company identification.

src/clj/auto_ap/parse.clj

@@ -2,30 +2,10 @@
   (:require [clojure.java.io :as io]
             [clojure.string :as str]
             [clojure.java.shell :as sh]
-            [auto-ap.parse.excel :as excel]))
+            [auto-ap.parse.excel :as excel]
+            [auto-ap.parse.templates :as t]))
 
-(def templates
-  [{:vendor "CHFW"
-    :keywords [#"CHEF'S WAREHOUSE"]
-    :extract {:total #"2 WKS C\.C\.\s+([\d.,]+)"
-              :customer-identifier #"\n([A-Z][A-Z ]+)\s{2,}"
-              :date #"\s+([0-9]+/[0-9]+/[0-9]+)"
-              :invoice-number #"\s+[0-9]+/[0-9]+/[0-9]+\s+([0-9]+)"}}
 
-   {:vendor "GGM"
-    :keywords [#"Golden Gate Meat"]
-    :extract {:total #"Invoice Total\:\s+\$([\d.,]+)"
-              :customer-identifier #"Bill To\s*:\s*([\w ]+)\s{2,}"
-              :date #"Printed:\s+([0-9]+/[0-9]+/[0-9]+)"
-              :invoice-number #"Invoice\s+[^\n]+\n[^\n]+\n\s+([0-9]+)"}}
-
-   {:vendor "CINTAS"
-    :keywords [#"CINTAS CORPORATION"]
-     :extract {:invoice-number #"INVOICE\s#\s+([\d.,]+)"
-               :customer-identifier #"BILL TO\s*:\s{2,}([\w ]+)\s{2,}"
-               :date #"INVOICE DATE\s*\n.*\s+([0-9]+/[0-9]+/[0-9]+)"
-               :total #"INVOICE TOTAL\s+([0-9.]+)"}
-    :multi #"\f\f"}])
 
 (defn template-applies? [text {:keys [keywords]}]
   (every? #(re-find % text) keywords))
@@ -45,13 +25,13 @@
             {:vendor (:vendor template)}))]))
 
 (defn parse [text]
-  (->> templates
+  (->> t/pdf-templates
        (filter (partial template-applies? text))
        first
        (extract-template text)))
 
 
-(defmulti parse-file (fn [file filename]  (last (str/split filename #"\." ))))
+(defmulti parse-file (fn [file filename]  (.toLowerCase (last (str/split filename #"\." )))))
 
 (defmethod parse-file
   "pdf"