解决AI图文解析偏差：CodeBuddy多模型交叉校验+腾讯地图Skill经纬度定位实战

W020260707346693894688.jpg
W020260707346694155028.jpg
W020260707346694416153.jpg
W020260707346694698336.jpg
W020260707346695046292.jpg
W020260707346695405020.jpg
W020260707346695679472.jpg
W020260707346695939204.jpg
W020260707346696188612.jpg
W020260707346696436334.jpg
W020260707346696673596.jpg 仔细对照图片内容和changsha_schools_sports_2026.json
 确认解析出来的信息是准确的，要求开放场地不要解析错误，学校名称不要解析错误。

import json
data = json.load(open('changsha_schools_sports_2026.json', 'r', encoding='utf-8'))
def print_range(start, end):
    for d in data[start-1:end]:
        no = d['No.']
        district = d['district']
        school = d['schoolName']
        venues = ', '.join(d['venues'])
        start_date = d['startDate']
        end_date = d['endDate']
        print(f"{no:>3} | {district:>6} | {school:<35} | {venues:<50} | {start_date} ~ {end_date}")

# 1-40
print("="*120)
print("序号 1-40 (市直+雨花区)")
print("="*120)
print_range(1, 40)
print()

# 41-80
print("="*120)
print("序号 41-80 (雨花区+开福区)")
print("="*120)
print_range(41, 80)
print()

# 81-120
print("="*120)
print("序号 81-120 (浏阳市)")
print("="*120)
print_range(81, 120)
print()

# 121-160
print("="*120)
print("序号 121-160 (浏阳市)")
print("="*120)
print_range(121, 160)
print()

# 161-200
print("="*120)
print("序号 161-200 (浏阳市)")
print("="*120)
print_range(161, 200)
print()

# 201-240
print("="*120)
print("序号 201-240 (浏阳市)")
print("="*120)
print_range(201, 240)
print()

# 241-280
print("="*120)
print("序号 241-280 (浏阳市+长沙县+湘江新区)")
print("="*120)
print_range(241, 280)
print()

# 281-320
print("="*120)
print("序号 281-320 (湘江新区)")
print("="*120)
print_range(281, 320)
print()

# 321-360
print("="*120)
print("序号 321-360 (湘江新区+宁乡市)")
print("="*120)
print_range(321, 360)
print()

# 361-400
print("="*120)
print("序号 361-400 (宁乡市+天心区)")
print("="*120)
print_range(361, 400)
print()

# 401-455
print("="*120)
print("序号 401-455 (天心区+望城区+芙蓉区)")
print("="*120)
print_range(401, 455)

import json
data = json.load(open('changsha_schools_sports_2026.json', 'r', encoding='utf-8'))
# Check items with special dates or venues
for idx in [289, 292, 332, 334, 340, 392, 403, 408, 413, 430, 440, 448]:
    d = data[idx]
    print(f"No.{d['No.']}: {d['schoolName']}")
    print(f"  venues: {d['venues']}")
    print(f"  dates: {d['startDate']} ~ {d['endDate']}")
    print()
# Check items with 2+ venues (potential splitting issues)
for idx in [271, 272, 273, 274, 275, 276, 277, 278, 279, 280, 281, 287, 288, 296, 297, 298, 299, 300, 301, 302, 303, 304, 305, 306, 307, 308, 309, 310, 312, 313, 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]:
    d = data[idx]
    if len(d['venues']) >= 3:
        print(f"No.{d['No.']}: {d['schoolName']}")
        print(f"  venues: {d['venues']}")
        print(f"  dates: {d['startDate']} ~ {d['endDate']}")
        print()

{
  "mcpServers": {
    "tencent-map": {
      "url": "https://mcp.map.qq.com/sse?key=你的腾讯地图Web端Key"
    }
  }
}

@skill:tencentmap-lbs-skill 请调用腾讯地图的服务，
将changsha_schools_sports_2026.json 中的schoolName进行位置检索，
返回对应的经纬度值，并添加到对应的json中。

import urllib.request
import urllib.parse
import json
import time
import os

KEY = '你的腾讯地图Web端Key'
INPUT_FILE = 'changsha_schools_sports_2026.json'
OUTPUT_FILE = 'changsha_schools_sports_2026.json'

def geocode(address):
    """调用腾讯地图地理编码API，返回(lng, lat)或(None, None)"""
    url = f'https://apis.map.qq.com/ws/geocoder/v1?address={urllib.parse.quote(address)}&key={KEY}'
    try:
        req = urllib.request.Request(url, headers={'User-Agent': 'Mozilla/5.0'})
        resp = urllib.request.urlopen(req, timeout=15).read().decode('utf-8')
        data = json.loads(resp)
        if data.get('status') == 0:
            loc = data.get('result', {}).get('location')
            if loc and 'lng' in loc and 'lat' in loc:
                return float(loc['lng']), float(loc['lat'])
        return None, None
    except Exception as e:
        return None, None

def clean_name(name):
    """去掉括号及其中内容，用于备选查询"""
    import re
    return re.sub(r'[（(].*?[）)]', '', name).strip()

def main():
    with open(INPUT_FILE, 'r', encoding='utf-8') as f:
        data = json.load(f)

    success = 0
    fail = 0
    failed_items = []

    total = len(data)
    for idx, item in enumerate(data):
        no = item.get('No.', idx + 1)
        name = item.get('schoolName', '')
        if not name:
            continue

        # 优先用原始名称查询
        lng, lat = geocode(name)

        # 如果失败，尝试去掉括号内容
        if lng is None:
            clean = clean_name(name)
            if clean and clean != name:
                lng, lat = geocode(clean)

        # 如果还失败，尝试加"长沙市"前缀（针对区县级学校）
        if lng is None and not name.startswith('长沙市'):
            lng, lat = geocode('长沙市' + name)

        if lng is not None:
            item['longitude'] = lng
            item['latitude'] = lat
            success += 1
        else:
            fail += 1
            failed_items.append((no, name))

        # 每10条打印一次进度
        if (idx + 1) % 10 == 0 or idx == total - 1:
            print(f'进度: {idx + 1}/{total}, 成功: {success}, 失败: {fail}')

        # 控制请求频率，避免限流
        time.sleep(0.15)

    # 写回JSON
    with open(OUTPUT_FILE, 'w', encoding='utf-8') as f:
        json.dump(data, f, ensure_ascii=False, indent=2)

    print(f'\n处理完成！总计: {total}, 成功: {success}, 失败: {fail}')
    if failed_items:
        print('\n失败的条目:')
        for no, name in failed_items:
            print(f'  No.{no}: {name}')

if __name__ == '__main__':
    main()