﻿<?xml version="1.0" encoding="utf-8"?><rss version="2.0"><channel><title>Ayende @ Rahien</title><link>http://ayende.com</link><description>Ayende @ Rahien</description><copyright>Copyright (C) Ayende Rahien  2004 - 2021 (c) 2026</copyright><ttl>60</ttl><item><title>Andrew commented on NugGet Perf, Part V&amp;ndash;Searching Packages</title><description>@Fujiy One reason to have tags inside the Query field and outside is to allow for a search all in one field (Query) and also support explicit tag searches (just want the tag 'Finance' not something with Finance in it's title)

It can also mean you can sort by tag (albeit the first one only)</description><link>http://ayende.com/158211/nugget-perf-part-v-searching-packages#comment5</link><guid>http://ayende.com/158211/nugget-perf-part-v-searching-packages#comment5</guid><pubDate>Tue, 04 Sep 2012 00:32:06 GMT</pubDate></item><item><title>Fujiy commented on NugGet Perf, Part V&amp;ndash;Searching Packages</title><description>Why are you indexing p.Tags inside the Query property and outside too?

RavenDB supports filtered indexes? It can achieve impressive performance boost sometimes, decreasing the index size from thousands or millions rows to dozens rows</description><link>http://ayende.com/158211/nugget-perf-part-v-searching-packages#comment4</link><guid>http://ayende.com/158211/nugget-perf-part-v-searching-packages#comment4</guid><pubDate>Mon, 03 Sep 2012 14:34:33 GMT</pubDate></item><item><title>Simon Skov Boisen commented on NugGet Perf, Part V&amp;ndash;Searching Packages</title><description>John,

Yes Lucene handles substring, regex and levensteihn string relations through the  FuzzyQuery and WildCardQuery. They are not as quick as pure index-lookups but they are quite quick after version 4.0 introduced a levensteihn automata to do the job of comparing an index term with a search term.</description><link>http://ayende.com/158211/nugget-perf-part-v-searching-packages#comment3</link><guid>http://ayende.com/158211/nugget-perf-part-v-searching-packages#comment3</guid><pubDate>Mon, 03 Sep 2012 11:33:06 GMT</pubDate></item><item><title>Rafal commented on NugGet Perf, Part V&amp;ndash;Searching Packages</title><description>They could implement poor man's keyword search in SQL by creating a table that maps keyword-&gt;package Id (an inverted index for packages). For such a simple database there shouldn't be any problem with performance or with keeping the index up to date.</description><link>http://ayende.com/158211/nugget-perf-part-v-searching-packages#comment2</link><guid>http://ayende.com/158211/nugget-perf-part-v-searching-packages#comment2</guid><pubDate>Mon, 03 Sep 2012 10:52:50 GMT</pubDate></item><item><title>John Landheer commented on NugGet Perf, Part V&amp;ndash;Searching Packages</title><description>Ah, but does this find a package named 'MyJQueryExtension' ? I understand that substring searches are not always a good idea, but in this case you're not replicating the search. Can RavenDb and/or Lucene handle substring searches?</description><link>http://ayende.com/158211/nugget-perf-part-v-searching-packages#comment1</link><guid>http://ayende.com/158211/nugget-perf-part-v-searching-packages#comment1</guid><pubDate>Mon, 03 Sep 2012 09:56:39 GMT</pubDate></item></channel></rss>