gistfile1.clj

(defn vec-cookies [s]
  "Return vector of cookies and their values."
  (vec
   (for [triplet (re-seq cookies s)]
     [(second triplet) (last triplet)]))

(defmapop logl [s]
  "Vector consisting of log line components, for a single line."
  [(re-find remote-addr s)
  (second (re-find timestamp s))
  (second (re-find method-url s))
  (last (re-find method-url s))
  (second (re-find response-status s))
  (second (re-find upstream-time s))
  (second (re-find request-time s))
  (second (re-find host-upstream-addr s))
  (last (re-find host-upstream-addr s))
  (vec-cookies s)])

(defn logdata [inpath]
  "Return parsed logfile from HFS tap."
  ;; Create a textline tap on the target logfile.
  (let [src (hfs-textline inpath)
        fields ["?remote-addr" "?timestamp" "?method-url1" "?method-url2" "?response-status"
                "?upstream-time" "?request-time" "?host-upstream-addr1" "?host-upstream-addr2"
                "?cookies"]]
    ;; Use another textline tap as output target.
    (<- fields (src ?c) (logl ?c :>> fields) (:distinct false))))

(defmapcatop explode [l] l)

(defn cookie-line [inpath]
  "Get cookie data from the logfile."
  (let [data (select-fields (logdata inpath) ["?cookie"])]
    (<- [?key ?val]
        (data ?cookie)
        (explode ?cookie :> ?key ?val)
        (:distinct false))))


(let [cl (cookie-line "resources/snippet.log")]
  (?<- (stdout) [?key ?count]
       (cl ?key _)
       (c/count ?count)))