Parsing flat files in Clojure
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
;; Code for blog article at | |
;; http://www.lexicallyscoped.com/2015/01/05/parsing-flat-files-in-clojure.html | |
(ns flatfile.core | |
(:use [clojure.java.io]) | |
(:import [java.io PushbackReader])) | |
(defn load-rulebook [file] | |
(with-open [r (reader file)] | |
(read (PushbackReader. r)))) | |
;; Use this version when not include data types, other version below. | |
(comment | |
(defn extract-str [s segment] | |
(let [start (:start-pos segment) | |
end (:end-pos segment)] | |
(subs s (dec start) end))) | |
) | |
(defn match-identifier? [row ident] | |
(let [s (extract-str row ident)] | |
(= s (:id-string ident)))) | |
(defn match-rule? [row idents] | |
(every? #(match-identifier? row %) idents)) | |
(defn find-rule [row rules] | |
(first (filter #(match-rule? row (:identifiers %)) rules))) | |
(defn parse-with-rule [row rule] | |
(let [segs (:segments rule) | |
extracted (reduce #(assoc %1 (keyword (:id %2)) (extract-str row %2)) {} segs)] | |
(assoc extracted :id (:id rule)))) | |
(defn parse-file [file rules] | |
(with-open [r (reader file)] | |
(doall | |
(map #(parse-with-rule % (find-rule % rules)) (line-seq r))))) | |
(comment | |
(def rb (load-rulebook "rules/dd.clj")) | |
(def result (parse-file "data/dd.txt" rb)) | |
) | |
(defn date-formatter [format timezone] | |
(let [d (java.text.SimpleDateFormat. format)] | |
(.setTimeZone d (java.util.TimeZone/getTimeZone timezone)) | |
d)) | |
;;(.parse (date-formatter "yyyyMMdd" "Sweden") "20141230") | |
(defmulti data-type :data-type) | |
(defmethod data-type :date [s] | |
(.parse (date-formatter "yyyyMMdd" "Sweden") (:value s))) | |
(defmethod data-type :datetime [s] | |
(.parse (date-formatter "yyyyMMddHHmmss" "Sweden") (:value s))) | |
(defmethod data-type :long [s] (Long/parseLong (:value s))) | |
(defmethod data-type :default [s] (:value s)) | |
(defmethod data-type :amount [s] | |
(let [v (:value s)] | |
{:kr (Long/parseLong (subs v 1 10)) | |
:ore (Long/parseLong (subs v 10))})) | |
(defn extract-str [s segment] | |
(let [start (:start-pos segment) | |
end (:end-pos segment) | |
s (subs s (dec start) end)] | |
(data-type (assoc segment :value s)))) | |
(comment | |
(prn | |
(parse-file "data/dd.txt" (load-rulebook "rules/dd.clj"))) | |
) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment