Question

我有4个表来执行索引。我的问题是，

我希望搜索结果不是按重量而是按表优先级。一世   平均表1匹配的结果先显示表2，依此类推。

目前，它首先按最高权重检索结果。

查询

Select UUID_SHORT() AS sphinxid, info as keyword, 'column_name' as type from table 1 
UNION ALL 
Select UUID_SHORT() AS sphinxid,  info as keyword, 'column_name' as type from table 2 
UNION ALL 
Select UUID_SHORT() AS sphinxid,  info as keyword, 'column_name' as type from table 3 
UNION ALL 
Select UUID_SHORT() AS sphinxid,  info as keyword, 'column_name' as type from table 4;

Answer 1

如果仍然使用原始的SphinxAPI，那么就有一个SetIndexWeights。 http://sphinxsearch.com/docs/current.html#api-func-setindexweights 虽然必须将数据拆分为单独的实际索引。

...... SphinxQL中没有这样的等价物。 AFAIK

向索引添加新属性可能最简单

sql_query =   \
Select ...,8 as multiplier from table1  \
UNION ALL  \
Select ...,4 as multiplier from table2  \
UNION ALL  \
...

sql_attr_uint = multiplier

然后在运行查询时，可以使用该因子

sphinxQL> SELECT id,WEIGHT()*multiplier AS w FROM index WHERE MATCH(..) ORDER BY w DESC;

如何设置表的优先级？

1 个答案: