mysql关于拥有和计数的查询问题

Question

在这个查询*中，我按纬度、经度、各种标准……从近到远或得分及其工作/列出数据列出并排序“mytableabc”中的行，我认为它足够好/足够快！？

select 
round((6371 * acos(cos(radians('40.123456')) * cos(radians(latitude)) * cos(radians(longtitude) - radians('29.123456')) + sin(radians(latitude)) * sin(radians('40.123456')))), (2)) as distance, 
    (match(colone,coltwo) against('searchkeywordabc' in boolean mode)) as score, 
id,colone,coltwo,latitude,longtitude,colthree 
        from mytableabc 
        where (colone='sampleforsomething') 
            and (match(colone,coltwo) against('searchkeywordabc' in boolean mode))
                having distance <= 5 order by distance asc limit 0,50

--问题

在不同的行和文件中，我只需要具有相同条件的返回元素的总数 count(id)... 我尝试添加计数（id），但它没有给出正确的计数总数和列表行，我只想要总数（就像通常计数查询一样）而不是数据行

select 
    count(id), 
round((6371 * acos(cos(radians('40.123456')) * cos(radians(latitude)) * cos(radians(longtitude) - radians('29.123456')) + sin(radians(latitude)) * sin(radians('40.123456')))), (2)) as distance, 
        (match(colone,coltwo) against('searchkeywordabc' in boolean mode)) as score, 
    id,colone,coltwo,latitude,longtitude,colthree 
            from mytableabc 
            where (colone='sampleforsomething') 
                and (match(colone,coltwo) against('searchkeywordabc' in boolean mode))
                    having distance <= 5 order by distance asc limit 0,50

--这个没有距离/纬度/经度标准的查询也很好用，只返回总数......但我需要在这个查询中添加距离查询

select count(id) from mytableabc where (colone='sampleforsomething') and (match(colone,coltwo) against('searchkeywordabc' in boolean mode))

我不是 sql 专家，是的，正如您所见，我需要一个。 谢谢。

Answer 1

A 计划：

将distance 检查移至WHERE 子句。这消除了对HAVING 的需要，同时过滤更多，因此COUNT 不会被夸大。

select  count(*)
    from  mytableabc
    where  (colone='sampleforsomething')
      and  (match(colone,coltwo) against('searchkeywordabc' in boolean mode))
      AND  (6371 * acos(cos(radians('40.123456')) * 
        cos(radians(latitude)) * cos(radians(longtitude) -
        radians('29.123456')) + sin(radians(latitude)) *
        sin(radians('40.123456')))) < 5;

我希望首先使用 FULLTEXT 索引，然后检查其他两个测试（= 和距离）以检查 FT 搜索返回的所有行。

B计划：

继续使用HAVING，但有一个子查询。 （这可能更慢。）