Search:

Well for one someone could just spam your error log by running javascript errors in the console. You could also end up with errors in extensions that people are running (nothing to do with your app)...

HTML is the basic building block of websites and web apps. Everything the browser renders (pretty much) must be, in the end, HTML. HTML5 is the latest version, learn it.

CSS is used to apply...

As was said: they use webkit, so the Document Object Model (DOM) API will be the same.

If your site is a static one I guess you could use a crawler to download it starting at the site index and just re-upload that, deleting anything else.

I wonder if it would be a better idea to create a mirror table under MyISAM holding only the fields that need to be full-text searched. This could also concatenate all the tags and habtm data into...

I don't know about free hosting, but nearlyfreespeech.net is really cheap and I use them for non-critical projects (otherwise I use ByteMark which is upward of £15/month).

NFSN are great.

Is sendmail configured on your machine?

Dear all,

I have quite a complex problem to deal with:

My database uses a lot of InnoDB related tables. I have a database with objects that need to be searched for a) full text matches b)...

if it is a simple regular polygon this is simple (I am assuming it is since you're using a radius).
Basically, look at the polygon like a circle. Vertices are placed regularly around the circle at...

Does anyone know why Sphinx is behaving like this? Maybe I have some weird config options or something? Please?

Hi all,

I'm having a problem with Sphinx Search where if I have a document with a title "Laser applications" and I use the command


search -i per_projects -a laser -q


I get no results yet...

You're welcome. Bonne chance.

Oh, I thought you wanted a specifically regex way of doing this, apologies. I've actually used http://simplehtmldom.sourceforge.net/ when I've wanted to parse potentially incorrect/fragmented html...

I wasn't serious, don't worry =]

Actually, better than that would be


<([A-Za-z]+)(\s[^>]+)?>.*<\/\1>


Which is far simpler, and I only thought of when posting.

Something like


<([A-Za-z]+)(\s([A-Za-z0-9"'_\-\ =]+))?>.*<\/\1>


Is a good start, though it has limitations like not matching anything with linebreaks. You could also make it more advanced...

Why is London not on this list?
London and the UK are quite different. I am, of course, from the former.

Okay, thanks.

After some testing it looks like JS is a bit limited in this respect because of the lack of a proper OO model.

It looks like I'm going to just pass the class and method names in...

@Kor:

on the first point: I don't really expect any beginners, on the second: I'm not sure that this is true because there are plenty of "experts" writing some really fairly impressive things...

bump

Okay thanks. I'll try standard MyISAM splitting and if I don't like it I'll try Sphinx.

Lucene and Sphinx are fairly equivalent, no? The difference being that Lucene is Java and Sphinx is in a C derivative which should make it faster shouldn't it? Or at least hog less memory?

Okay, thanks. I only really envision my queries using a fulltext search on the descriptions table along with a join to the main table and conditions involving related n:n tables. Shouldn't be too...

And the thing about mem alloc issues with MyISAM + InnoDB? Or is this actually fallacious? I wasn't sure.

@Mindzai:

I called it a "bad" solution a) because of the need to synchronise data across two tables (which seems, to me, like an unnecessary overhead and potential pain if something goes awry) and...