,
            "total_prompt_tokens": usage.get("prompt_tokens", 0),
            "total_completion_tokens": usage.get("completion_tokens", 0),
            "llm_provider": usage.get("provider", ""),
            "llm_model": usage.get("model", ""),
        }
    except Exception as e:
        logger.error(f"补货建议 LLM 分析失败: {e}")
        return {"replenishment_summary_analysis": {"error": str(e)}}


def build_report_llm_subgraph() -> StateGraph:
    """
    构建并发 LLM 分析子图

    四个 LLM 节点从 START fan-out 并发执行,结果 fan-in 汇总到 END。
    """
    graph = StateGraph(ReportLLMState)

    # 添加四个并发节点
    graph.add_node("inventory_overview_llm", _node_inventory_overview)
    graph.add_node("sales_analysis_llm", _node_sales_analysis)
    graph.add_node("inventory_health_llm", _node_inventory_health)
    graph.add_node("replenishment_summary_llm", _node_replenishment_summary)

    # fan-out: START → 四个节点
    graph.add_edge(START, "inventory_overview_llm")
    graph.add_edge(START, "sales_analysis_llm")
    graph.add_edge(START, "inventory_health_llm")
    graph.add_edge(START, "replenishment_summary_llm")

    # fan-in: 四个节点 → END
    graph.add_edge("inventory_overview_llm", END)
    graph.add_edge("sales_analysis_llm", END)
    graph.add_edge("inventory_health_llm", END)
    graph.add_edge("replenishment_summary_llm", END)

    return graph.compile()


# ============================================================
# 主节点函数
# ============================================================


def _serialize_stats(stats: dict) -> dict:
    """将统计数据中的 Decimal 转换为 float,以便 JSON 序列化"""
    result = {}
    for k, v in stats.items():
        if isinstance(v, Decimal):
            result[k] = float(v)
        elif isinstance(v, dict):
            result[k] = _serialize_stats(v)
        elif isinstance(v, list):
            result[k] = [
                _serialize_stats(item) if isinstance(item, dict) else (float(item) if isinstance(item, Decimal) else item)
                for item in v
            ]
        else:
            result[k] = v
    return result


def generate_analysis_report_node(state: dict) -> dict:
    """
    分析报告生成主节点

    串联流程:
    1. 统计计算(四大板块)
    2. 并发 LLM 分析(LangGraph 子图)
    3. 汇总报告
    4. 写入数据库

    单板块 LLM 失败不影响其他板块。

    Args:
        state: AgentState 字典

    Returns:
        更新后的 state 字典
    """
    from .state import AgentState
    from ..models import AnalysisReport
    from ..services.result_writer import ResultWriter

    logger.info("[AnalysisReport] ========== 开始生成分析报告 ==========")
    start_time = time.time()

    part_ratios = state.get("part_ratios", [])
    part_results = state.get("part_results", [])

    # ---- 1. 统计计算 ----
    logger.info(f"[AnalysisReport] 统计计算: part_ratios={len(part_ratios)}, part_results={len(part_results)}")

    inventory_overview_stats = calculate_inventory_overview(part_ratios)
    sales_analysis_stats = calculate_sales_analysis(part_ratios)
    inventory_health_stats = calculate_inventory_health(part_ratios)
    replenishment_summary_stats = calculate_replenishment_summary(part_results)

    # 序列化统计数据(Decimal → float)
    io_stats_serialized = _serialize_stats(inventory_overview_stats)
    sa_stats_serialized = _serialize_stats(sales_analysis_stats)
    ih_stats_serialized = _serialize_stats(inventory_health_stats)
    rs_stats_serialized = _serialize_stats(replenishment_summary_stats)

    # ---- 2. 并发 LLM 分析 ----
    logger.info("[AnalysisReport] 启动并发 LLM 分析子图")

    statistics_date = state.get("statistics_date", "")

    subgraph = build_report_llm_subgraph()
    llm_state: ReportLLMState = {
        "inventory_overview_stats": io_stats_serialized,
        "sales_analysis_stats": sa_stats_serialized,
        "inventory_health_stats": ih_stats_serialized,
        "replenishment_summary_stats": rs_stats_serialized,
        "statistics_date": statistics_date,
        "inventory_overview_analysis": None,
        "sales_analysis_analysis": None,
        "inventory_health_analysis": None,
        "replenishment_summary_analysis": None,
        "total_prompt_tokens": 0,
        "total_completion_tokens": 0,
        "llm_provider": None,
        "llm_model": None,
    }

    try:
        llm_result = subgraph.invoke(llm_state)
    except Exception as e:
        logger.error(f"[AnalysisReport] LLM 子图执行异常: {e}")
        llm_result = llm_state  # 使用初始状态(所有分析为 None)

    # ---- 3. 汇总报告 ----
    inventory_overview_data = {
        "stats": io_stats_serialized,
        "llm_analysis": llm_result.get("inventory_overview_analysis") or {"error": "未生成"},
    }
    sales_analysis_data = {
        "stats": sa_stats_serialized,
        "llm_analysis": llm_result.get("sales_analysis_analysis") or {"error": "未生成"},
    }
    inventory_health_data = {
        "stats": ih_stats_serialized,
        "chart_data": ih_stats_serialized.get("chart_data"),
        "llm_analysis": llm_result.get("inventory_health_analysis") or {"error": "未生成"},
    }
    replenishment_summary_data = {
        "stats": rs_stats_serialized,
        "llm_analysis": llm_result.get("replenishment_summary_analysis") or {"error": "未生成"},
    }

    total_tokens = (
        (llm_result.get("total_prompt_tokens") or 0)
        + (llm_result.get("total_completion_tokens") or 0)
    )
    execution_time_ms = int((time.time() - start_time) * 1000)

    # ---- 4. 写入数据库 ----
    report = AnalysisReport(
        task_no=state.get("task_no", ""),
        group_id=state.get("group_id", 0),
        dealer_grouping_id=state.get("dealer_grouping_id", 0),
        dealer_grouping_name=state.get("dealer_grouping_name"),
        brand_grouping_id=state.get("brand_grouping_id"),
        inventory_overview=inventory_overview_data,
        sales_analysis=sales_analysis_data,
        inventory_health=inventory_health_data,
        replenishment_summary=replenishment_summary_data,
        llm_provider=llm_result.get("llm_provider") or "",
        llm_model=llm_result.get("llm_model") or "",
        llm_tokens=total_tokens,
        execution_time_ms=execution_time_ms,
        statistics_date=state.get("statistics_date", ""),
    )

    try:
        writer = ResultWriter()
        report_id = writer.save_analysis_report(report)
        writer.close()
        logger.info(f"[AnalysisReport] 报告已保存: id={report_id}, tokens={total_tokens}, 耗时={execution_time_ms}ms")
    except Exception as e:
        logger.error(f"[AnalysisReport] 报告写入数据库失败: {e}")

    # 返回更新后的状态
    return {
        "analysis_report": report.to_dict(),
        "llm_provider": llm_result.get("llm_provider") or state.get("llm_provider", ""),
        "llm_model": llm_result.get("llm_model") or state.get("llm_model", ""),
        "llm_prompt_tokens": llm_result.get("total_prompt_tokens") or 0,
        "llm_completion_tokens": llm_result.get("total_completion_tokens") or 0,
        "current_node": "generate_analysis_report",
        "next_node": "end",
    }