3

I have a table named 'comp' with a column 'compName', the compName contain different country's Characters, I am using mysql5.7 with ngram parser, now it is fine to search the Chinese word, but the it brings me the bad result when i searched English word. According to INFORMATION_SCHEMA.INNODB_FT_INDEX_CACHE table , i found it participle English word by the character, like: abc will be participled as ab,bc, But as we understand, the English will be participled as "SPACE", right ? So how to resolve this kinds of case when using ngram parser in mysql5.7.

Kris
  • 151
  • 1
  • 2
  • 12
  • I think, i resolved this problem by myself , We should separated the compName as two column ( One is englishName, another is chineseName), and then apply the different fulltext index into the corresponding column, After that we also need to separated the keyword of user search into english and chinese string,and apply the same into the column against their index. – Kris Dec 30 '16 at 06:34
  • My upon solution will bring another problem is two columns matchs will reduce the matched-degree : like : MATCH (title) AGAINST ('公司 ') and MATCH (title) AGAINST ('郑州 ') ) . So, any other solutions ? – Kris Jan 03 '17 at 00:56

0 Answers0