jQuery UI comes with an autocomplete module that you can supply a source to 
(the source just outputs JSON for records that match the given thing typed). 
I've heard of using Redis, some cron-job-initiated indexing, and a fast service 
layer to populate it quickly enough that you can't tell it's coming from a 
request - of course, you'd have to have a redis server big enough to store 
however many entries you have. If that level of speed isn't important to you 
then the source can always come from a typical database.

Trey

-----Original Message-----
From: Code for Libraries [mailto:[email protected]] On Behalf Of Cary 
Gordon
Sent: Monday, July 08, 2013 8:06 AM
To: [email protected]
Subject: Re: [CODE4LIB] Lightweight Autocomplete Application

I may be missing some subtlety here, but it sounds like you are looking for a 
jQuery (Javascript) autocomplete script with an AJAX back end. This would load 
the vocabulary in the background, typically starting at page load.

This works fine as long as the vocabulary is of a reasonable size. If we are 
talking tens of millions of entries, it is not going to work so well, and if 
you have to look up terms in real time as you type, I don't know of any 
client/server tool that will be fast enough.

Thanks,

Cary

On Jul 8, 2013, at 7:37 AM, "Anderson, David (NIH/NLM) [E]" 
<[email protected]> wrote:

> I'm looking for a lightweight autocomplete application for data entry. Here's 
> what I'd like to be able to do:
> 
> 
> *         Import large controlled vocabularies into the app
> 
> *         Call up the app with a macro wherever I'm entering data
> 
> *         Begin typing in a term from the vocabulary, get a list of 
> suggestions for terms
> 
> *         Select a term from the list and have it paste automatically into my 
> data entry field
> 
> Ideally it would load and suggest terms quickly. I've looked around, but 
> nothing really stands out. Anyone using anything like this?
> 
> Thanks,
> David
> 
> 
> --
> David Anderson
> Systems Librarian
> Technical Services Division
> National Library of Medicine
> National Institutes of Health
> Email: [email protected]<mailto:[email protected]>
> Phone: 301-402-0033

Reply via email to