This post originated from an RSS feed registered with .NET Buzz
by Scott Watermasysk.
Original Post: .Text Search
Feed Title: ScottW's ASP.NET WebLog
Feed URL: /error.aspx?aspxerrorpath=/blog/rss.aspx
Feed Description: ASP.NET and Blogging
I managed to put together the basics for .Text's search feature this weekend. I will post later how it all works, but for now I would love some feedback.
The index contains the 18,000 or so posts and articles from both http://weblogs.asp.net and http://blogs.msdn.com . This will NOT be the permanent home for the index. To make the index live, I would have to push a new build of .Text and I am not ready for that yet (probably next weekend).
Traditional boolean searches using && and || (btw, would you rather use "and" and "or")
Single term (Scott) or phrases ("Scott Watermasysk")
Field Searches. When adding a search term, just include the field name and a colon. Ex: blog:brada && enum should return all of the posts by Brad Abrams with the term enum in the post. (blog:brada && title:enum would display Brad's posts with enum in the title.) By default only the first word after the colon will be included. To search for multiple words, include them in quotes. (title:"enum versioning"). Also note, if you want to limit a search by a field, you must explicitly specify "&&".
blog
author
title
body (default field if not specified)
link (links found in the body of each post, not the url itself)
domain
Wildcard searches using "*" (0 or more characters) and "?" (one character). Ex: title:cach* should return post titles with both cache and caching
Fuzzy searches using "~" Ex: Term~
Proximity searches allow you to search for words with in a specific range to each other. Ex: "scott .Text"~5 should return results where "scott" and ".Text" are within 5 words of each other
Term boosting allows you to "weight" a particular word in a search. Ex: Scott .Text^3 will return results with scott or .Text, but will lean towards .Text