Hi Erick, The "fq" did the trick. This basically solved my need, and I can call it a day (now that it is late Friday)
The reason why I'm using two (and there will be move) handlers vs "qf" in the URL, is due to the GET limit. The list of fields will be large (nearing 1000) and each field name can be long (up to 40 characters). If Solr will accept POST, than I can pass the list via "qf" and call it a day (I might have to worry a bit about the larger than normal network traffic per search request, but I can deal with that). So, do you know or does any one know if Solr supports POST request? If so, what's the body format that I need to send (Shawn asked this question too)? Thanks Steve On Wed, May 13, 2015 at 8:13 PM, Erick Erickson <erickerick...@gmail.com> wrote: > Two things: > > > 1> There's really no need to define two request handlers here. The > <defaults> section is exactly that, defaults which can be overridden > by the URL. So rather than have select_group_b, use something like > > ... solr/collection/select_group_a?q=whatever&qf=F2,F3,F5 > > 2> When you add a field qualifier to an edismax query, it overrides > all the qf definitions. Consider using an fq (filter query) clause by > adding &fq=type:(PDF OR DOC OR TXT) > > > Best, > Erick > > On Wed, May 13, 2015 at 6:15 PM, Steven White <swhite4...@gmail.com> > wrote: > > Looks like I got it working (however I still have an outstanding issue, > see > > end of my email). > > > > Here is what I have done: > > > > 1) In my solrconfig.xml, I created: > > > > <requestHandler name="/select_group_a" class="solr.SearchHandler"> > > <lst name="defaults"> > > <str name="echoParams">explicit</str> > > <int name="rows">20</int> > > <str name="defType">edismax</str> > > <str name="qf">F1 F2 F3</str> > > <str name="fl">type,id,score</str> > > <str name="wt">xml</str> > > <str name="indent">true</str> > > </lst> > > </requestHandler> > > > > And > > > > <requestHandler name="/select_group_b" class="solr.SearchHandler"> > > <lst name="defaults"> > > <str name="echoParams">explicit</str> > > <int name="rows">20</int> > > <str name="defType">edismax</str> > > <str name="qf">F2 F3 F5</str> > > <str name="fl">type,id,score</str> > > <str name="wt">xml</str> > > <str name="indent">true</str> > > </lst> > > </requestHandler> > > > > 2) My search URL is now: > > http://localhost:8983/solr/db/select_group_a?q.op=OR&q=<search string> > and > > http://localhost:8983/solr/db/select_group_b?q.op=OR&q=<search string> > > > > This all works, BUT when I use "q=type:(PDF OR DOC OR TXT)" so that I can > > further narrow down search to within, for example, file-extensions, this > > doesn't seem to work. Is this because using "qf" with edismax ends > doesn't > > parse the string the same way as the default defType? > > > > Steve > > > > On Wed, May 13, 2015 at 6:11 PM, Steven White <swhite4...@gmail.com> > wrote: > > > >> Thanks for the quick reply Shawn. I will dig into dismax and edismax > and > >> come back with questions if I cannot figure it out. I avoided them > >> thinking they are for faceting use only, my need is generic search (all > the > >> features I get via solr.SearchHandler) but limited to a set of fields. > >> > >> Steve > >> > >> On Wed, May 13, 2015 at 5:58 PM, Shawn Heisey <apa...@elyograg.org> > wrote: > >> > >>> On 5/13/2015 3:36 PM, Steven White wrote: > >>> > <requestHandler name="/select_group_a" class="solr.SearchHandler"> > >>> > <lst name="defaults"> > >>> > <str name="echoParams">explicit</str> > >>> > <int name="rows">20</int> > >>> > <str name="df">F2,F3,F5</str> > >>> > <str name="fl">id,score</str> > >>> > </lst> > >>> > </requestHandler> > >>> > > >>> > However, this isn't working because whatever is in "df" is being > >>> treated as > >>> > single field name. > >>> > >>> The df parameter is shorthand for "default field." It is, by > >>> definition, a single field -- it is the field searched by default when > >>> you don't specify a field directly in a query handled by the default > >>> (lucene) query parser. The default parser doesn't search multiple > >>> fields for your search terms. > >>> > >>> What you're going to want to do here is use a different query parser -- > >>> dismax or edismax -- and put your field list in the qf field, separated > >>> by spaces rather than commas. The qf parameter means "query fields" > and > >>> is specific to the dismax/edismax parsers. Depending on your exact > >>> needs, you may also want to define the pf parameter as well (phrase > >>> fields). > >>> > >>> There is a LOT of detail on these parsers, so I'll give you the > >>> documentation links rather than try and explain everything: > >>> > >>> > https://cwiki.apache.org/confluence/display/solr/The+DisMax+Query+Parser > >>> > >>> > https://cwiki.apache.org/confluence/display/solr/The+Extended+DisMax+Query+Parser > >>> > >>> Thanks, > >>> Shawn > >>> > >>> > >> >
