Files
orc-order-v2/app/services/order_service.py
T

585 lines
25 KiB
Python
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
"""
订单服务模块
---------
提供订单处理服务,协调Excel处理和订单合并流程。
"""
import os
import json
import re
import shutil
from datetime import datetime
from pathlib import Path
from typing import Dict, List, Optional, Tuple, Union, Any, Callable
from app.config.settings import ConfigManager
from app.core.utils.log_utils import get_logger
from app.core.excel.processor import ExcelProcessor
from app.core.excel.merger import PurchaseOrderMerger
from app.core.db.product_db import ProductDatabase
from app.core.db.order_metadata_db import OrderMetadataDB
from app.core.ocr.metadata_extractor import OrderMetadataExtractor, sanitize_for_filename
logger = get_logger(__name__)
class OrderService:
"""
订单服务:协调Excel处理和订单合并流程
"""
def __init__(self, config: Optional[ConfigManager] = None, missing_barcodes_cb: Optional[Callable[[List[str]], None]] = None):
"""
初始化订单服务
Args:
config: 配置管理器,如果为None则创建新的
missing_barcodes_cb: 缺失条码的回调函数
"""
logger.info("初始化OrderService")
self.config = config or ConfigManager()
self.missing_barcodes_cb = missing_barcodes_cb
# 创建共享的商品数据库实例
db_path = self.config.get_path('Paths', 'product_db', fallback='data/product_cache.db') if hasattr(self.config, 'get_path') else 'data/product_cache.db'
tpl_folder = self.config.get('Paths', 'template_folder', fallback='templates')
item_data = self.config.get('Templates', 'item_data', fallback='商品资料.xlsx')
tpl_path = os.path.join(tpl_folder, item_data)
self.product_db = ProductDatabase(db_path, tpl_path)
# 创建Excel处理器和采购单合并器
self.excel_processor = ExcelProcessor(
self.config,
product_db=self.product_db,
missing_barcodes_cb=self.missing_barcodes_cb
)
self.order_merger = PurchaseOrderMerger(self.config)
# 元信息识别器 + 单据元信息库
self.extractor = OrderMetadataExtractor()
self.metadata_db = OrderMetadataDB(db_path)
logger.info("OrderService初始化完成")
def get_latest_excel(self) -> Optional[str]:
"""
获取最新的Excel文件
Returns:
最新Excel文件路径,如果未找到则返回None
"""
return self.excel_processor.get_latest_excel()
def process_excel(self, file_path: Optional[str] = None, progress_cb: Optional[Callable[[int], None]] = None) -> Optional[str]:
"""
处理Excel订单文件,生成标准采购单
Args:
file_path: Excel文件路径,如果为None则处理最新的文件
Returns:
输出采购单文件路径,如果处理失败则返回None
"""
if not file_path:
file_path = self.excel_processor.get_latest_excel()
if not file_path:
logger.warning("未找到可处理的Excel文件")
return None
logger.info("OrderService开始处理最新Excel文件")
else:
logger.info(f"OrderService开始处理指定Excel文件: {file_path}")
# 检查是否需要特殊的供应商预处理(如杨碧月)
try:
from app.services.special_suppliers_service import SpecialSuppliersService
special_service = SpecialSuppliersService(self.config)
# 尝试识别并预处理(注意:这里不再传入 progress_cb 避免无限递归或重复进度条,
# 或者我们在 special_service 内部逻辑中处理完后直接返回结果)
# 为了避免循环调用,我们在 SpecialSuppliersService 内部不再调用 process_excel
# 而是让 process_excel 识别后自己决定是否处理预处理后的文件。
# 我们新增一个 check_and_preprocess 方法
preprocessed_path = self._check_special_preprocess(file_path)
if preprocessed_path:
logger.info(f"检测到特殊供应商,已生成预处理文件: {preprocessed_path}")
file_path = preprocessed_path
except Exception as e:
logger.error(f"检查特殊预处理时出错: {e}")
result_path = self.excel_processor.process_specific_file(file_path, progress_cb=progress_cb)
if not result_path:
return None
# 应用单据元信息识别 + 重命名 result 与原图
try:
meta = self._extract_and_save_metadata(file_path)
if meta:
result_path = self._apply_metadata_to_filenames(
result_path, file_path, meta
)
except Exception as e:
logger.error(f"应用单据元信息失败(不影响 result 文件): {e}")
return result_path
def _check_special_preprocess(self, file_path: str) -> Optional[str]:
"""检查并执行特殊的预处理(支持杨碧月、烟草公司、蓉城易购)"""
try:
from app.core.utils.file_utils import smart_read_excel
import pandas as pd
import re
# 仅读取前 50 行进行智能识别 (header=None 确保能读到第一行内容)
df_head = smart_read_excel(file_path, nrows=50, header=None)
df_str = df_head.astype(str)
# 1. 识别:烟草公司 (Tobacco)
# 特征:内容中包含“专卖证号”或特定证号“510109104938”
is_tobacco = df_str.apply(lambda x: x.str.contains('专卖证号|510109104938')).any().any()
if is_tobacco:
logger.info("识别到烟草公司订单,执行专用预处理...")
from app.services.tobacco_service import TobaccoService
tobacco_svc = TobaccoService(self.config)
return tobacco_svc.preprocess_tobacco_order(file_path)
# 2. 识别:蓉城易购 (Rongcheng Yigou)
# 特征:内容中包含单号标识“RCDH”
is_rongcheng = df_str.apply(lambda x: x.str.contains('RCDH')).any().any()
if is_rongcheng:
logger.info("识别到蓉城易购订单,执行专用预处理...")
from app.services.special_suppliers_service import SpecialSuppliersService
special_svc = SpecialSuppliersService(self.config)
return special_svc.preprocess_rongcheng_yigou(file_path)
# 3. 识别:杨碧月 (Yang Biyue)
# 特征:经手人列包含“杨碧月”
handler_col = None
for col in df_head.columns:
# 在前50行中搜索“经手人”关键字
if df_head[col].astype(str).str.contains('经手人').any():
handler_col = col
break
if handler_col is not None:
# 检查该列是否有“杨碧月”
if df_head[handler_col].astype(str).str.contains('杨碧月').any():
logger.info("识别到杨碧月订单,执行专用预处理...")
from app.services.special_suppliers_service import SpecialSuppliersService
special_svc = SpecialSuppliersService(self.config)
return special_svc.process_yang_biyue_only(file_path)
except Exception as e:
logger.warning(f"智能预处理识别失败: {e}")
return None
def get_purchase_orders(self) -> List[str]:
"""
获取采购单文件列表
Returns:
采购单文件路径列表
"""
return self.order_merger.get_purchase_orders()
def merge_purchase_orders(self, file_paths: List[str], progress_cb: Optional[Callable[[int], None]] = None) -> Optional[str]:
"""
合并指定的采购单文件
Args:
file_paths: 采购单文件路径列表
Returns:
合并后的采购单文件路径,如果合并失败则返回None
"""
logger.info(f"OrderService开始合并指定采购单: {file_paths}")
return self.merge_orders(file_paths, progress_cb)
def merge_all_purchase_orders(self, progress_cb: Optional[Callable[[int], None]] = None) -> Optional[str]:
"""
合并所有可用的采购单文件
Returns:
合并后的采购单文件路径,如果合并失败则返回None
"""
logger.info("OrderService开始合并所有采购单")
return self.merge_orders(None, progress_cb)
def merge_orders(self, file_paths: Optional[List[str]] = None, progress_cb: Optional[Callable[[int], None]] = None) -> Optional[str]:
"""
合并采购单
Args:
file_paths: 采购单文件路径列表,如果为None则处理所有采购单
Returns:
合并后的采购单文件路径,如果合并失败则返回None
"""
if file_paths:
logger.info(f"OrderService开始合并指定采购单: {file_paths}")
else:
logger.info("OrderService开始合并所有采购单")
return self.order_merger.process(file_paths, progress_cb)
def validate_unit_price(self, result_path: str) -> List[str]:
"""
校验采购单单价与商品资料进货价的差异
Args:
result_path: 待校验的采购单路径
Returns:
差异信息列表,无差异返回空列表
"""
try:
import pandas as pd
from app.core.utils.file_utils import smart_read_excel
from app.core.handlers.column_mapper import ColumnMapper as CM
# 使用共享的商品数据库实例
product_db = self.product_db
# 读取待校验的采购单
df_res = smart_read_excel(result_path)
res_barcode_col = CM.find_column(list(df_res.columns), 'barcode')
res_price_col = CM.find_column(list(df_res.columns), 'unit_price')
if not res_barcode_col or not res_price_col:
logger.warning("未能在采购单中找到条码或单价列")
return []
# 批量查询进货价
barcodes = df_res[res_barcode_col].astype(str).str.strip().tolist()
item_prices = product_db.get_prices(barcodes)
results = []
for _, row in df_res.iterrows():
bc = str(row[res_barcode_col]).strip()
if bc not in item_prices:
continue
try:
res_price = float(row[res_price_col])
except (ValueError, TypeError):
continue
item_price = item_prices[bc]
diff = abs(res_price - item_price)
if diff > 1.0:
results.append(f"条码 {bc}: 采购单价={res_price} vs 进货价={item_price} 差异={diff:.2f}")
return results
except Exception as e:
logger.error(f"单价校验过程中发生错误: {e}")
return []
# ══════════════════════════════════════════════════════════════
# 单据元信息识别 + 文件重命名
# ══════════════════════════════════════════════════════════════
def _extract_and_save_metadata(self, ocr_excel_path: str) -> Optional[Any]:
"""从 OCR 输出的 xlsx 同目录的 .meta.json 提取元信息,写入 SQLite。
Returns:
OrderMetadata 或 None(失败时)
"""
try:
base = Path(ocr_excel_path)
file_hash = base.stem
meta_path = base.with_suffix('.meta.json')
ocr_text = ''
ocr_rows: List[List[str]] = []
source_image = ''
if meta_path.exists():
try:
payload = json.loads(meta_path.read_text(encoding='utf-8'))
# 区分表格 OCR 文本和通用识别文本
ocr_text = payload.get('ocr_text', '') or ''
general_text = payload.get('general_text', '') or ''
ocr_rows = payload.get('ocr_rows', []) or []
source_image = payload.get('image_path', '') or ''
except Exception as e:
logger.warning(f"读 meta.json 失败: {e}")
# 兜底:从 xlsx 拼文本(与 OCRService._write_meta_json 的兜底一致)
if not ocr_text and not general_text:
try:
import xlrd
rb = xlrd.open_workbook(str(ocr_excel_path))
ws = rb.sheet_by_index(0)
lines = []
for r in range(ws.nrows):
row_vals = [str(ws.cell_value(r, c)) for c in range(ws.ncols)]
ocr_rows.append(row_vals)
lines.append(' '.join(row_vals))
ocr_text = '\n'.join(lines)
except Exception as e:
logger.debug(f"从 xlsx 拼 OCR 文本失败: {e}")
meta = self.extractor.extract(ocr_text, ocr_rows, general_text=general_text)
self.metadata_db.save(
file_hash=file_hash,
supplier=meta.supplier,
bill_date=meta.bill_date,
total_amount=meta.total_amount,
raw_supplier_text=meta.raw_supplier_text,
source_image=source_image,
)
logger.info(
f"元信息识别: hash={file_hash} supplier={meta.supplier!r} "
f"bill_date={meta.bill_date!r} total_amount={meta.total_amount:.2f}"
)
return meta
except Exception as e:
logger.error(f"_extract_and_save_metadata 失败: {e}")
return None
def _apply_metadata_to_filenames(self, result_path: str,
ocr_excel_path: str,
meta) -> str:
"""应用新文件名规则(按供应商+日期):
- result (xls){供应商}_{日期}.xls
- output xlsx {供应商}_{日期}.xlsx
- 原图: {供应商}_{日期}.{ext}
冲突时加 _2 / _3 ...;任一步骤失败不影响整体。
Returns:
新 result 路径(无论重命名是否成功都返回;失败时返回原路径)
"""
try:
# 1. 提取并清理元数据
supplier_clean = sanitize_for_filename(meta.supplier) or '未知供应商'
date_part = meta.bill_date or '未知日期'
# 2. 构造新的基础文件名:[供应商]_[日期]
# 按照用户要求:按照供应商名称加日期进行修改,且不含“采购单”等冗余字眼
base_name = f"{supplier_clean}_{date_part}"
file_hash = Path(ocr_excel_path).stem # 仅用作 SQLite PK,不入文件名
logger.info(f"生成标准化文件名: {base_name}")
# ── 1. result 重命名 ──
new_result_path = self._safe_rename(
result_path, f"{base_name}.xls"
)
# ── 2. output xlsx 重命名(保证 output/ 与 result/ 文件名一致)──
new_output_path = self._safe_rename(
ocr_excel_path, f"{base_name}.xlsx"
)
# 同步 meta.json
old_meta = Path(ocr_excel_path).with_suffix('.meta.json')
if old_meta.exists():
new_meta = Path(new_output_path).with_suffix('.meta.json')
if str(new_meta) != str(old_meta):
try:
if new_meta.exists(): os.remove(new_meta)
os.rename(old_meta, new_meta)
except Exception as e:
logger.warning(f"meta.json 重命名失败: {e}")
# ── 2.5 同步 processed_files.json(避免下次 OCR 重新识别)──
if str(new_output_path) != str(ocr_excel_path):
self._sync_processed_record(Path(ocr_excel_path), Path(new_output_path))
# 也要同步 ExcelProcessor 的记录
self._update_excel_process_record(Path(ocr_excel_path), Path(new_output_path))
# ── 3. 原图重命名 ──
try:
# 尝试从 processed_files.json 反查原图路径
src_image = self._find_input_image_for_output(str(ocr_excel_path))
# 如果没找到,再从 DB 拿
if not src_image:
meta_row = self.metadata_db.get(file_hash)
src_image = (meta_row or {}).get('source_image', '')
if src_image and os.path.exists(src_image):
src_p = Path(src_image)
ext = src_p.suffix
# 严禁包含原文件名,统一格式: 采购单_YYYYMMDD_供应商.ext
new_image_name = f"{base_name}{ext}"
new_image_path = self._safe_rename(src_image, new_image_name)
# ── 3.5 同步 processed_files.json 的 Key (原图路径) ──
if str(new_image_path) != str(src_p):
self._sync_processed_key(src_p, Path(new_image_path))
# 更新 source_image 路径到 DB
if str(new_image_path) != str(src_p):
self.metadata_db.save(
file_hash=file_hash,
supplier=meta.supplier,
bill_date=meta.bill_date,
total_amount=meta.total_amount,
raw_supplier_text=meta.raw_supplier_text,
source_image=str(new_image_path),
)
else:
logger.warning(f"找不到原图,跳过重命名: {src_image}")
except Exception as e:
logger.warning(f"原图重命名失败: {e}")
return str(new_result_path)
except Exception as e:
logger.error(f"_apply_metadata_to_filenames 失败: {e}")
return result_path
@staticmethod
def _safe_rename(src: str, new_basename: str) -> str:
"""把 src 重命名为 src.parent / new_basename。
- 目标已存在时加 _2 / _3 ... 后缀避免覆盖
- 失败返回原路径,不抛异常
"""
try:
src_p = Path(src)
if not src_p.exists():
return src
new_p = src_p.parent / new_basename
if str(new_p) == str(src_p):
return src
if new_p.exists():
# 冲突去重
stem, suffix = new_p.stem, new_p.suffix
n = 2
while True:
cand = src_p.parent / f"{stem}_{n}{suffix}"
if not cand.exists():
new_p = cand
break
n += 1
os.rename(src_p, new_p)
logger.info(f"重命名: {src_p.name} -> {new_p.name}")
return str(new_p)
except Exception as e:
logger.warning(f"重命名失败 {src} -> {new_basename}: {e}")
return src
@staticmethod
def _dedup_path(p: Path) -> Path:
"""路径冲突时加 _N 后缀。"""
stem, suffix = p.stem, p.suffix
parent = p.parent
n = 1
while True:
cand = parent / f"{stem}_{n}{suffix}"
if not cand.exists():
return cand
n += 1
def _find_input_image_for_output(self, output_xlsx: str) -> Optional[str]:
"""从 processed_files.json 反查生成该 output_xlsx 的原图路径。"""
try:
output_dir = self.config.get_path('Paths', 'output_folder', fallback='data/output')
record_file = os.path.join(output_dir, 'processed_files.json')
if not os.path.exists(record_file):
logger.debug(f"记录文件不存在: {record_file}")
return None
with open(record_file, 'r', encoding='utf-8') as f:
data = json.load(f)
target_output = os.path.normpath(output_xlsx)
found_img = None
# 1. 尝试直接匹配和规范化路径匹配
for img_path, out_path in data.items():
if os.path.normpath(out_path) == target_output:
found_img = img_path
break
if found_img:
if os.path.exists(found_img):
return found_img
# 2. 模糊匹配:如果记录中的路径不存在,尝试在同目录下找 stem 匹配的文件
# (解决因之前重命名导致的路径不一致问题)
img_p = Path(found_img)
parent = img_p.parent
if parent.exists():
stem = img_p.stem
logger.debug(f"尝试模糊匹配原图: stem={stem[:20]}... in {parent}")
# 尝试寻找以原 stem 开头的文件
for cand in parent.iterdir():
if cand.is_file() and cand.stem.startswith(stem[:20]):
logger.info(f"模糊匹配成功: {cand.name}")
return str(cand)
logger.debug("模糊匹配失败")
else:
logger.debug(f"在 processed_files.json 中未找到输出文件 {output_xlsx} 对应的原图记录")
except Exception as e:
logger.debug(f"反查原图路径失败: {e}")
return None
def _sync_processed_record(self, old_output: Path, new_output: Path) -> None:
"""重命名 output xlsx 后,把 processed_files.json 里所有 value 为旧路径的项改成新路径。"""
self._update_processed_json(old_path=old_output, new_path=new_output, is_key=False)
def _sync_processed_key(self, old_input: Path, new_input: Path) -> None:
"""重命名原图后,把 processed_files.json 里的 key 从旧路径改为新路径。"""
self._update_processed_json(old_path=old_input, new_path=new_input, is_key=True)
def _update_excel_process_record(self, old_xlsx: Path, new_xlsx: Path) -> None:
"""重命名 xlsx 后,同步 excel_process_records.json (ExcelProcessor 使用的记录)。"""
self._update_processed_json(
old_path=old_xlsx,
new_path=new_xlsx,
is_key=True,
filename='excel_process_records.json'
)
def _update_processed_json(self, old_path: Path, new_path: Path, is_key: bool = False, filename: str = 'processed_files.json') -> None:
"""更新处理记录 JSON。
Args:
old_path: 旧路径
new_path: 新路径
is_key: True 更新 key, False 更新 value
filename: JSON 文件名
"""
try:
# 确定文件路径
output_dir = self.config.get_path('Paths', 'output_folder', fallback='data/output')
record_file = os.path.join(output_dir, filename)
if not os.path.exists(record_file):
return
with open(record_file, 'r', encoding='utf-8') as f:
data = json.load(f)
old_str = str(old_path)
new_str = str(new_path)
changed = False
# 规范化路径以便匹配
old_norm = os.path.normpath(old_str)
if is_key:
# 更新 Key
for k in list(data.keys()):
if k == old_str or os.path.normpath(k) == old_norm:
data[new_str] = data.pop(k)
changed = True
break
else:
# 更新 Value
for k, v in list(data.items()):
if v == old_str or os.path.normpath(v) == old_norm:
data[k] = new_str
changed = True
if changed:
with open(record_file, 'w', encoding='utf-8') as f:
json.dump(data, f, ensure_ascii=False, indent=2)
logger.info(f"{filename} 已同步 ({'Key' if is_key else 'Value'}): {old_str} -> {new_str}")
except Exception as e:
logger.warning(f"更新 {filename} 失败: {e}")