正则筛选字数&GBK转码

正则:

    //1.去除Html语句<>
    NSString * preStr1 =@"<([0-9a-zA-Z\"=#:\\/\\.\\_\\-\\? 　])*>|<\\/([0-9a-zA-Z\"=#:\\/\\.\\_\\-\\? 　])*>|<([0-9a-zA-Z\"=#:\\/\\.\\_\\-\\? 　])*\\/>";
    NSString * replacement = @"";
    
    NSRegularExpression *regExp = [[NSRegularExpression alloc] initWithPattern:preStr1
                                                                       options:NSRegularExpressionCaseInsensitive
                                                                         error:nil];
    //替换
    newLengthStr = [regExp stringByReplacingMatchesInString:newLengthStr
                                                 options:NSMatchingReportProgress
                                                   range:NSMakeRange(0, newLengthStr.length)
                                            withTemplate:replacement];

• 用到的正则:1.特殊符号

"/(?<=[^x00-xff！…“”《》—、’])( |　)（？＝［＾x00-xff！…＂＂《》－、＇］）／"

• 2.\r\n 如果连续大于3个的情况

"(\r\n){3,}"

根据GBK计算长度

    //GBK转码
    NSStringEncoding gbkEncoding = CFStringConvertEncodingToNSStringEncoding(kCFStringEncodingGB_18030_2000);
    NSData *data=[newLengthStr dataUsingEncoding:gbkEncoding];
    NSInteger length = 0;
    if (data.length % 2 == 0) {
        length = data.length /2;
    }else{
        length = data.length /2 + 1;
    }