Could you provide some more details on your use case? This sounds like an XY problem (see http://people.apache.org/~hossman/#xyproblem). The reason I say this is that you're probably going to shoot yourself in the foot if you require such symbols, leading to an "interesting" user experience.
That said, you can pre-process your data for both indexing and seaching by, say applying a regex that strategically removes the spaces you care about and using, say, the whitespacetokenizer. I'll also apply a lowercasefilter. HTH Erick On Thu, Jul 29, 2010 at 10:25 AM, <markus.rietz...@rzf.fin-nrw.de> wrote: > hi, > what is the best way to deal with searches with special chars like § > (paragraph), € (euro), @ (at in emails), % and so forth. > i think that the WordDelimiterFilters is working on such chars (on > index-time and on query-time). > > the greatest problem i see is, that there can be an optional space between > those chars and numbers, like 50% or 50 %, or §235 or § 235 and so on. > so even if i get the WordDelimiter (or any other filter) right and working > with those chars i think there is no way to deal with the optional spaces. > > anyone have a solution for this. > > > > markus >
