One client I used two search indices and combined them.  Oddly enough it was faster for me to do this in XSLT.

<xsl:variable name="resultsRaw">
   <results>
     <xsl:for-each select="msxml:node-set(umbraco.page:Search($searchTerm))/nodes/node [string(hideFromSearch) != '1']" >
       <xsl:copy-of select="." />
     </xsl:for-each>
     <xsl:for-each select="msxml:node-set(umbraco.page:SearchPDF($searchTerm))/nodes/node" >
       <xsl:copy-of select="." />
     </xsl:for-each>
   </results>
 </xsl:variable>
 <xsl:variable name="results" select="msxml:node-set($resultsRaw)/results" />
<xsl:for-each select="$results/node">          
   <xsl:sort select="./@score" order="descending" data-type="number" />
 </xsl:for-each>    

Jon,

Was the reason because you had issue with multi searcher?

I can do 2 searches then combine and sort with linq.

Regards

Ismail

Honestly, this was done several years ago and I was still doing nearly all my macro development in XSLT on the 4.5+ platform. It was not a Multisearcher.

<add name="EnforcerIndexer" type="UmbracoExamine.UmbracoContentIndexer, UmbracoExamine"
     runAsync="true"
     supportUnpublished="false"
     supportProtected="false"
     interval="10"
     analyzer="Lucene.Net.Analysis.Standard.StandardAnalyzer, Lucene.Net"
     IndexSet="EnforcerIndexSet" />

<add name="PDFIndexer" type="UmbracoExamine.PDF.PDFIndexer, UmbracoExamine.PDF" extensions=".pdf" umbracoFileProperty="umbracoFile"/>

So, I didn't answer your question, but maybe my lackof Examine knowledge at the time might offer an alternative solution if you need proper scoring until you get a proper answer.

I felt the search was quick enough I didn't need to combine into a single searcher.

You are welcome to see the performance of the implementation. One requirement was to display any text available in the PDF; unfortunately, the client has not modified the annotations so you see garbage in the results

Honestly, this was done several years ago and I was still doing nearly all my macro development in XSLT on the 4.5+ platform. It was not a Multisearcher.

<add name="EnforcerIndexer" type="UmbracoExamine.UmbracoContentIndexer, UmbracoExamine"
     runAsync="true"
     supportUnpublished="false"
     supportProtected="false"
     interval="10"
     analyzer="Lucene.Net.Analysis.Standard.StandardAnalyzer, Lucene.Net"
     IndexSet="EnforcerIndexSet" />

<add name="PDFIndexer" type="UmbracoExamine.PDF.PDFIndexer, UmbracoExamine.PDF" extensions=".pdf" umbracoFileProperty="umbracoFile"/>

So, I didn't answer your question, but maybe my lackof Examine knowledge at the time might offer an alternative solution if you need proper scoring until you get a proper answer.

I felt the search was quick enough I didn't need to combine into a single searcher.

You are welcome to see the performance of the implementation. One requirement was to display any text available in the PDF; unfortunately, the client has not modified the annotations so you see garbage in the results

Jon,

Ok I have already answered this many moons ago http://our.umbraco.org/forum/developers/extending-umbraco/13963-Examine-wildcard-search-score doh!! In both instances I am doing wildcard so in both instances the score is 1. Either I update examine source or take out the wild card.

Regards

Ismail

Do the update to source effect non-wildcard queries?  It might be something worth recommending to be added to core.  Glad you found your answer [again].

Jon,

That old post did not fix the issue.

Regards

Ismail

is working on a reply...