clj-web-crawler allows you to easily crawl the web from the comfort of your newfound Lisp
dialect that runs on the JVM, Clojure. This Clojure script is a wrapper around the Apache
commons-client Java library.
; The crawl macro also accepts a client, a method and a body.
; In the body of crawl, you can examine various things like
; the status code, the response, etc.
(let [clj-ws (client "http://www.clojure.org")
home (method "/")]
(crawl clj-ws home
(println (.getStatusCode home))
(println (response-str home))))
; If you need to login to a website you can do that too and
; verify any cookies are set to validate the login
(let [site (client "http://www.example.com")
login (method "/accounts/login" :post {:login "doctor_no" :password "clojure"})]
(crawl site login
(if (assert-cookie-names site "username" "logged-in")
(println "yeah, I'm in")
(println "I can't remember my password again!"))))
; the crawl-response function is for quick and dirty things and just returns the
; response as a string from the server
(println (crawl-response "http://www.clojure.org"))
; prints the HTML of the clojure.org/api page and always defaults to a GET request
(println (crawl-response "http://www.clojure.org" "/api"))
; you can also pass in a real method object
(println (crawl-response "http://www.example.com" (method "/do_login" :post {:login "me"}))
I’ve only implemented some basic functionality to make the commons-client
more in line with functional programming. There are lots of things that
could be added to this Clojure script.
Since I’m calling commons-client under the covers I’ve using the underlying
naming conventions. The client function returns a
org.apache.commons.httpclient.HttpClient and the method function returns
an implementation of org.apache.commons.httpclient.HttpMethodBase. Please
see the commons-client API docs for reference.
- clojure-contrib
- Apache commons-client 3.1 and its dependencies (only tested with the 3.1 version)
Any comments, suggestions, improvements are welcome!