Hi,-
I hope everyone is doing great.
i have a question regarrding ComplexPhraseQueryParser class.
This class can handle this queryText case very well:
"term1 erm2 abcd term3*"~2
(last term3 has * at the end and the whole phrase has slop value 2)
The term1, term2 and term3 are all in the
the standard phrasequery cannot do this, but you can prefilter the invalid
term(abcd) out by using MultiTerms api.
Also, I have found that “a b c”~2 phrase query does not really match “a x x b x
x c” by its implementation……
陈志祥
阿里巴巴 地图引擎核心算法工程师
电话：057128223456-81124100
邮箱：
Thanks Zhixiang. Yes, it cant find when there is an unrelated term in the
middle that is not indexed.
Similar to what You suggested:
i can try the queryText by excluding one term at a time with the
ComplexPhraseQueryParser and see best matches.
But, i'd rather this is embedded into a Lucene api
