Userscripts.org

This is my fix for when the image has extra attributes (i.e. border) between "img" and "src".
Just modify the pattern and add "[^>]*" before "src",
e.g. /img [^>]*src=\"([^\"]+)\"/gi
The reason I didn't use .* is that some posts have multiple images on the same line and I don't think javascript supports non greedy matches.

I've also added a few modifications to suit my needs.
I'm looking for rentals an San Francisco's North Bay. There are a few towns I'm sure I don't want to move to so I've added a list of ignoredTowns and the code yanks those out of the page.

Also, some realtors include logos and other crap. I'm filtering those images out too, but I need to add to that list every now and then (i.e. requires some maintenance)

The last thing I'm still working on is to filter out over posting by some realtors.
The only consistent commonality between the posts are the images, so I'm storing their md5 and removing subsequent matches.

All this is hardcoded. I might release my fixes if/when I expose all that in the UI.
That is, if there's any interest.

You missed a place when adding the regex pattern. You need to also add it to line 89.
s = s.replace(/img [^>]*src=/g,"");

I also added the space between the img and the [^ on line 73.
if (m = details.responseText.match(/img [^>]*src=\"([^\"]+)\"/gi)) {

Can thee be an option to not show pictures with width larger than 2x the height (obvious banner ads, etc)? that way its only photographs being shown?

I'm currently living in Seoul Korean and I changed the js's ".org"s to ".co.kr" to make it work over here.

Something like this might make it more universal?
if (link.href && link.href.match(/.*craigslist.*\/\d+\.html$/)) {
return loc.match(/http:\/\/\w+\.craigslist.*\/\w+\/?/);