Switch branches/tags
Nothing to show
Find file
Fetching contributors…
Cannot retrieve contributors at this time
62 lines (40 sloc) 3.16 KB
MVC is an attempt to create components with self contained models controllers and views
Hopefeully these packages can be used to create interrelated applications and shared problem solvers
Noogle relies on node v.0.2.6. Higher installs have changed the way that post parameters are interpreted.
It uses the latest Mongo v.1.6. Upcoming Mongos might change the way that Map/Reduce works so they aren't
guaranteed to be compatible with Noogle.
Noogle is not currently real time - data must be loaded manually.
First use the "init" job to add the index page.
Run the app (node index.js); click pages, then click on the (only) page in the list; click "Spider" -
that will add all the links on that page as subpages.
Back on the console, browse to the "jobs" folder and run in order, ONE AT A TIME:
node reindex.js
node word_freq.js
node load_nicks.js
node update_nick_line_count.js
node auto_alias.js
Index the lines database on words, then nick (two seperate insureIndex() calls)
Noogle as it stands pretty much does the fundamental task of allowing you to find a conversation about a given topic. There are a few more wish list
features I have and a few roads I am currently developing.
0: Add Dynamic Updating
Right now the Noogle database updates when I upload a new data file set to the server. This is obviously not how a search engine is supposed to work.
Re-scripting the data spider to update incrementally is going to be a big win.
1: Nick Centric Features
I already have a view nick system (not live) in git, with a nick page that lists lines per day graphically. Because IRC often forces you to use a synonym
if you are on multiple machines I am trying to reconcile bingomanatee with bingomanatee_, bingomanatee__ to give a true aggregate report.
Once that is complete I also want to generate a "personal tag cloud" of words that each nick uses. (watch your swears boys!).
* in process - nicks and logins done, work needs to be done in coordinating them.
* TODO: an ACL system to lock out non - me people from playing with the pages.
1.1 Logins
I also want to build up a login system around nicks so people can "Claim" identities. This would give you the option of adding user pics, getting an RSS feed, contributing an article, et all.
2. Interactive Node.js chatting
. It would be great if I can build Noogle up into an IRC chat engine where people can go, chat, save snippets of text for future reference, etc., so that, for instance, when you type a question, you get instant noogle results and/or FAQ hits.
3. FAQ builder
I would like the ability to "sculpt" a search result into a FAQ, by removing nonparticipant nicks from a conversation, cropping it, and highlighting useful answer comments from the general text flow. That way we can have a FAQ page that browsers can build up, rate answers, and build a dictionary
of node help.
4. Convert the search engine to SOLR
I built up the current system in MongoDB partly because I wanted to get under the hood of Christkov-native. That being done, SOLR is a much better
text search engine and I would like to export the search function to it ASAP.