mirror of
https://github.com/unti-io/go-utils.git
synced 2025-12-24 12:57:50 +08:00
325 lines
9.3 KiB
Go
325 lines
9.3 KiB
Go
package utils
|
||
|
||
import (
|
||
"fmt"
|
||
HTML "html"
|
||
"net/url"
|
||
"regexp"
|
||
"strings"
|
||
"sync"
|
||
|
||
"github.com/spf13/cast"
|
||
)
|
||
|
||
// 读写锁 - 防止并发写入
|
||
type wrLock struct {
|
||
Lock *sync.RWMutex
|
||
Data map[string]any
|
||
}
|
||
|
||
// Get - 获取数据
|
||
func (this *wrLock) get(key string) any {
|
||
defer this.Lock.Unlock()
|
||
this.Lock.Lock()
|
||
return this.Data[key]
|
||
}
|
||
|
||
// Set - 设置数据
|
||
func (this *wrLock) set(key string, val any) {
|
||
defer this.Lock.Unlock()
|
||
this.Lock.Lock()
|
||
this.Data[key] = val
|
||
}
|
||
|
||
// Has - 判断是否存在
|
||
func (this *wrLock) has(key string) (ok bool) {
|
||
defer this.Lock.Unlock()
|
||
this.Lock.Lock()
|
||
_, ok = this.Data[key]
|
||
return ok
|
||
}
|
||
|
||
// Parse - 解析
|
||
var Parse *ParseClass
|
||
|
||
type ParseClass struct {}
|
||
|
||
// ParamsBefore - 解析参数
|
||
// 把 Content-Type = application/x-www-form-urlencoded 的参数解析成 object.deep.age = 10 的格式
|
||
func (this *ParseClass) ParamsBefore(params url.Values) (result map[string]any) {
|
||
|
||
wg := sync.WaitGroup{}
|
||
wr := wrLock{
|
||
Data: make(map[string]any, 3),
|
||
Lock: &sync.RWMutex{},
|
||
}
|
||
worker := func(keys string, value []string, wg *sync.WaitGroup) {
|
||
|
||
defer wg.Done()
|
||
// 加锁
|
||
wr.Lock.Lock()
|
||
// 解锁
|
||
defer wr.Lock.Unlock()
|
||
|
||
// 判断是否为 [] 结尾 - 普通数据
|
||
if strings.HasSuffix(keys, "[]") {
|
||
// 将 [] 替换为 .
|
||
key := strings.Replace(keys, "[]", ".", -1)
|
||
wr.Data[key] = value
|
||
return
|
||
}
|
||
keys = strings.Replace(keys, "[", ".", -1)
|
||
keys = strings.Replace(keys, "]", "", -1)
|
||
|
||
// 正则匹配末尾是否为 .[0-9]+ - 【一维数组】
|
||
reg := regexp.MustCompile(`\.[0-9]+$`)
|
||
if reg.MatchString(keys) {
|
||
|
||
// 取到 .[0-9]+ 的内容
|
||
index := reg.FindString(keys)
|
||
// 去除 . 只要数字
|
||
index = strings.Replace(index, ".", "", -1)
|
||
|
||
// 将 .[0-9]+ 替换为 .
|
||
key := reg.ReplaceAllString(keys, ".")
|
||
// 判断 key 是否存在，如果不存在，则创建 - 【初始化一维数组】
|
||
if _, ok := wr.Data[key]; !ok {
|
||
|
||
item := make([]any, cast.ToInt(index)+1)
|
||
item[cast.ToInt(index)] = value[0]
|
||
wr.Data[key] = item
|
||
|
||
} else {
|
||
|
||
// 如果存在，则追加，先判断 index 是否超出数组长度 - 【追加一维数组】
|
||
if cast.ToInt(index) >= len(wr.Data[key].([]any)) {
|
||
|
||
// 超出长度，需要扩容
|
||
item := make([]any, cast.ToInt(index)+1)
|
||
// 将原来的数据复制到新数组
|
||
copy(item, wr.Data[key].([]any))
|
||
// 将新数据追加到新数组
|
||
item[cast.ToInt(index)] = value[0]
|
||
wr.Data[key] = item
|
||
|
||
} else {
|
||
|
||
// 未超出长度，直接追加
|
||
wr.Data[key].([]any)[cast.ToInt(index)] = value[0]
|
||
}
|
||
}
|
||
return
|
||
}
|
||
// 如果不是 [] 结尾，也不是 .[0-9]+ 结尾，则直接赋值 - 【普通数据】
|
||
wr.Data[keys] = value[0]
|
||
}
|
||
|
||
for key, value := range params {
|
||
wg.Add(1)
|
||
go worker(key, value, &wg)
|
||
}
|
||
|
||
wg.Wait()
|
||
|
||
return wr.Data
|
||
}
|
||
|
||
// Params - 解析参数
|
||
// 把 Content-Type = application/x-www-form-urlencoded 的参数解析成 map[string]any
|
||
func (this *ParseClass) Params(params map[string]any) (result map[string]any) {
|
||
|
||
wg := sync.WaitGroup{}
|
||
wr := wrLock{
|
||
Data: make(map[string]any, 3),
|
||
Lock: &sync.RWMutex{},
|
||
}
|
||
|
||
var worker func(params map[string]any, key string, val any)
|
||
worker = func(params map[string]any, key string, val any) {
|
||
|
||
// 加锁
|
||
wr.Lock.Lock()
|
||
defer wg.Done()
|
||
// 解锁
|
||
defer wr.Lock.Unlock()
|
||
|
||
// 如果 key 的末尾为 . ，表示该 key 为一维数组
|
||
if strings.HasSuffix(key, ".") {
|
||
key = strings.TrimSuffix(key, ".")
|
||
}
|
||
|
||
// 通过 . 分割 key
|
||
keys := strings.Split(key, ".")
|
||
|
||
// 判断是否为最后一个元素
|
||
if len(keys) == 1 {
|
||
params[keys[0]] = val
|
||
return
|
||
}
|
||
|
||
// 如果不是最后一个元素，则继续递归
|
||
if _, ok := params[keys[0]]; !ok {
|
||
params[keys[0]] = make(map[string]any, 3)
|
||
}
|
||
wg.Add(1)
|
||
go worker(cast.ToStringMap(params[keys[0]]), strings.Join(keys[1:], "."), val)
|
||
return
|
||
}
|
||
|
||
for key, val := range params {
|
||
wg.Add(1)
|
||
go worker(wr.Data, key, val)
|
||
}
|
||
|
||
wg.Wait()
|
||
|
||
return wr.Data
|
||
}
|
||
|
||
// Domain - 解析域名
|
||
func (this *ParseClass) Domain(value any) (domain string) {
|
||
URL, err := url.Parse(cast.ToString(value))
|
||
if err != nil {
|
||
return ""
|
||
}
|
||
return URL.Hostname()
|
||
}
|
||
|
||
// HtmlToText - 去除 HTML 标签、解码实体、压缩空白，并按 Unicode 字符截取前 length 个字符
|
||
func (this *ParseClass) HtmlToText(html string, length int, isLine bool) (text string) {
|
||
|
||
// 首先，对HTML实体进行转义，使像< >这样的符号变为实际字符
|
||
content := HTML.UnescapeString(html)
|
||
// 将非换行空格规范化为普通空格
|
||
content = strings.ReplaceAll(content, "\u00A0", " ")
|
||
|
||
// 把 <br> 和 <br />（不区分大小写）替换为换行符
|
||
br := regexp.MustCompile(`(?i)<br\s*/?>`)
|
||
content = br.ReplaceAllString(content, "\n")
|
||
|
||
// 把常见的块级元素的起始/结束标签替换为换行符（例如 <p>, <div>, <h1>.. 等）
|
||
// 这样可以保留它们原本的换行语义
|
||
block := regexp.MustCompile(`(?i)</?(p|div|h[1-6]|li|ul|ol|blockquote|address|article|section|header|footer|nav|pre|table|tr|td|th|hr)[^>]*>`)
|
||
content = block.ReplaceAllString(content, "\n")
|
||
|
||
// 去掉剩余的标签
|
||
tag := regexp.MustCompile(`(?s)<[^>]*>`)
|
||
content = tag.ReplaceAllString(content, "")
|
||
|
||
// 统一 CRLF -> LF
|
||
content = strings.ReplaceAll(content, "\r\n", "\n")
|
||
content = strings.ReplaceAll(content, "\r", "\n")
|
||
|
||
// 压缩连续的空格/制表符为单个空格（保留换行符），并压缩连续换行为单个换行
|
||
reSpace := regexp.MustCompile(`[ \t]+`)
|
||
content = reSpace.ReplaceAllString(content, " ")
|
||
reNewline := regexp.MustCompile(`\n+`)
|
||
content = reNewline.ReplaceAllString(content, "\n")
|
||
|
||
// 去除每行前后空白并整体 Trim
|
||
lines := strings.Split(content, "\n")
|
||
for i := range lines {
|
||
lines[i] = strings.TrimSpace(lines[i])
|
||
}
|
||
content = strings.TrimSpace(strings.Join(lines, "\n"))
|
||
|
||
if !isLine {
|
||
// 去除所有换行符，变成单行文本
|
||
content = strings.ReplaceAll(content, "\n", " ")
|
||
}
|
||
|
||
// length == 0 表示不截断，返回全部过滤后的内容
|
||
if length == 0 { return content }
|
||
|
||
// 按 rune 截取，避免切断多字节字符
|
||
runes := []rune(content)
|
||
if len(runes) <= length { return content }
|
||
|
||
// 超出长度则截取前 n 个字符并追加省略号
|
||
return string(runes[:length]) + "..."
|
||
}
|
||
|
||
// HtmlToTextRetainImg - 去除 HTML 标签、解码实体、压缩空白，保留 img 标签并添加 emoji 类，并按 Unicode 字符截取前 length 个字符
|
||
func (this *ParseClass) HtmlToTextRetainImg(html string, length int, isLine bool) (text string) {
|
||
|
||
// 首先，对HTML实体进行转义，使像< >这样的符号变为实际字符
|
||
content := HTML.UnescapeString(html)
|
||
// 将非换行空格规范化为普通空格
|
||
content = strings.ReplaceAll(content, "\u00A0", " ")
|
||
|
||
// 把 <br> 和 <br />（不区分大小写）替换为换行符
|
||
br := regexp.MustCompile(`(?i)<br\s*/?>`)
|
||
content = br.ReplaceAllString(content, "\n")
|
||
|
||
// 把常见的块级元素的起始/结束标签替换为换行符
|
||
block := regexp.MustCompile(`(?i)</?(p|div|h[1-6]|li|ul|ol|blockquote|address|article|section|header|footer|nav|pre|table|tr|td|th|hr)[^>]*>`)
|
||
content = block.ReplaceAllString(content, "\n")
|
||
|
||
// 处理 img 标签：保留并添加/追加 emoji 类
|
||
imgRegex := regexp.MustCompile(`(?i)<img\b([^>]*)>`)
|
||
content = imgRegex.ReplaceAllStringFunc(content, func(imgTag string) string {
|
||
// 检查是否已有 class 属性
|
||
classRegex := regexp.MustCompile(`(?i)\bclass\s*=\s*(["'])(.*?)\1`)
|
||
|
||
if match := classRegex.FindStringSubmatch(imgTag); match != nil {
|
||
// 已有 class 属性
|
||
quote := match[1]
|
||
classes := match[2]
|
||
|
||
// 检查是否已有 emoji 类
|
||
if !regexp.MustCompile(`\bemoji\b`).MatchString(classes) {
|
||
// 追加 emoji 类
|
||
newClasses := strings.TrimSpace(classes) + " emoji"
|
||
return classRegex.ReplaceAllString(imgTag, fmt.Sprintf(`class=%s%s%s`, quote, newClasses, quote))
|
||
}
|
||
return imgTag
|
||
} else {
|
||
// 没有 class 属性，添加 class="emoji"
|
||
endRegex := regexp.MustCompile(`(\s*/?>)`)
|
||
if endRegex.MatchString(imgTag) {
|
||
return endRegex.ReplaceAllString(imgTag, ` class="emoji"$1`)
|
||
}
|
||
}
|
||
return imgTag
|
||
})
|
||
|
||
// 去掉除了 img 之外的所有标签
|
||
tag := regexp.MustCompile(`(?i)<(?!img\b)[^>]*>`)
|
||
content = tag.ReplaceAllString(content, "")
|
||
|
||
// 统一 CRLF -> LF
|
||
content = strings.ReplaceAll(content, "\r\n", "\n")
|
||
content = strings.ReplaceAll(content, "\r", "\n")
|
||
|
||
// 压缩连续的空格/制表符为单个空格（保留换行符），并压缩连续换行为单个换行
|
||
reSpace := regexp.MustCompile(`[ \t]+`)
|
||
content = reSpace.ReplaceAllString(content, " ")
|
||
reNewline := regexp.MustCompile(`\n+`)
|
||
content = reNewline.ReplaceAllString(content, "\n")
|
||
|
||
// 去除每行前后空白并整体 Trim
|
||
lines := strings.Split(content, "\n")
|
||
for i := range lines {
|
||
lines[i] = strings.TrimSpace(lines[i])
|
||
}
|
||
content = strings.TrimSpace(strings.Join(lines, "\n"))
|
||
|
||
if !isLine {
|
||
// 去除所有换行符，变成单行文本
|
||
content = strings.ReplaceAll(content, "\n", " ")
|
||
}
|
||
|
||
// length == 0 表示不截断，返回全部过滤后的内容
|
||
if length == 0 {
|
||
return content
|
||
}
|
||
|
||
// 按 rune 截取，避免切断多字节字符
|
||
runes := []rune(content)
|
||
if len(runes) <= length {
|
||
return content
|
||
}
|
||
|
||
// 超出长度则截取前 n 个字符并追加省略号
|
||
return string(runes[:length]) + "..."
|
||
} |