t-ob

;; Before:
;; (http/get "https://tv.eurosport.com/")
;; Error: AbstractVerifier.java:228 org.apache.http.conn.ssl.AbstractVerifier.verify
;;
;; Instead:
;; (with-connection-pool {:insecure? true} (http/get "https://tv.eurosport.com/"))
;; ignores bad certificate

(defn request*
  "HTTP request with a constrained time and body size."

t-ob

(ns ldnclj-sep-2012.core)

(defn split-seq [n s]
  (loop [ln n ls s r1 []]
    (cond
     (= ln 0) (vector r1 ls)
     :else (recur (dec ln) (rest ls) (conj r1 (first ls))))))

t-ob

Exception in thread "main" java.lang.NoSuchMethodError: org.slf4j.spi.LocationAwareLogger.log(Lorg/slf4j/Marker;Ljava/lang/String;ILjava/lang/String;[Ljava/lang/Object;Ljava/lang/Throwable;)V
	at org.eclipse.jetty.util.log.JettyAwareLogger.log(JettyAwareLogger.java:601)
	at org.eclipse.jetty.util.log.JettyAwareLogger.warn(JettyAwareLogger.java:425)
	at org.eclipse.jetty.util.log.Slf4jLog.warn(Slf4jLog.java:74)
	at org.eclipse.jetty.util.component.AbstractLifeCycle.setFailed(AbstractLifeCycle.java:199)
	at org.eclipse.jetty.util.component.AbstractLifeCycle.start(AbstractLifeCycle.java:69)
	at ring.adapter.jetty$run_jetty.invoke(jetty.clj:85)
	at ring.server.standalone$serve$fn__1551.invoke(standalone.clj:98)
	at ring.server.standalone$try_port.invoke(standalone.clj:16)
	at ring.server.standalone$serve.doInvoke(standalone.clj:95)

t-ob

(def sample-tweets
  [["1234567890" (str "[\"" "http://t.co/09hWanj8" "\",\"" "http://t.co/lozKUbnY" "\"]")]
   ["1029384756" (str "[\"" "http://t.co/1O8YAA21" "\",\"" "http://t.co/ONYmi6gK" "\"]")]])

;; tap/hbase
;; (defn hbase [table-name key-field column-family & value-fields]
;;   (let [scheme (HBaseScheme. (w/fields key-field) column-family (w/fields value-fields))]
;;     (HBaseTap. table-name scheme)))

;; This works

t-ob

select count(*), tweet_id, content, max(screen_name), max(created_at), min(screen_name), min(created_at)
from tweets_with_id
group by tweet_id
having count(*) = 2

t-ob

import org.apache.hadoop.hbase.filter.CompareFilter
import org.apache.hadoop.hbase.filter.SingleColumnValueFilter
import org.apache.hadoop.hbase.filter.SubstringComparator
import org.apache.hadoop.hbase.util.Bytes
scan 'tweets', { COLUMNS => 'base:content', FILTER => SingleColumnValueFilter.new(Bytes.toBytes('base'),Bytes.toBytes('content'), CompareFilter::CompareOp.valueOf('EQUAL'),SubstringComparator.new('burn')) }

t-ob

#!/bin/bash
NOW=$(date +"%Y-%m-%d-%T")

USERS10="users-10.$NOW.csv"
EVENTS10="events-10.$NOW.csv"
USERS20="users-20.$NOW.csv"
EVENTS20="events-20.$NOW.csv"

echo $USERS10;

t-ob

(defn limited-output
  [keywords-str limit]
  (let [tweets-tap (tweets-new)
        urls-tap (short-urls)
        jdbc-sink db/jdbc-tap]
    (?- (stdout)
        (c/first-n
         (<- [ ?args ]
             ; query
         limit))))

t-ob

(defn tweets
  []
  (let [scheme (HBaseScheme. (Fields. (into-array String ["id"]))
                             (into-array String ["base" "raw"])
                             (into-array [(Fields. (into-array String ["tweet_id"
                                                          "screen_name"
                                                          "content"
                                                          "created_at"
                                                          "urls"]))
                                          (Fields. (into-array String ["topsy_url"]))]))

t-ob

(def short-urls
  [["http://t.co/yERArQn0"]
   ["http://t.co/gI8TjreI"]
   ["http://t.co/CBsucpNm"]
   ["http://t.co/F74GG1oN"]
   ["http://t.co/hyoXObbU"]])

(def longified-urls
  [["http://t.co/yERArQn0" "http://news.sky.com/story/964624/olympic-lanes-open-traffic-delays-in-london"]
   ["http://t.co/CBsucpNm" "http://media.tumblr.com/tumblr_m65uu5UsrQ1r7h1lt.gif"]