群晖NAS使用Docker部署WordPress

搭建网站的工具和过程降到低门槛和低成本，WordPress使用小型文件数据库SQLite组建Sandbox系统模拟

朗读全文

有什么用

创建工作室/公司网站, 网店; 发布文章, 分享经验等

SEO友好

展示: https://n.carlzeng.com

怎么用

version: '3'
services:
  wordpress:
    image: soulteary/sqlite-wordpress
    restart: always
    ports:
      - 8104:80
    volumes:
      - ./:/var/www/html
      #用于对接兰空图床批量上传的图库, CSV中含文件名对应到产品
      - /volume2/KingchuxingSSD512G/MacBookPro_Skitch/storage/app/uploads/2026/05:/var/www/html/wp-content/uploads/lskyupload202605

后来验证: 这种方式最后用于sandbox的建设, 因为网店再导入商品和图片后下单会报错, 切换使用官方版本后解决; 用这种轻量级的sandbox好处有: 随便安装插件等, 不怕污染了主数据库(数据的清理干净很麻烦, 耗时, 易出错)

演示/实践 https://w.carlzeng.com

相关内容

WooCommerce导入产品

安装插件: WebToffee Import Export

如何导入产品照片?

需求: 导入所有几千张照片, 然后关联到产出

WP的SQLlite数据库文件在哪里?

WordPress 用 SQLite 时（官方插件 sqlite-database-integration 或 wp-sqlite-db），数据库文件默认在：

wp-content/database/.ht.sqlite

最后图片还是要使用WP原始的图库功能;

上传到图库, 但由于兰空图床的目录使用的docker的映射方式, 被我直接映射到docker目录

/volume2/KingchuxingSSD512G/MacBookPro_Skitch/storage/app/uploads/2026/05

这样; Addd From Server Reloaded 就能加载到这些1170+张照片了

导入后: 图库

故障: 很多可以在兰空中看到的照片, 无法访问, 不知为何(是否和docker wordpress映射 并上传后有关? )
https://img.carlzeng.com:3/i/2026/05/O1CN01bBq8Wx1smZl0yPj7B_!!4611686018427385985-53-fleamarket.jpg

O1CN01fP0ICM1smZqe9QMrt_!!4611686018427385985-53-xy_item.jpg

成功链接到使用免费版的WebToffee Import Export; 导入所有产品和相关联的图片(先导入到图库中)

照片可以导入格式:
/wp-content/uploads/O1CN013Dpd011smZsfIaiIb_4611686018427385985-53-xy_item.jpg

可以导入多张照片, 用 , 空格分开

-5:09pm- 用,分隔即可

    修改全部的csv文件, 转化
    | 为 ,
    !! 去除
    格式如上

优化docker-compose 的目录映射 添加:ro, 确保兰空图床中的文件, 不要受影响/删除

导入详细商品目录

用DBeaver, 导入wp_posts表数据, 和 wp_postmeta表数据

绝大部分数据成功/轻松导入.

发现问题:

有的货品的SKU和价格丢失

修复步骤:

1. 查询到sku 或 price为空的行, 一共60个; 复制出p.ID

SELECT 
    p.ID as product_id, 
    p.post_title as product_name,
    sku.meta_value as sku,
    price.meta_value as price
FROM 
    wp_posts p
LEFT JOIN 
    wp_postmeta sku ON p.ID = sku.post_id AND sku.meta_key = '_sku'
LEFT JOIN 
    wp_postmeta price ON p.ID = price.post_id AND price.meta_key = '_price'
WHERE 
    p.post_type = 'product' 
    AND p.post_status = 'publish'
    AND (
        sku.meta_value IS NULL OR sku.meta_value = ''
        OR price.meta_value IS NULL OR price.meta_value = ''
    );

2. 到Sandbox正常的数据库中提取这些wp_postmeta相关的行数据

SELECT 
    post_id,
    meta_key,
    meta_value
FROM 
    wp_postmeta
WHERE 
    post_id IN (9, 240, 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 254, 255, 256, 257, 258, 259, 260, 281, 282, 283, 284, 285, 286, 287, 288, 289, 290, 291, 292, 293, 294, 295, 296, 297, 298, 299, 300, 321, 322, 323, 324, 325, 326, 327, 328, 329, 330, 331, 332, 333, 334, 335, 336, 337, 338)

3. 使用DBeaver导出这些数据, 然后目标选择到MariaDB的wp_postmeta 表,

   注意: 不要输出meta_id, 列, 不然会存在PK冲突的问题, 导致数据无法正常导入

   导入最后一步选择 Insert IGNORE

至此60个有显示故障的商品wp_postmeta被重新导入修复完毕!

下一步: 继续设置商品的分类和品牌, 标签等

解锁媒体库目前上传限制2MB

媒体库目前上传限制时2MB, 如何改成4M?

docker-compose中添加

WORDPRESS_CONFIG_EXTRA: |
 @ini_set( 'upload_max_filesize', '4M' );  
 @ini_set( 'post_max_size', '4M' );
 @ini_set( 'max_execution_time', '300' );

无法解决问题, 媒体上传限制还是2M

@ini_set( 'upload_max_size' , '4M' );
@ini_set( 'post_max_size', '4M');

解决办法: 新建一个插件

<?php
/* Plugin Name: Increase Upload Limit */

add_filter( 'upload_size_limit', 'b5f_increase_upload' );

function b5f_increase_upload( $bytes )
{
    return 4194304; // 4m; 5m=5*1024*1024
}

把这个文件保存, 然后压缩成.zip文件; 比如: increase-upload-limit.zip

然后在安装插件页面上传, 启用, 解决!

而且编辑是动态的, 也就是说 想要修改这个限制; 直接ssh过去服务器中, 修改这个值; 立即生效, 简单直接.

在wordpress 根目录 创建一个新的php.ini文件

使用代码或文本编辑器，将以下代码添加到现有或新的php.ini文件中：

upload_max_filesize = 5M 
post_max_size = 5M 
memory_limit = 128M

本小节感谢: WordPress上传限制,最大上传文件大小,最大上传限制 文字教程

修改商店页面标题:

同时也添加到了vi increase-upload-limit.php 这个文件中

add_filter( 'document_title_parts', function( $parts ) {
    // 商店主页、产品分类、产品标签
    if ( is_shop() || is_product_category() || is_product_tag() ) {
        // 把原标题前面加上「案例」
        if ( ! empty( $parts['title'] ) ) {
            $parts['title'] = '案例 - ' . $parts['title'];
        } else {
            // 兜底：万一没拿到，直接给「案例」
            $parts['title'] = '案例';
        }
    }
    return $parts;
}, 999 );

如何批量导入MD博客文章

插件: Import Markdown 似乎只能单独, 单个文件的导入?

各种研究, 最后在一个博主的站点上找到了解决方案, 用一个python脚本 一键批量上传目录中441篇文章到WordPress

简直太高效了, 分享源码:

import os  # 用来遍历文件路径
import sys
import ssl
ssl._create_default_https_context = ssl._create_unverified_context


"""
需要安装如下pip包
pip3 install python-frontmatter
pip3 install markdown
pip install 'Markdown'
pip3 install python-wordpress-xmlrpc
"""
# 1 导入frontmatter模块
import frontmatter

# 2 导入markdown模块
import markdown

# 3 导入wordpress_xmlrpc模块
from wordpress_xmlrpc import Client, WordPressPost
from wordpress_xmlrpc.methods.posts import NewPost


def make_post(filepath, metadata):
    """
    make a WordPressPost for Client call
    :param filepath: 要发布的文件路径
    :param metadata: 字典类型
             包括 metadata['category']: 文章分类
                  metadata['tag']: 文章标签
                  metadata['status']: 有publish发布、draft草稿、private隐私状态可选
    :return WordPressPost: if success
            None: if failure
    """
    filename = os.path.basename(filepath)  # 例如：test(2021.11.19).md
    filename_suffix = filename.split('.')[-1]  # 例如：md
    filename_prefix = filename.replace('.' + filename_suffix, '')  # 例如：test(2021.11.19)；注意：这种替换方法要求文件名中只有一个".md"

    # 目前只支持 .md 后缀的文件
    if filename_suffix != 'md':
        return None

    # 1 通过frontmatter.load函数加载读取文档里的信息，包括元数据
    post_from_file = frontmatter.load(filepath)

    # 2 markdown库导入内容
    post_content_html = markdown.markdown(post_from_file.content, extensions=['markdown.extensions.fenced_code'])
    post_content_html = post_content_html.encode("utf-8")

    # 3 将本地post的元数据暂存到metadata中
    metadata['title'] = filename_prefix  # 将文件名去掉.md后缀，作为标题
    # metadata['slug'] = metadata['title']  # 别名
    metadata_keys = metadata.keys()
    # 如果post_from_file.metadata中的属性key存在，那么就将metadata[key]替换为它
    for key in metadata_keys:
        if key in post_from_file.metadata:  # 若md文件中没有元数据'category'，则无法调用post.metadata['category']
            metadata[key] = post_from_file.metadata[key]

    # 4 将metadata中的属性赋值给post的对应属性
    post = WordPressPost()  # 要返回的post
    post.content = post_content_html
    post.title = metadata['title']
    # post.slug = metadata['slug']
    post.post_status = metadata['status']
    post.terms_names = {
        'category': metadata['category'],
        'post_tag': metadata['tag']
    }
    post.comment_status = 'open'  # 开启评论
    return post


def push_post(post, client):
    """
    上传post到WordPress网站
    :param post: 要发布的文章（WordPressPost类型），由make_post函数得到
    :param client: 客户端
    :return True: if success
    """
    return client.call(NewPost(post))


def get_filepaths(path):
    """
    如果path是目录路径，递归遍历path目录下的所有文件，将所有文件路径存入filepaths
    如果path是文件路径，直接将单个文件路径存入filepaths
    :param path: 你要上传的目录路径或文件路径（绝对路径）
    :return filepaths: 该目录下的所有子文件或单个文件的绝对路径
            None: wrong path
    """
    filepaths = []
    if os.path.isdir(path):  # 当前路径是目录
        for now_dirpath, child_dirnames, child_filenames in os.walk(path):
            # ========== 新增：排除包含 secret 的目录 ==========
            if 'secret' in now_dirpath.lower():  # 不区分大小写：Secret / SECRET 都排除
                continue
                
            for filename in child_filenames:
                filepath = os.path.join(now_dirpath, filename)
                filepaths.append(filepath)
        filepaths.sort()  # <-- 按英文字母 / 数字顺序排序（就加这一行）
        return filepaths
    elif os.path.isfile(path):  # 当前路径是文件
        return [path]
    else:  # wrong path
        return None


if __name__ == '__main__':
    # User Configuration
    path = '/Users/carlzeng/App/blog/source/_posts'  # e.g. D:/PythonCode/post-wordpress-with-markdown/doc
    domain_list = [
        'https://wp.carlzeng.com:3/'
    ]  # e.g. https://example.com（配置了SSL证书就用https，否则用http）
    username = '*******'
    password = '*******'

    for domain in domain_list:
        # Optional Configuration
        post_metadata = {
            'category': ['未分类'],  # 文章分类
            'tag': ['手柄一修哥'],  # 文章标签
            'status': 'publish'  # 可选publish发布、draft草稿、private隐私状态
        }

        # Start Work
        print('----------------------------------------------START----------------------------------------------')
        filepaths = get_filepaths(path)
        if filepaths is None:
            print('FAILURE: wrong path')
            sys.exit(1)

        client = Client(domain + '/xmlrpc.php', username, password)  # 客户端

        md_cnt = 0
        all_cnt = len(filepaths)
        process_number = 0
        failpaths = []  # 存储上传失败的文件路径
        for filepath in filepaths:
            process_number = process_number + 1
            filename = os.path.basename(filepath)
            try:
                # 把可能报错的逻辑全部放进 try
                post = make_post(filepath, post_metadata)
                
                if post is not None:
                    push_post(post, client)
                    md_cnt = md_cnt + 1
                    print('Process number: %d/%d  SUCCESS: Push "%s" completed!' % (process_number, all_cnt, filename))
                else:
                    failpaths.append(filepath)
                    print('Process number: %d/%d  WARNING: Can\'t push "%s" because it\'s not Markdown file.' % (
                    process_number, all_cnt, filename))
            except Exception as e:
                # 捕获解析/发布失败，不中断程序
                failpaths.append(filepath)
                print('Process number: %d/%d  ERROR: Failed to push "%s" | Reason: %s' % (process_number, all_cnt, filename, str(e)))

        print('-----------------------------------------------END-----------------------------------------------')
        print('SUCCESS: %d files have been pushed to {0}.'.format(domain) % md_cnt)

        if len(failpaths) > 0:
            print('WARNING: %d files haven\'t been pushed to {0}.'.format(domain) % len(failpaths))
            print('\nFailure to push these file paths:')
            for failpath in failpaths:
                print(failpath)

感谢作者: 一键上传markdown 格式文章到wordpress

修改的点有:

1. 第137行 post = make_post(filepath, post_metadata) 这个需要try catch
2. 添加排序功能: 目前加载目录下的文件没有排序, 请按文件的英文字母顺序排序
3. 请排除目录名中包含secret的目录

运行后输出:

----------------------------------------------START----------------------------------------------
Process number: 8/418  SUCCESS: Push "1099-MISC是什么-CarlZeng.md" completed!
Process number: 9/418  SUCCESS: Push "A-feature-in-Netsuite-Reports-Financial-Balance-Sheet-CarlZeng.md" completed!
Process number: 10/418  SUCCESS: Push "AI脸部识别，人脸比对，人脸识别-CarlZeng.md" completed!
Process number: 11/418  SUCCESS: Push "AJAX提交数据时-中文处理-以及js-url-中文处理-CarlZeng.md" completed!
Process number: 12/418  SUCCESS: Push "Bar-codes-in-NetSuite-Saved-Searches-transport-reprint-活用NetSuite搜索页面显示可扫描条码-CarlZeng.md" completed!

...

Process number: 417/418  SUCCESS: Push "高手写出的是天使，而新手写的，可能是魔鬼！-Javascript这样的脚本语言，由于太灵活-CarlZeng.md" completed!
Process number: 418/418  SUCCESS: Push "高效的服务端生成QRCODE二维码方案-Docker搭建.md" completed!
-----------------------------------------------END-----------------------------------------------
SUCCESS: 411 files have been pushed to https://wp.carlzeng.com:3/.
WARNING: 7 files haven't been pushed to https://wp.carlzeng.com:3/.

Failure to push these file paths:
/Users/carlzeng/App/blog/source/_posts/.DS_Store

遇到的bug/enhancement:

1. 手动完成文章所有分类,

2. 标签和照片无法导入; (用python, 等待更新, 提取这些信息到WP)

3. 下一步:

Woocommerce下单错误

点击下单报错:

Uncaught Error: Call to a member function get_total() on false in /var/www/html/wp-content/plugins/woocommerce/includes/gateways/cod/class-wc-gateway-cod.php:314\nStack trace:\n#0 /var/www/html/wp-content/plugins/woocommerce/src/StoreApi/Legacy.php(56): WC_Gateway_COD->process_payment(4131)\n#1 /var/www/html/wp-includes/class-wp-hook.php(341): Automattic\\WooCommerce\\StoreApi\\Legacy->process_legacy_payment(Object(Automattic\\WooCommerce\\StoreApi\\Payments\\PaymentContext), Object(Automattic\\WooCommerce…wp-includes/class-wp-hook.php(365): WP_Hook->apply_filters(NULL, Array)\n

wp_woocommerce_order_items 等等, 这些表都有的

为了解决woocommerce的故障, 需要让wordpress使用官方的docker image, 切换数据库(mssql, mariadb)方案

从另外一个wordpress数据库导入文章

通过工具链接到MariaDB, 将查询导出为CSV; 然后导入到在线的文件型数据库中: .ht.sqlite

SELECT p.* 
FROM wp_posts p
JOIN wp_term_relationships tr ON p.ID = tr.object_id
JOIN wp_term_taxonomy tt ON tr.term_taxonomy_id = tt.term_taxonomy_id
WHERE 
  tt.term_id = 22          -- 分类ID 22
  AND tt.taxonomy = 'category'  -- 只查文章分类，不查标签/自定义分类
  AND p.post_type = 'post'      -- 只查文章
  AND p.post_status = 'publish' -- 只查已发布
ORDER BY p.post_date DESC;

继续在新的系统中编辑这些文章

修改域名, 和导入的文章URL需要修改

-- 3. 文章/页面里的URL（含GUID）
SELECT 'wp_posts' AS src, CONCAT(ID,':',post_title) AS name, guid 
FROM wp_posts 
WHERE guid LIKE 'http%'

UNION ALL

-- 4. 文章自定义字段里的URL
SELECT 'wp_postmeta' AS src, meta_key, meta_value 
FROM wp_postmeta 
WHERE meta_value LIKE 'http%';

UPDATE wp_posts 
SET post_content=REPLACE(post_content,'http://192.168.6.203:8108','https://netsuite.carlzeng.com'),
    guid=REPLACE(guid,'http://192.168.6.203:8108','https://netsuite.carlzeng.com');
    
UPDATE wp_posts 
SET post_content=REPLACE(post_content,'https://w.carlzeng.com','https://netsuite.carlzeng.com'),
    guid=REPLACE(guid,'https://w.carlzeng.com','https://netsuite.carlzeng.com');
    
--手动更新了一行 
-- 4. 文章自定义字段里的URL
SELECT 'wp_postmeta' AS src, meta_key, meta_value, * 
FROM wp_postmeta 
WHERE meta_value LIKE 'http%';

如何修复错误 “Too Many Redirects” WordPress 报错 (ERR_TOO_MANY_REDIRECTS)

ERR_TOO_MANY_REDIRECTS

使用本地部署后, 局域网可访问, 如: http://192.168.6.* 那么NPM反代以后, 使用 https://wp.yourdomain.com:3/wp-admin可正常访问后台; 下一步: 如何让远程的CDN直接使用无需携带端口的访问方式?

1. 首先要去处理的是本地访问

2. 然后反向代理, 测试反向代理后的后台访问是否正常, css都正常, 登录也正常

3. 确保CDN后者cloudflare中的映射是通的; 具体详见“博客迁移说明” CDN章节

//define(‘WP_HOME’,’https://w.carlzeng.com‘);
define(‘WP_SITEURL’,’https://w.carlzeng.com‘);

#这个用来应急时调整, 获得后台的访问权限; 主要是要先把CDN那边端口映射先做好, 映射到本地的NPM 如: *.yourdomain.com:4443(这个端口目前只有后台的访问权限/wp-admin

切换到官方Docker Image

Production生成环境

mkdir wordpresshtml && mkdir db 
vi docker-compose.yaml

services:

  wordpress:
    image: wordpress
    restart: always
    ports:
      - 8106:80
    environment:
      WORDPRESS_DB_HOST: db
      WORDPRESS_DB_USER: example user
      WORDPRESS_DB_PASSWORD: example pass
      WORDPRESS_DB_NAME: wordpress
    volumes:
      - ./wordpresshtml:/var/www/html

  db:
    image: mysql:8.0
    restart: always
    environment:
      MYSQL_DATABASE: wordpress
      MYSQL_USER: example user
      MYSQL_PASSWORD: example pass
      MYSQL_RANDOM_ROOT_PASSWORD: '1'
    volumes:
      - ./db:/var/lib/mysql

volumes:
  wordpress:
  db:

chown mysql /volume2/KingchuxingSSD512G/docker/compose/wordpress2/db -R

chmod +777 /volume2/KingchuxingSSD512G/docker/compose/wordpress2/db -R

chown mysql /volume2/KingchuxingSSD512G/docker/compose/wordpress2/wordpresshtml -R

chmod +777 /volume2/KingchuxingSSD512G/docker/compose/wordpress2/wordpresshtml -R

必须要运行这个目录权限, 否者就是各种报错.

woocommerce Connection Information

upload a WooCommerce .zip file and install the plugin on the local host. Every time I click install button, WordPress is asking for connection for FTP

Add this line in wp-config.php

define( 'FS_METHOD', 'direct' );

//define('WP_HOME','https://w.carlzeng.com');  
define('WP_SITEURL','https://w.carlzeng.com');                                             #这个用来应急时调整, 获得后台的访问权限; 主要是要先把CDN那边端口映射先做好, 映射到本地的NPM 如: *.yourdomain.com:4443(这个端口目前只有后台的访问权限/wp-admin                                 
                                                                                                                            
define('WP_CONTENT_URL', '/wp-content');  

还是决定切换到MariaDB

services:

  wordpress:
    image: wordpress
    restart: always
    ports:
      - 8106:80
    environment:
      WORDPRESS_DB_HOST: db
      WORDPRESS_DB_USER: example user
      WORDPRESS_DB_PASSWORD: example pass
      WORDPRESS_DB_NAME: wordpress
    volumes:
      - ./wordpresshtml:/var/www/html

  db:
    image: mariadb:jammy
    restart: always
    ports:
      - 8107:3306
    environment:
      MYSQL_DATABASE: wordpress
      MYSQL_USER: example user
      MYSQL_PASSWORD: example pass
      MYSQL_RANDOM_ROOT_PASSWORD: '1'
    volumes:
      - ./db:/var/lib/mysql

volumes:
  wordpress:
  db:

初始化安装好wordpress, 只需要这些表:

1. 手工导入wp_posts (直接从数据库到数据库)

2. 然后用csv导入wp_postmeta

裁剪您的图片时发生错误。

原来是操作不了PNG; JPG可以的, 无所谓, 无伤大雅

SQL修复文章分类

wp_term_taxonomy 这个基础表维护的是: 分类ID和分类名称; 比如: 产品分类、文章分类、产品类型、模版类型部分区域

wp_term_relationships 这个表保存文章/产品 与分类的关系; 比如: 文章#1属于类名#1

步骤一: 插入wp_term_relationships

INSERT IGNORE INTO wp_term_relationships (object_id, term_taxonomy_id, term_order)

select ID, 22, 0 from wp_posts where post_type = 'post' and post_content like '%netsuite%' or post_title like '%netsuite'

note: 用IGNORE支持多次执行, 会避开重复的行, 输出

Duplicate entry '205-22' for key 'PRIMARY'
Duplicate entry '3298-22' for key 'PRIMARY'
Duplicate entry '3306-22' for key 'PRIMARY'
Duplicate entry '3312-22' for key 'PRIMARY'
Duplicate entry '3640-22' for key 'PRIMARY'
Duplicate entry '3662-22' for key 'PRIMARY'

步骤二: 插入wp_term_taxonomy

#关键后续：更新 WordPress 分类计数器
UPDATE wp_term_taxonomy 
SET count = (
    SELECT COUNT(*) 
    FROM wp_term_relationships 
    WHERE wp_term_relationships.term_taxonomy_id = wp_term_taxonomy.term_taxonomy_id
)
WHERE term_taxonomy_id = 22

添加剩余的文章都分类到: 一修哥工作室

INSERT IGNORE INTO wp_term_relationships (object_id, term_taxonomy_id, term_order)
select ID, 23, 0 from wp_posts where post_type = 'post' and ID not in (
select object_id from wp_term_relationships
)

UPDATE wp_term_taxonomy 
SET count = (
    SELECT COUNT(*) 
    FROM wp_term_relationships 
    WHERE wp_term_relationships.term_taxonomy_id = wp_term_taxonomy.term_taxonomy_id
)
WHERE term_taxonomy_id = 23

导入部分硬件工作记录/成功案例

以准备好成功案例的md聊天记录导出, 导入展现. 此md来源为: 闲鱼APP中聊天记录导出

批量, 加载目录下md文件时, 需要按md文件的创建时间顺序来加载

Sanbox测试得知: 待修复 视频,如何支持在线点击查看?
md文件中如何显示视频播放器?

import os
import re

# 匹配 Markdown 格式的 [视频](url地址)
pattern = r'\[\s*视频\s*\]\(\s*(https?://[^\s\)]+|[^\s\)]+\.mp4|[^\s\)]+)\s*\)'


def get_file_creation_time(file_path):
    """兼容多系统的获取文件创建时间函数"""
    stat = os.stat(file_path)
    try:
        # macOS / FreeBSD 优先使用 birthtime
        return stat.st_birthtime
    except AttributeError:
        # Windows 上 ctime 即为创建时间；Linux 上为元数据修改时间
        return stat.st_ctime


def convert_md_files_by_order():
    current_dir = os.getcwd()

    # 1. 过滤出当前目录下所有的 .md 文件
    md_files = [f for f in os.listdir(current_dir) if f.endswith(".md")]

    # 2. 【核心修改】将文件列表按照创建时间进行“升序排序”（从旧到新）
    # 如果想改为降序（最新创建的先处理），可以加上 reverse=True
    md_files.sort(key=lambda x: get_file_creation_time(os.path.join(current_dir, x)))

    success_count = 0
    print("开始按创建时间【升序】处理文件...\n" + "-" * 40)

    # 3. 按排序后的顺序遍历文件
    for filename in md_files:
        file_path = os.path.join(current_dir, filename)

        with open(file_path, "r", encoding="utf-8") as f:
            content = f.read()

        if re.search(pattern, content):
            # 替换并强制添加前后空行激活 HTML 渲染
            replaced_content = re.sub(
                pattern,
                r"\n\n<video src='\1' controls width='100%' height='auto' loop='loop'></video>\n\n",
                content,
            )

            with open(file_path, "w", encoding="utf-8") as f:
                f.write(replaced_content)

            # 获取文件创建时间用于终端打印展示
            c_time = get_file_creation_time(file_path)
            from datetime import datetime

            time_str = datetime.fromtimestamp(c_time).strftime(
                "%Y-%m-%d %H:%M:%S"
            )

            print(f"[{time_str}] 成功转换: {filename}")
            success_count += 1

    print("-" * 40 + f"\n处理完成！共依序转换了 {success_count} 个文件。")


if __name__ == "__main__":
    convert_md_files_by_order()

----------------------------------------------START----------------------------------          
Process number: 1/30  SUCCESS: Push "聊天记录_tbNick_n9c42_2026-05-17.md" completed!
....

对隐私/敏感信息的屏蔽保护

SELECT ID, post_title, post_date 
FROM wp_posts 
WHERE post_content LIKE '%13261977480%';

-- 1. 替换文章正文内容（最核心）
UPDATE wp_posts 
SET post_content = REPLACE(post_content, '13261977480', '1326****480')
WHERE post_content LIKE '%13261977480%';

UPDATE wp_posts 
SET post_content = REPLACE(post_content, '后北营家园一区10号楼3-130', '后北营家园一区10号(大运河果蔬便民店外)')
WHERE post_content LIKE '%后北营家园一区10号楼3-130%';

UPDATE wp_posts 
SET post_content = REPLACE(post_content, 'VPN', 'V*P*N')
WHERE post_content LIKE '%VPN%';

-- 替换文章标题
UPDATE wp_posts 
SET post_title = REPLACE(post_title, '_2026-05-17', '_手柄一修哥')
WHERE post_title LIKE '%_2026-05-17%';

维护商品分类

INSERT IGNORE INTO wp_term_relationships (object_id, term_taxonomy_id, term_order)
SELECT ID, 17, 0
FROM wp_posts 
WHERE post_type = 'product' 
AND post_title like '%PS5%'

and ID not in (
select object_id from wp_term_relationships)

INSERT IGNORE INTO wp_term_relationships (object_id, term_taxonomy_id, term_order)
SELECT ID, 27, 0
FROM wp_posts 
WHERE post_type = 'product' 
AND post_title like '%switch%' 


INSERT IGNORE INTO wp_term_relationships (object_id, term_taxonomy_id, term_order)
SELECT ID, 27, 0
FROM wp_posts 
WHERE post_type = 'product' 
AND post_title like '%joycon%' 

#totally 43 row, success 23, duplicated

Updated rows 23; 
Duplicate entry '242-27' for key 'PRIMARY'
Duplicate entry '432-27' for key 'PRIMARY'
Duplicate entry '443-27' for key 'PRIMARY'
Duplicate entry '455-27' for key 'PRIMARY'
Duplicate entry '487-27' for key 'PRIMARY'
Duplicate entry '506-27' for key 'PRIMARY'
Duplicate entry '259-27' for key 'PRIMARY'
Duplicate entry '293-27' for key 'PRIMARY'
Duplicate entry '323-27' for key 'PRIMARY'
Duplicate entry '578-27' for key 'PRIMARY'
Duplicate entry '580-27' for key 'PRIMARY'
Duplicate entry '606-27' for key 'PRIMARY'
Duplicate entry '608-27' for key 'PRIMARY'
Duplicate entry '615-27' for key 'PRIMARY'
Duplicate entry '631-27' for key 'PRIMARY'
Duplicate entry '652-27' for key 'PRIMARY'
Duplicate entry '659-27' for key 'PRIMARY'
Duplicate entry '682-27' for key 'PRIMARY'
Duplicate entry '684-27' for key 'PRIMARY'
Duplicate entry '695-27' for key 'PRIMARY'

​ 17 ps5 93
​ 27 Switch 133 joycon; duplicate lite, oled, 大气层

​ 28 Xbox 35
​ 29 SteamDeck 27
​ 30 Quest3 6
​ 31 PS4 6
​ 32 IT支持 10 远程 10 ; 上门 4

​ 33 NetSuite 1
​ 34 掌机 3

未分类: 24

SELECT ID, 27, 0
FROM wp_posts 
WHERE post_type = 'product' 
and  wp_posts.ID not in (
select  object_id from wp_term_relationships
)


#更新分类的统计表
UPDATE wp_term_taxonomy 
SET count = (
    SELECT COUNT(*) 
    FROM wp_term_relationships 
    WHERE wp_term_relationships.term_taxonomy_id = wp_term_taxonomy.term_taxonomy_id
)
WHERE term_taxonomy_id = 34

Next 搭建: 网站客服系统 docker

比如: https://zhuanlan.zhihu.com/p/23677870488

用于新增一个潜在客户直接语言沟通的机会, 直接在网站中聊天

快速上线的,但是无拓展功能(只能去付费拓展的)

https://www.tawk.to/software/live-chat/

https://www.twt.com/en/chat/pricing

https://github.com/chaskiq/chaskiq

https://github.com/chatwoot/chatwoot

version: '3.8'

services:
  # 内部轻量缓存（不可省略）
  chaskiq-redis:
    image: redis:7-alpine
    container_name: chaskiq-redis
    restart: always

  # Chaskiq 客服核心服务
  chaskiq-app:
    image: chaskiq/chaskiq:latest
    container_name: chaskiq-app
    restart: always
    ports:
      - "3000:3000"
    environment:
      - HOST=http://localhost:3000
      - SECRET_KEY_BASE=Another_very_long_random_string_64_chars
      - REDIS_URL=redis://chaskiq-redis:6379/1        DATABASE_URL=postgresql://umami:[email protected]:5434/chaskiq_prod
      
    # 首次运行前需手动进入容器执行 bundle exec rails db:schema:load 完成建表

方案一：正确配置 Chatwoot Community 纯开源版（解除限制）

version: '3.8'

services:
  cskefu-redis:
    image: redis:7-alpine
    container_name: cskefu-redis
    restart: always

  chatwoot-web:
    image: chatwoot/chatwoot:latest
    container_name: chatwoot-web
    restart: always
    ports:
      - "3000:3000"
    environment:
      - FRONTEND_URL=http://localhost:3000
      - SECRET_KEY_BASE=a_very_long_random_string_here_32_chars
      
      # ⚠️ 关键配置：关闭企业版限制，强制切换为无限制社区版
      - ENABLE_ACCOUNT_LIMITS=false
      - INSTALLATION_ENV=docker
      
      # 直连您已有的 Postgres 15 数据库
      - POSTGRES_HOST=172.17.0.1  # 宿主机桥接IP
      - POSTGRES_PORT=5432
      - POSTGRES_DATABASE=chatwoot_prod
      - POSTGRES_USERNAME=postgres
      - POSTGRES_PASSWORD=your_existing_postgres_password
      - REDIS_URL=redis://cskefu-redis:6379/0
    command: bundle exec rails s -p 3000 -b '0.0.0.0'
    depends_on:
      - cskefu-redis

  chatwoot-worker:
    image: chatwoot/chatwoot:latest
    container_name: chatwoot-worker
    restart: always
    environment:
      - ENABLE_ACCOUNT_LIMITS=false
      - INSTALLATION_ENV=docker
      - POSTGRES_HOST=172.17.0.1
      - POSTGRES_PORT=5432
      - POSTGRES_DATABASE=chatwoot_prod
      - POSTGRES_USERNAME=postgres
      - POSTGRES_PASSWORD=your_existing_postgres_password
      - REDIS_URL=redis://cskefu-redis:6379/0
    command: bundle exec sidekiq -g chatwoot -C config/sidekiq.yml
    depends_on:
      - cskefu-redis

SEO优化, 尝试提交Google

下载 wordpress sitemap generator 用于生成网站地图

下载zip, 上传安装后, 自动生成了http://w.carlzeng.com/sitemap.xml

状态: Processing data, please check again in a day or so

待下周重新查阅一下Google收录情况

2. https://ziyuan.baidu.com/linksubmit/url

3. Bing

   

提交了网站地图 http://w.carlzeng.com/sitemap.xml

Google Page Indexing issue

这个很简单, 按链接提示, 把货品图片补上即可.

Missing field “hasMerchantReturnPolicy” (in “offers”)

This is a non-critical issue. Items with these issues are valid, but could be presented with more features or be optimized for more relevant queries

这个已经在商品的详细页面添加了退款和退货策略 还是无法通过校验. 请高人指点.

總結：WPSSO Merchant Return Policy Manager 外掛可以幫助你輕鬆管理Google Merchant 列表的任何數量的退貨政策，同時修復遺失”hasMerchantReturnPolicy“ 的警告。

页面加载速度优化

去除首页的woocommerce区块, 避免照片的3M载入

下面这段是最能减负的, 让css减轻了首页和非commerce页面的负担

/**
 * 禁用 WooCommerce 所有的 CSS 样式和 JS 脚本
 */
add_action( 'template_redirect', 'remove_woocommerce_styles_scripts', 999 );
function remove_woocommerce_styles_scripts() {
    if ( function_exists( 'is_woocommerce' ) ) {
        if ( ! is_woocommerce() && ! is_cart() && ! is_checkout() ) {
            remove_action('wp_enqueue_scripts', [WC_Frontend_Scripts::class, 'load_scripts']);
	    remove_action('wp_print_scripts', [WC_Frontend_Scripts::class, 'localize_printed_scripts'], 5);
	    remove_action('wp_print_footer_scripts', [WC_Frontend_Scripts::class, 'localize_printed_scripts'], 5);
        }
    }
}

解决问题: wordpress首页并没有放woocommerce组件 页面很多woocommerce的css

速度检测网站: https://tools.pingdom.com

用CDN把JS给缓存起来, 能有效的提升网站访问速度

优化n.carlzeng.com访问速度

20260620 优化了所有照片, 剪裁尺寸到500px - 1000px (宽度最大值)

但是cloudflare缓存了之前大照片的, 大文件;

TODO: 等待cloudflare刷新缓存, 看需要耗时多久?

半小时左右还是有问题的:

0621第二天了, 用tools.pingdom.com来查询整个网址 还有1.8MB; 照片还有257kb的, 说明CDN的缓存还在生效, 怎么刷新cloudflare的cdn?

操作步骤

1. 登录 dash.cloudflare.com，选中你的域名

2. 左侧菜单：Caching（缓存） → Configuration（配置）

3. 找到 Purge Cache 区域，点击 Custom Purge（自定义清除）

4. 可以purge特定的URL资源, 或者整个站点清除, 或者 全部清除.

开启「开发模式 Development Mode」临时禁用缓存

1. 域名首页 Overview 页面
2. 找到 Development Mode 开关打开（持续 3 小时）
3. 此时所有资源直接回源，不会走 CDN 缓存，用来验证源站图片是否真的替换成功

1:15pm 已经提交过了好几分钟, 重新用浏览器还是看到了旧的图片大小内容; Monitor 要多久才能更新缓存?

0622 又是一个; purge后第二天, clashflare扯淡的, 不会刷新不回purge(用域名的方式已尝试)
重新尝试使用 URL prefix的方式 purge n.carlzeng.com/wp-content; 再一次拭目以待

9am URL访问源站点, 肯定已经清除到最新的小图片了; 开了develpment mode 持续3小时

0623 等了一天, 依旧无法刷新缓存. 验证方法

curl -I https://n.carlzeng.com/wp-content/uploads/20
26/06/image-3.png

域名 Overview 页面点击「Pause Cloudflare on Site」，完全停用 CDN 测试图片

pause后马上测试, 依旧是无法生效的, 依旧是 HIT cf-cache-status

分享媒体库中的照片

chmod 644 media-images.php

/volume2/KingchuxingSSD512G/docker/compose/wordpress2/wordpresshtml# vi getimages.php 

<?php
$dir = __DIR__;
for ($i = 0; $i < 5; $i++) {
    if (file_exists($dir . '/wp-load.php')) {
        require_once($dir . '/wp-load.php');
        break;
    }
    $dir = dirname($dir);
}

// 获取媒体库所有图片
$images = get_posts([
    'post_type'      => 'attachment',
    'post_mime_type' => 'image',
    'post_status'    => 'inherit',
    'posts_per_page' => -1,
]);
?>

<!DOCTYPE html>
<html lang="zh-CN">
<head>
    <meta charset="UTF-8">
    <title>案例图片 - 手柄一修哥</title>
    <meta name="viewport" content="width=device-width, initial-scale=1">

    <!-- Cloudflare CDN: baguetteBox CSS -->
    <link rel="stylesheet"
          href="https://cdnjs.cloudflare.com/ajax/libs/baguettebox.js/1.11.1/baguetteBox.min.css">

    <style>
        *{margin:0;padding:0;box-sizing:border-box}
        body{background:#f5f5f5;padding:20px}
        .grid{display:grid;grid-template-columns:repeat(auto-fill,minmax(180px,1fr));gap:15px;margin-top:20px}
        .card{background:#fff;border-radius:8px;overflow:hidden}
        img{width:100%;height:140px;object-fit:cover;display:block;background:#f0f0f0}
        /* 懒加载淡入 */
        img.lazy{opacity:0;transition:opacity 0.3s}
        img.lazy.loaded{opacity:1}
            
        h3 a { text-decoration: none !important; }
    </style>
</head>
<body>
    <h3>✅ <a href="https://w.carlzeng.com" target='_blank'>案例图库：<?php echo count($images); ?> 张（手柄一修哥工作室w.carlzeng.com）</a></h3>

    <div class="grid" id="gallery">
        <?php foreach($images as $img):
            $img_url = wp_get_attachment_image_url($img->ID, 'medium');
            $full_url = wp_get_attachment_image_url($img->ID, 'full');
        ?>
        <div class="card">
            <a href="<?php echo $full_url; ?>" class="gallery-item">
                <img class="lazy" data-src="<?php echo $img_url; ?>" alt="">
            </a>
        </div>
        <?php endforeach; ?>
    </div>

    <script src="https://cdnjs.cloudflare.com/ajax/libs/vanilla-lazyload/17.1.0/lazyload.min.js"></script>
    <script src="https://cdnjs.cloudflare.com/ajax/libs/baguettebox.js/1.11.1/baguetteBox.min.js"></script>

    <script>
        // 懒加载初始化
        new LazyLoad({ elements_selector: ".lazy" });

        // 灯箱初始化
        baguetteBox.run('#gallery', {
            animation: 'fadeIn',
            noScrollbars: true
        });
    </script>
    <!-- Matomo -->
<script>
  var _paq = window._paq = window._paq || [];
  _paq.push(["setDocumentTitle", document.domain + "/" + document.title]);
  _paq.push(["setCookieDomain", "*.carlzeng.com"]);
  _paq.push(['trackPageView']);
  _paq.push(['enableLinkTracking']);
  (function() {
    var u="//statcounter.carlzeng.com/";
    _paq.push(['setTrackerUrl', u+'matomo.php']);
    _paq.push(['setSiteId', '1']);
    var d=document, g=d.createElement('script'), s=d.getElementsByTagName('script')[0];
    g.async=true; g.src=u+'matomo.js'; s.parentNode.insertBefore(g,s);
  })();
</script>
<noscript><p><img src="//statcounter.carlzeng.com/matomo.php?idsite=1&amp;rec=1" style="border:0;" alt="" /></p></noscript>
<!-- End Matomo Code -->
    <script defer="" src="https://umami.carlzeng.com:3/script.js" data-website-id="84261200-b106-4d11-8bc3-ad3b395a3f31"></script>
</body>
</html>

修正文章中的URL地址

由于切换使用CDN等缘故, 修改源文章中的链接

1. Mariadb中更新

select * from wp_posts where 
post_content LIKE '%https://query.carlzeng.com:3/%'

UPDATE wp_posts 
SET post_content=REPLACE(post_content,'https://query.carlzeng.com:3/','https://query.carlzeng.com/')

2. 下载sqllite的数据库, 同样在数据库中更新

灵感来源与项目计划

WordPress 告别 MySQL：Docker SQLite WordPress

我写了个从 Obsidian 自动发布文章到微信公众号的工具

(请输入关键字以后回车，检索内容涵盖全局站点)

最新文章

• 奥克掌机更换霍尔摇杆及校准
• DSE Shoulder R1/L1 and Bumper R2/L2 Omoron Switch Buttons
• wordpress转支付宝小程序
• 硬件PCB设计及编程实现学习项目pixl.js-模拟Amiibo
• NetSuite的gzip解压缩在接口开发中的应用
• 浅谈部署闲鱼自动回复管理系统xianyu-auto-reply
• NetSuite接口开发之JSON源数据发送实现gzip压缩
• ESXI之N5105倍控小主机体验及现实应用
• Steam客户端不识别PS5手柄不识别SwitchPro手柄不识别Xbox和国产手柄

个性化需求沟通 扫客服加V加群：