If Solr's persian configuration is not sufficient, you could look into commercial implementation from Basis Tech (I haven't tested it): http://www.basistech.com/text-analytics/rosette/base-linguistics/for-arabic/ (says it supports persian at the bottom of the page).
I would also open a JIRA with example of what works and what does not. Even if it does not get implemented very soon, it would be clear record of what to aim for. Regards, Alex. ---- Solr Analyzers, Tokenizers, Filters, URPs and even a newsletter: http://www.solr-start.com/ On 1 September 2015 at 08:27, sara hajili <hajili.s...@gmail.com> wrote: > i used pre-built persian analyzer but it isn't enough for me. > this work just for search exacly indexed work! > like search 'go' and search result is 'go' > as i said in english we have stemming,synonym,etc filters that help us to > have flexible search. > and i want some filter for persian. > that pre-built text_fa doesn't satisfied me.have you better perisan filter > than that?or a soulotion to have this filter in persian? > tnx. > > On Tue, Sep 1, 2015 at 5:21 AM, Alexandre Rafalovitch <arafa...@gmail.com> > wrote: > >> http://blog.griddynamics.com/2015/08/scoring-join-party-in-solr-53.html >> shows how to keep updates in a separate core. Notice that it is an >> intermediate-level article for query syntax. >> >> For persian text analysis, there is a pre-built analyser defiition in >> the techproducts example, start from that. It is in the schema.xml in >> server/solr/configsets/sample_techproducts_configs/conf and is one of >> the example configsets. >> >> Regards, >> Alex. >> ---- >> Solr Analyzers, Tokenizers, Filters, URPs and even a newsletter: >> http://www.solr-start.com/ >> >> >> On 1 September 2015 at 08:07, sara hajili <hajili.s...@gmail.com> wrote: >> > and another question is: >> > my docs are persian and i use text_fa for fieldType but i wanna to have a >> > persian textfield that handle search problem such as stemming. >> > word distance,synonyms etc >> > like english types. >> > as i said i handle "میخواهم " and "خواستن" and so on. >> > can you suggest me a fieldtype for handle this issues in persian field. >> > tnx >> > >> > On Tue, Sep 1, 2015 at 3:16 AM, sara hajili <hajili.s...@gmail.com> >> wrote: >> > >> >> i'm really confused:| >> >> i'm really anxious about cost of update like count. >> >> and as you said: >> >> <field name="like_count" type="int" indexed="true" stored="true" >> >> docValues="true"/> >> >> you indexed like_count field .and i think it cost alot to update and >> index >> >> again docs. >> >> because like count change more and more >> >> so isn't better to indede="false" that this field name??!! >> >> >> >> On Tue, Sep 1, 2015 at 3:08 AM, Upayavira <u...@odoko.co.uk> wrote: >> >> >> >>> you don't need to use a dynamic field, just a normal field will work >> for >> >>> you. But, you *will* want to index it, and you may benefit from >> >>> docValues, so: >> >>> >> >>> <field name="like_count" type="int" indexed="true" stored="true" >> >>> docValues="true"/> >> >>> >> >>> Upayavira >> >>> >> >>> On Tue, Sep 1, 2015, at 10:59 AM, sara hajili wrote: >> >>> > my solr version is 5.2.1 >> >>> > i have a question. >> >>> > if i create 2 core .one for post and one for like . i must index like >> >>> > count? >> >>> > i mean in schema for like core i must write: >> >>> > <dynamicField name="like_count" type="int" indexed="false" >> >>> > stored="true"/> >> >>> > >> >>> > am i true? >> >>> > >> >>> > On Tue, Sep 1, 2015 at 2:42 AM, Upayavira <u...@odoko.co.uk> wrote: >> >>> > >> >>> > > So you want to be able to sort by the "number of likes" value for a >> >>> > > post? >> >>> > > >> >>> > > What version of Solr are you using? How many posts do you have? >> >>> > > >> >>> > > There's a neat feature in Solr 5.2.1 (I'm pretty sure it is there, >> not >> >>> > > 5.3) called score joins. Using that you can have two cores, one >> >>> > > containing your posts, and another containing your likes. >> >>> > > >> >>> > > You cannot *sort* on these values, but you can include your likes >> into >> >>> > > the score, which might even be better. >> >>> > > >> >>> > > If this sounds good, I can dig up some syntax for such a query. >> >>> > > >> >>> > > Upayavira >> >>> > > >> >>> > > On Tue, Sep 1, 2015, at 10:36 AM, sara hajili wrote: >> >>> > > > hi. >> >>> > > > at first i.m sorry for my bad english! >> >>> > > > i have a social app.i want to use solr for searching in this app. >> >>> > > > i have many document (in my case people text that posted on my >> >>> social >> >>> > > > app). >> >>> > > > and i indexed this. >> >>> > > > but i'm have 1 issue and it is : >> >>> > > > >> >>> > > > i have very doc(post) and they have a property "like" is it good >> >>> > > > approach >> >>> > > > to index like count (people can like eachother post in my social >> >>> app)? >> >>> > > > likecount change more and more in one day.(so as i know it must >> be >> >>> set >> >>> > > > dynamic field) >> >>> > > > and if i indexed it ,i think it costs alot , to update and index >> >>> > > > likecount >> >>> > > > more and more even i use bach update. >> >>> > > > so is it approach to didn't index one field in solr but i could >> >>> sort my >> >>> > > > search result according to that unindexed field? >> >>> > > > >> >>> > > > tnx >> >>> > > >> >>> >> >> >> >> >>
