Current Status: Usable. Please file bugs!
HandsomeSoup is the library I wish I had when I started parsing HTML in Haskell.
It is built on top of HXT and adds a few functions that make is easier to work with HTML.
Most importantly, it adds CSS selectors to HXT. The goal of HandsomeSoup is to be a complete CSS2 selector parser for HXT.
cabal install HandsomeSoup
Nokogiri, the HTML parser for Ruby, has an example showing how to scrape Google search results. This is easy in HandsomeSoup:
main = do doc <- fromUrl "http://www.google.com/search?q=egon+schiele" links <- runX $ doc >>> css "h3.r a" ! "href" mapM_ putStrLn links
What can HandsomeSoup do for you?
Easily parse an online page using
doc <- fromUrl "http://example.com"
Or a local page using
contents <- readFile [filename] let doc = parseHtml contents
Easily extract elements using
Here are some valid selectors:
doc <<< css "a" doc <<< css "*" doc <<< css "a#link1" doc <<< css "a.foo" doc <<< css "p > a" doc <<< css "p strong" doc <<< css "#container h1" doc <<< css "img[width]" doc <<< css "img[width=400]" doc <<< css "a[class~=bar]" doc <<< css "a:first-child"
Easily get attributes using
doc <<< css "img" ! "src" doc <<< css "a" ! "href"
Find Haddock docs on Hackage.
I also wrote The Complete Guide To Parsing HXT With Haskell.
Made by Adit.