extract.clj

;; This buffer is for notes you don't want to save, and for Lisp evaluation.
;; If you want to create a file, visit that file with C-x C-f,
;; then enter the text in that file's own buffer.

(defn extract-render-listener
  "A RenderListener implementation that extracts images from a PDF and
writes them to disk."
  [path]
  (reify RenderListener
    (renderImage [_ render-info]
      (let [image (.getImage render-info)
            ref-number (-> render-info (.getRef) (.getNumber))
            file-type (.getFileType image)
            file-name (str path ref-umber "." file-type)]
        (with-open [os (clojure.java.io/output-stream file-name)]
          (.write os (.getImageAsBytes image))
          (.flush os))))
    (beginTextBlock [_] nil)
    (endTextBlock [_] nil)
    (renderText [_ render-info] nil)))
 
(defn matrix-render-listener
  "A RenderListener implementation that updates shared mutable state with
nested vectors of image coordinate data."
  [page]
  (reify RenderListener
    (renderImage [_ render-info]
      (let [matrix (.getImageCTM render-info)
            [x y w h :as coords] (for [i [6 7 0 4]] (.get matrix i))
            pagekey (keyword (str page))]
        (dosync (alter coords update-in [pagekey] conj coords))))
    (beginTextBlock [_] nil)
    (endTextBlock [_] nil)
    (renderText [_ render-info] nil)))