Decoding base64 encoded file back to original format using Clojure

Question

How does one convert a file that has been base64 encoded back to its original format and write it to disk? For instance I have a pdf file which has been mime64 encoded. The file starts with:

data:application/pdf;base64,JVBER

I would like to write this out to disk in the proper format. I have tried several libraries (e.g. ring.util.codec) that decode the string into a byte-array, but if I write the resulting byte-array out to a file (using spit) the file appears corrupted.

UPDATE:

The PHP function base64_decode appears to be doing what I am looking for, as it returns a string. What is the equivalent in Java?

Have you looked on the internet for base64 tools? Or in Linux have you searched your package repository? — Kerrek SB
– Kerrek SB, Commented Aug 21, 2011 at 22:37
stackoverflow.com/questions/469695/decode-base64-data-in-java — Nicolas Buduroi
– Nicolas Buduroi, Commented Aug 21, 2011 at 22:41
I have and I also read the question referenced above. I can decode the string into a byte.array, but how do I write this to a file in a way that turns the contents into the original file format? — mac
– mac, Commented Aug 21, 2011 at 22:47

nha · Accepted Answer · 2016-02-22 15:04:08Z

In Clojure, there is data.codec (formerly in clojure-contrib).

Using Java interoperability :

with Java8, see java.util.base64
with Java6, see javax.xml.bind.DatatypeConverter/parseBase64Binary and javax.xml.bind.DatatypeConverter/printBase64Binary
with Google Guava, see com.google.common.io.BaseEncoding
with the Apache Commons Library 1.1.0, see org.apache.commons.codec.binary.Base64

So those are the helper functions I used for images when using data.codec :

(require '[clojure.data.codec.base64 :as b64-codec])

(defn write-img! [id b64]
  (clojure.java.io/copy
   (decode-str (chop-header b64))
   (java.io.File. (str "/Users/nha/tmp/" id "." (b64-ext b64)))))

(defn decode-str [s]
  (b64-codec/decode (.getBytes s)))

(defn in?
  "true if the seq coll contains the element el"
  [coll el]
  (some #(= el %) coll))

(defn b64-ext [s]
  (if-let [ext (second (first (re-seq #"data:image/(.*);base64.*" s)))]
    (if (in? ["png" "jpeg"] ext)
      ext
      (throw (Exception. (str "Unsupported extension found for image " ext))))
    (throw (Exception. (str "No extension found for image " s)))))

(defn chop-header [s]
  (nth (first (re-seq #"(data:image/.*;base64,)(.*)" s)) 2))

Nicolas Modrzyk · Accepted Answer · 2011-08-22 02:05:21Z

3

Any java library should work (here's one, from Apache Commons, here's one totally in Clojure from Clojure-contrib

I suspect the content is modified somehow, meaning bytes may be converted to string using some encoding, and then trying to read this string back to bytes using a different encoding.

The first step may be to check you have the exact same number of bytes in the file on the server side, and the file you are trying to read. Also, try to confirm the checksum (MD5) is the same.

In any case, a PDF file is a binary file, so you should NOT convert it to string anywhere, but straight bytes.

answered Aug 22, 2011 at 2:05

Nicolas Modrzyk

14.2k2 gold badges39 silver badges40 bronze badges

3 Comments

mac Over a year ago

I did check the integrity of the file and it has not been corrupted. It also can be converted using the PHP base64_decode function without any issues.

Nicolas Modrzyk Over a year ago

can you make the raw bytes available somewhere ?

mac Over a year ago

I solved it. Apparently the datauri header was messing up the decoding. If I chop off the header "data:application/pdf;base64," it works.

Collectives™ on Stack Overflow

Decoding base64 encoded file back to original format using Clojure

2 Answers 2

Comments

3 Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

3 Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related