283 lines
13 KiB
Python
283 lines
13 KiB
Python
from enum import Enum
|
||
|
||
from rapidfuzz import process, fuzz
|
||
import re
|
||
import json
|
||
from pypinyin import lazy_pinyin
|
||
|
||
# 数字转换表（1-20，常见数字）
|
||
digit_to_chinese = {
|
||
"1": "一", "2": "二", "3": "三", "4": "四", "5": "五",
|
||
"6": "六", "7": "七", "8": "八", "9": "九", "10": "十",
|
||
"11": "十一", "12": "十二", "13": "十三", "14": "十四",
|
||
"15": "十五", "16": "十六", "17": "十七", "18": "十八",
|
||
"19": "十九", "20": "二十"
|
||
}
|
||
|
||
def arabic_to_chinese_number(text):
|
||
"""
|
||
将文中阿拉伯数字转换为中文数字
|
||
:param text: 输入文本
|
||
:return: 转换后的文本
|
||
"""
|
||
cn_to_arabic = {'一': '1', '二': '2', '三': '3', '四': '4', '五': '5',
|
||
'六': '6', '七': '7', '八': '8', '九': '9', '零': '0'}
|
||
arabic_to_cn = {v: k for k, v in cn_to_arabic.items()} # 反向映射
|
||
|
||
for num, cn in arabic_to_cn.items():
|
||
text = text.replace(num, cn)
|
||
|
||
return text
|
||
def text_to_pinyin(text):
|
||
"""将文本转换为拼音字符串"""
|
||
return ''.join(lazy_pinyin(text))
|
||
|
||
def load_standard_data(path):
|
||
with open(path, "r", encoding="utf-8") as f:
|
||
return json.load(f)
|
||
|
||
|
||
def extract_number(text):
|
||
"""
|
||
提取项目部中的数字（支持阿拉伯数字和中文数字），并转换为统一格式（中文数字）。
|
||
"""
|
||
match = re.search(r'(第?[一二三四五六七八九十百千零\d]+)', text)
|
||
if match:
|
||
num_str = match.group(1).replace("第", "")
|
||
if num_str.isdigit():
|
||
return digit_to_chinese.get(num_str, num_str) # 阿拉伯数字转中文
|
||
return num_str # 中文数字直接返回
|
||
return None
|
||
|
||
|
||
def standardize_company_and_project(input_company, input_project, standard_data):
|
||
"""
|
||
将口语化的公司名和项目部名转换为标准化名称。
|
||
|
||
参数:
|
||
input_company (str): 用户输入的公司名（可能是口语化或不完整的名称）。
|
||
input_project (str): 用户输入的项目部名（可能是口语化或不完整的名称）。
|
||
standard_data (dict): 标准化的公司名和项目部名数据，格式为 {公司名: [项目部名1, 项目部名2, ...]}。
|
||
|
||
返回:
|
||
tuple: (标准化公司名, 匹配的项目部名列表)。如果无法匹配，返回 (None, None)。
|
||
"""
|
||
# **1. 标准化公司名**
|
||
company_match = process.extractOne(input_company, standard_data.keys(), scorer=fuzz.ratio)
|
||
if not company_match or company_match[1] < 65: # 相似度低于 70 可能匹配错误
|
||
return None, None
|
||
standard_company = company_match[0]
|
||
|
||
# **2. 先尝试直接匹配最相似的项目名**
|
||
project_match = process.extractOne(input_project, standard_data[standard_company], scorer=fuzz.ratio)
|
||
print(f"项目部名称最相似：{project_match[0]},{project_match[1]}", flush=True)
|
||
if project_match and project_match[1] >= 86:
|
||
return standard_company, [project_match[0]] # 直接返回匹配的项目名
|
||
|
||
# **3. 提取项目部的数字部分**
|
||
query_number = extract_number(input_project)
|
||
|
||
# **4. 过滤所有符合数字的项目部**
|
||
matched_projects = []
|
||
for project in standard_data[standard_company]:
|
||
project_number = extract_number(project)
|
||
if query_number and query_number == project_number:
|
||
matched_projects.append(project)
|
||
|
||
return standard_company, matched_projects
|
||
|
||
def standardize_company_and_projectDepartment(input_company, input_project, origianl_company_list , company_project_department_map, company_pinyin_to_original_map = None):
|
||
"""
|
||
将口语化的公司名和项目部名转换为标准化名称。
|
||
|
||
参数:
|
||
input_company (str): 用户输入的公司名（可能是口语化或不完整的名称）。
|
||
input_project (str): 用户输入的项目部名（可能是口语化或不完整的名称）。
|
||
company_project_department_map (dict): 标准化的公司名和项目部名数据，格式为 {公司名: [项目部名1, 项目部名2, ...]}。
|
||
pinyin_to_original_map:分公司拼音和分公司原始名的映射
|
||
|
||
返回:
|
||
tuple: (标准化公司名, 匹配的项目部名列表)。如果无法匹配，返回 (None, None)。
|
||
"""
|
||
try:
|
||
# **1. 标准化公司名**
|
||
best_company_match = multiple_standardize_single_name(input_company, origianl_company_list,list(company_pinyin_to_original_map.keys()),company_pinyin_to_original_map,60,85,True)
|
||
|
||
if not best_company_match:
|
||
return None, None
|
||
else:
|
||
standard_company = best_company_match[0]
|
||
|
||
# **2. 先尝试直接匹配最相似的项目名**
|
||
project_match = process.extract(input_project, company_project_department_map[standard_company], scorer=fuzz.token_sort_ratio, limit=len(company_project_department_map[standard_company]))
|
||
# project_match = process.extractOne(input_project, company_project_department_map[standard_company], scorer=fuzz.ratio)
|
||
print(f"项目部名称最相似：{project_match[0]},{project_match[1]}", flush=True)
|
||
if project_match and project_match[1] >= 86:
|
||
return standard_company, [project_match[0]] # 直接返回匹配的项目名
|
||
|
||
# **3. 提取项目部的数字部分**
|
||
query_number = extract_number(input_project)
|
||
|
||
# **4. 过滤所有符合数字的项目部**
|
||
matched_projects = []
|
||
for project in company_project_department_map[standard_company]:
|
||
project_number = extract_number(project)
|
||
if query_number and query_number == project_number:
|
||
matched_projects.append(project)
|
||
|
||
return standard_company, matched_projects
|
||
except Exception as e:
|
||
print(f"standardize_company_and_projectDepartment：{e}", flush=True)
|
||
return None,None
|
||
|
||
#弃用
|
||
def standardize_single_name(input_name, name_list, lower_score=70, high_score=85):
|
||
"""
|
||
将输入的名称（可能是口语化或不完整的名称）转换为标准化名称。
|
||
|
||
参数:
|
||
input_name (str): 用户输入的名称（可能是口语化或不完整的名称）。
|
||
name_list (list): 标准化的名称列表。
|
||
lower_score (int): 匹配的最低相似度阈值，默认值为 70。
|
||
high_score (int): 匹配的高置信度阈值，默认值为 85。
|
||
|
||
返回:
|
||
list: 匹配的标准化名称列表。如果未找到匹配项，返回 None。
|
||
"""
|
||
match_results = process.extract(input_name, name_list, scorer=fuzz.token_sort_ratio, limit=len(name_list))
|
||
|
||
# 找到所有相似度 > 80 的匹配项
|
||
high_confidence_matches = [(match[0], match[1]) for match in match_results if match[1] > lower_score]
|
||
print(f"standardize_single_name, high_confidence_matches:{high_confidence_matches}", flush=True)
|
||
if not high_confidence_matches:
|
||
return None # 没有找到匹配项
|
||
# 返回匹配结果
|
||
|
||
best_match = max(high_confidence_matches, key=lambda x: x[1], default=None)
|
||
|
||
print(f"best_match: {best_match}", flush=True)
|
||
if best_match and best_match[1] >= high_score:
|
||
return [best_match[0]] # 直接返回最高相似度的单个匹配项
|
||
|
||
return [match[0] for match in high_confidence_matches]
|
||
|
||
|
||
def multiple_standardize_single_name(origin_input_name, origin_name_list, pinyin_name_list = None, pinyin_to_original_map = None, lower_score=70, high_score=85, isArabicNumConv = False):
|
||
"""
|
||
使用拼音 + rapidfuzz 进行关键词模糊匹配，并返回原始的标准名
|
||
:param input_name: 口语化的名称（中文）
|
||
:param name_list: 关键词列表（中文）
|
||
:pinyin_name_list: 关键词列表（拼音）
|
||
:param pinyin_to_original_map: 拼音到原始标准名的映射
|
||
:param lower_score: 低匹配分数阈值（默认70）
|
||
:param high_score: 高匹配分数阈值（默认85）
|
||
:return: 最匹配的原始关键词，或 None
|
||
"""
|
||
#First round, 原始标准名的匹配性查找，能找到直接返回
|
||
if isArabicNumConv:
|
||
origin_input_name = arabic_to_chinese_number(origin_input_name)
|
||
|
||
match_results = process.extract(origin_input_name, origin_name_list, scorer=fuzz.token_sort_ratio, limit=len(origin_name_list))
|
||
|
||
# 找到所有相似度 > 80 的匹配项
|
||
original_high_confidence_matches = [(match[0], match[1]) for match in match_results if match[1] >= lower_score]
|
||
print(f"standardize_pinyin_single_name 原始名匹配, high_confidence_matches:{original_high_confidence_matches[:3]}", flush=True)
|
||
|
||
combined_low_confidence_matches = []
|
||
if original_high_confidence_matches:
|
||
origin_best_match = max(original_high_confidence_matches, key=lambda x: x[1], default=None)
|
||
|
||
# 直接返回最高相似度的单个匹配项
|
||
# print(f"原始名匹配: {origin_best_match}", flush=True)
|
||
if origin_best_match and origin_best_match[1] >= high_score:
|
||
return [origin_best_match[0]]
|
||
|
||
else:
|
||
combined_low_confidence_matches = [match[0] for match in original_high_confidence_matches[:3]]
|
||
else:
|
||
if not pinyin_name_list or not pinyin_to_original_map:
|
||
return None #
|
||
|
||
#第二轮, 拼音名的匹配性查找，能找到直接返回
|
||
pinyin_input_name = text_to_pinyin(origin_input_name)
|
||
match_results = process.extract(pinyin_input_name, pinyin_name_list, scorer=fuzz.ratio, limit=len(pinyin_name_list))
|
||
|
||
# 筛选出匹配分数 > lower_score 的结果
|
||
pinyin_high_confidence_matches = [(match[0], match[1]) for match in match_results if match[1] >= lower_score]
|
||
print(f"standardize_pinyin_single_name 拼音匹配, input_name：{pinyin_input_name}, high_confidence_matches:{pinyin_high_confidence_matches[:3]}", flush=True)
|
||
|
||
if not pinyin_high_confidence_matches:
|
||
return combined_low_confidence_matches # 没有找到匹配项
|
||
|
||
# 选择最高相似度的匹配项
|
||
pinyin_best_match = max(pinyin_high_confidence_matches, key=lambda x: x[1], default=None)
|
||
|
||
if pinyin_best_match and pinyin_best_match[1] >= high_score:
|
||
return [pinyin_to_original_map[pinyin_best_match[0]]] # 直接返回最高相似度的原始工程名
|
||
|
||
combined_low_confidence_matches.extend(
|
||
[pinyin_to_original_map[match[0]] for match in pinyin_high_confidence_matches[:3]]
|
||
)
|
||
# 返回所有匹配项对应的原始名，最多返回最低匹配项的前5个
|
||
return list(dict.fromkeys(combined_low_confidence_matches))
|
||
|
||
def generate_project_prompt(matched_projects, original_name = "", type="项目部名"):
|
||
"""
|
||
生成提示信息，用于让用户确认匹配的项目名或分公司名或项目名。
|
||
|
||
参数:
|
||
matched_projects (list): 匹配的项目或分公司名称列表。
|
||
type (str): 提示信息的类型（例如 "项目名" 或 "分公司名"），默认值为 "项目名"。
|
||
|
||
返回:
|
||
str: 生成的提示信息。如果未找到匹配项，返回提示用户提供更准确信息的字符串。
|
||
"""
|
||
if not matched_projects:
|
||
return f"未找到匹配的{type}:{original_name}，请提供更准确的{type}信息。"
|
||
|
||
prompt = f"您说的{type}可能是："
|
||
for idx, project in enumerate(matched_projects, start=1):
|
||
prompt += f"第{idx}个： {project}，"
|
||
|
||
prompt += "请确认您要选择哪一个？"
|
||
return prompt
|
||
|
||
|
||
def load_standard_name(file_path: str):
|
||
"""
|
||
从指定文件中加载标准化的名称列表。
|
||
|
||
参数:
|
||
file_path (str): 文件路径，文件应包含标准化的名称列表，每行一个名称。
|
||
|
||
返回:
|
||
list: 从文件中读取的标准化名称列表。
|
||
|
||
异常:
|
||
FileNotFoundError: 如果文件不存在，抛出此异常。
|
||
Exception: 如果读取文件时发生其他错误，抛出此异常。
|
||
"""
|
||
try:
|
||
with open(file_path, 'r', encoding='utf-8') as file:
|
||
lines = [line.strip() for line in file if line.strip()]
|
||
return lines
|
||
except FileNotFoundError:
|
||
print(f"错误：文件 {file_path} 不存在", flush=True)
|
||
raise FileNotFoundError(f"错误：文件 {file_path} 不存在")
|
||
except Exception as e:
|
||
print(f"读取文件时发生错误：{e}", flush=True)
|
||
raise Exception(f"错误：文件 {file_path} 不存在")
|
||
|
||
|
||
class CheckResult(Enum):
|
||
NO_MATCH = 0 # 不符合检查条件
|
||
MATCH_FOUND = 1 # 匹配到了值
|
||
NEEDS_MORE_ROUNDS = 2 # 需要多轮
|
||
|
||
|
||
class StandardType(Enum):
|
||
#工程名检查
|
||
PROJECT_CHECK = 0
|
||
#项目名检查
|
||
PROGRAM_CHECK = 1
|