i solve this problem by developing of DocumentBuilder.toDocument() and DocumentBuilder.addField() functions.i don't use multiValue feature and in shcema i change the condition of check for multi value fields(skip it in colde) then in necessary position i put my custom gap as a value for textField.
now i have a question in schema. why we set a field as multi value? in other word what kind of situation caused use of this feature because i think it was possible to let to the developer to add multiple value for a field then it will be multi value or not then it will be single value. thanks, Amin On Tue, Dec 19, 2017 at 9:59 AM, Amin Raeiszadeh <amin24march1...@gmail.com> wrote: > Shawn i think your way is good. i will study more about it. > thanks, > Amin > > > On Tue, Dec 19, 2017 at 9:57 AM, Amin Raeiszadeh > <amin24march1...@gmail.com> wrote: >> Erick, in your example if first entry contains 10 term then i need to >> start the second entry position from 100 not from 110. >> thanks, >> Amin >> >> On Tue, Dec 19, 2017 at 3:25 AM, Shawn Heisey <apa...@elyograg.org> wrote: >>> On 12/18/2017 12:29 AM, Amin Raeiszadeh wrote: >>>> thanks too much Erick and mikhail. >>>> i change SloppyPhraseScorer class for my custom behavior with some fields. >>>> so i need to index some fields with customized gap between terms of fields. >>>> i'm not profession with solr and i think with schema.xml only i can >>>> set fixed gap increment between terms of field but i need to calculate >>>> base on some formula and then decide how much gap should be set >>>> between each term. >>>> i think i should use some thing like plug-in or inject some code to >>>> index flow of solr as you described. >>> >>> Most likely what you would need is a custom TokenFilter implementation >>> that can determine what custom gaps are needed and implement them. You >>> would then need to make sure Solr knows about the jar containing that >>> plugin, and add that filter as a step in the analysis chain in the >>> schema for your core/collection. >>> >>> I don't know anything about how to write a TokenFilter, but there are >>> lots of implementations in the Lucene code. >>> >>> Thanks, >>> Shawn >>>
