KGPython/python/icd_parse.py

import pandas as pd
import json
import numpy as np
import re

# === 配置 ===
EXCEL_PATH = r"C:\Users\hanyuqing\Desktop\最新国家医保ICD编码\最新国家医保ICD编码\ICD-10医保版数据.xlsx"
OUTPUT_JSON_PATH = "icd10_tree_with_level.json"

# === 章 code 格式化 ===
def extract_chapter_number(text):
    if not text:
        return None
    text = str(text).strip()
    match = re.search(r'第([一二三四五六七八九十百\d]+)章', text)
    if match:
        num_str = match.group(1)
        chinese_num_map = {
            '一': 1, '二': 2, '三': 3, '四': 4, '五': 5,
            '六': 6, '七': 7, '八': 8, '九': 9, '十': 10,
            '十一': 11, '十二': 12, '十三': 13, '十四': 14, '十五': 15,
            '十六': 16, '十七': 17, '十八': 18, '十九': 19, '二十': 20,
            '二十一': 21, '二十二': 22
        }
        if num_str in chinese_num_map:
            return chinese_num_map[num_str]
        elif num_str.isdigit():
            return int(num_str)
    if text.isdigit():
        return int(text)
    if text.endswith('.') and text[:-1].isdigit():
        return int(text[:-1])
    return None

def format_chapter_code(original):
    num = extract_chapter_number(original)
    if num is not None:
        return f"第{num}章"
    return str(original).strip() if original else ""

def get_clean(val):
    if val is None or str(val).strip().lower() in ("", "nan", "none"):
        return None
    return str(val).strip()

def make_fields(code, name):
    code_str = code or ""
    label_str = name or ""
    title_str = f"{code_str} {label_str}" if code_str and label_str else (code_str or label_str)
    return code_str, label_str, title_str

# === 带 level 的树节点 ===
class TreeNode:
    def __init__(self, code="", label="", title="", level=""):
        self.code = code
        self.label = label
        self.title = title
        self.level = level  # 新增字段
        self.children = []
        self._child_key_set = set()  # (code, label)

    def add_or_get_child(self, code, label, title, level):
        key = (code, label)
        if key in self._child_key_set:
            for child in self.children:
                if child.code == code and child.label == label:
                    return child
        else:
            new_child = TreeNode(code=code, label=label, title=title, level=level)
            self.children.append(new_child)
            self._child_key_set.add(key)
            return new_child
        return None

# === 构建树 ===
root = TreeNode()
chapter_map = {}

df = pd.read_excel(
    EXCEL_PATH,
    header=1,
    dtype=str,
    engine='openpyxl'
)
df.columns = df.columns.astype(str).str.strip()
df = df.replace({np.nan: None})

for idx, row in df.iterrows():
    raw_chapter = get_clean(row.get("章"))
    chapter_name = get_clean(row.get("章的名称"))

    section_code = get_clean(row.get("节代码范围"))
    section_name = get_clean(row.get("节名称"))

    category_code = get_clean(row.get("类目代码"))
    category_name = get_clean(row.get("类目名称"))

    subcategory_code = get_clean(row.get("亚目代码"))
    subcategory_name = get_clean(row.get("亚目名称"))

    diagnosis_code = get_clean(row.get("条目（诊断）代码"))
    diagnosis_name = get_clean(row.get("条目（诊断）名称"))

    if not raw_chapter and not chapter_name:
        continue

    # === 章 ===
    chapter_code = format_chapter_code(raw_chapter)
    chap_label = chapter_name or chapter_code
    chap_title = f"{chapter_code} {chapter_name}" if chapter_name else chapter_code
    chap_key = (chapter_code, chap_label)

    if chap_key not in chapter_map:
        chapter_node = TreeNode(
            code=chapter_code,
            label=chap_label,
            title=chap_title,
            level="chapter"
        )
        root.children.append(chapter_node)
        chapter_map[chap_key] = chapter_node
    else:
        chapter_node = chapter_map[chap_key]

    current = chapter_node

    # === 节 ===
    if section_code or section_name:
        sec_code, sec_label, sec_title = make_fields(section_code, section_name)
        current = current.add_or_get_child(sec_code, sec_label, sec_title, "section")

    # === 类目 ===
    if category_code or category_name:
        cat_code, cat_label, cat_title = make_fields(category_code, category_name)
        current = current.add_or_get_child(cat_code, cat_label, cat_title, "category")

    # === 亚目 ===
    if subcategory_code or subcategory_name:
        sub_code, sub_label, sub_title = make_fields(subcategory_code, subcategory_name)
        current = current.add_or_get_child(sub_code, sub_label, sub_title, "subcategory")

    # === 条目 ===
    if diagnosis_code or diagnosis_name:
        diag_code, diag_label, diag_title = make_fields(diagnosis_code, diagnosis_name)
        current.add_or_get_child(diag_code, diag_label, diag_title, "diagnosis")

# === 转为带 id 的 dict ===
next_id = 1

def node_to_dict(node):
    global next_id
    item = {
        "id": next_id,
        "code": node.code,
        "label": node.label,
        "title": node.title,
        "level": node.level  # 新增
    }
    next_id += 1

    if node.children:
        item["children"] = [node_to_dict(child) for child in node.children]

    return item

treeData = [node_to_dict(chap) for chap in root.children]

# === 保存 ===
with open(OUTPUT_JSON_PATH, 'w', encoding='utf-8') as f:
    json.dump(treeData, f, ensure_ascii=False, indent=2)

print(f"✅ 树形结构已生成，共 {len(treeData)} 个章节点")
print(f"📄 输出文件: {OUTPUT_JSON_PATH}")