On 20.01.2010, at 15:50, Lukas Kahwe Smith wrote: > > On 19.01.2010, at 22:52, Lukas Kahwe Smith wrote: > >>>> I also want to match multiple fields at once. >>> >>> Can you give an example? >> >> >> I enter "Kreuz" but this could either be part of a persons name or of a >> street name, which are separate fields in my index mainly because they >> analyzed differently (person name using doublemetaphone and street name >> using word splitting to extract relevant parts for better matching). > > > This is still really the kicker for me. Having to use TC on multiple fields > myself and adding up the results seems needlessly tedious. Plus, though I am > not sure if solr could do any better job here, adding up could produce dupes > (like a company using the city name as part of their company name).
Sorry about my little monolog here. Please free to jump in if you have some insights :) Obviously one crazy approach could be to do a termscomponents search on all the fields and in order to get proper numbers run a dismax search for each term. Another approach that I find a bit more feasible at this point is combining all fields together as the field to use for termscomponents. obviously this means that i cannot apply all of the various field specific analyzers that way, but its just auto complete. if people know what they want they will often cut&paste or just blindly type things out anyways. This however means that the numbers can be lower for a given term (since for example the doublemetaphone analyzer would not be applied to the name nor the word splitter for the street names). So if I search for "Kreuz" I would not see people whos names are "Creuz" or streetnames that are "Oberkreuzstrasse". But I would see people whos name is "Kreuz" or who live at "Kreuzstrasse". So I will probably just represent the values shown in the auto suggest with a plus sign next to them or maybe even round them down to ranges (1+, 10+, 50+, 100+, 1000+, 10000+ ..) regards, Lukas Kahwe Smith m...@pooteeweet.org