Mailing-List: contact dev-help@pdfbox.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@pdfbox.apache.org
Date: Fri, 9 Mar 2012 07:11:57 +0000 (UTC)
From: =?utf-8?Q?Andreas_Lehmk=C3=BChler_=28Commented=29_=28JIRA=29?=
 <jira@apache.org>
To: dev@pdfbox.apache.org
Message-ID: 
 <871541213.42562.1331277117464.JavaMail.tomcat@hel.zones.apache.org>
In-Reply-To: 
 <1739338213.5666.1329829594316.JavaMail.tomcat@hel.zones.apache.org>
Subject: [jira] [Commented] (PDFBOX-1232) FlateDecoder in stream mode
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable


    [ https://issues.apache.org/jira/browse/PDFBOX-1232?page=3Dcom.atlassia=
n.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=3D132=
25903#comment-13225903 ]=20

Andreas Lehmk=C3=BChler commented on PDFBOX-1232:
--------------------------------------------

Just to be sure, what is the expected result of the decompression of the gi=
ven 5 value stream?
               =20
> FlateDecoder in stream mode
> ---------------------------
>
>                 Key: PDFBOX-1232
>                 URL: https://issues.apache.org/jira/browse/PDFBOX-1232
>             Project: PDFBox
>          Issue Type: Bug
>            Reporter: Dave Smith
>
> The zlib (the unlying spec for Flate compression) does not require an Z_S=
TREAM_END to terminate the compression. The Java InflateInputStream is real=
ly assuming that you are reading a zip or gzip file which will always have =
a Z_STREAM_END (Z_STREAM_END is a constant in the zlib library which Java c=
alls natively) . So the following chunk decodes fine using  the jcraft zlib=
 decoder, but fails using the InflateInputStream.
> 3 0 obj
> <<
> /Type /XObject
> /Subtype /Form
> /FormType 1
> /Resources << /Font 4 0 R
> /ProcSet [/PDF /ImageC /Text]>>
> /BBox [0 0 595 842]
> /Matrix [1 0 0 1 0 0]
> /Filter /FlateDecode
> /Length 5 >>
> stream
> H<89>^C^@
> endstream
> endobj
> The blob is 72, -119, 3, 0, 13 decimal. It decodes to an empty string.
> The fix is to use Inflater and check to see if it has consumed all of the=
 input buffer and make sure it has nothing to write into the output buffer.
> protected ByteArrayOutputStream decompress(InputStream in)
>       throws IOException, DataFormatException
>   {
>       ByteArrayOutputStream out =3D new ByteArrayOutputStream();
>       byte buf[] =3D new byte[1000];
>       Inflater inflater =3D new Inflater();
>       int read =3D in.read(buf);
>       if(read =3D=3D 0)
>       {
>               return out;
>       }
>       inflater.setInput(buf,0,read);
>       byte res[] =3D new byte[1000];
>       while(true)
>       {
>               int resRead =3D inflater.inflate(res);
>               if(resRead !=3D0)
>               {
>                       out.write(res,0,resRead);
>                       continue;
>               }
>               if(inflater.finished() || inflater.needsDictionary() ||  in=
.available()=3D=3D0)
>               {
>                       out.close();
>                       return out;
>               }
>              read =3D in.read(buf);
>              inflater.setInput(buf,0,read);
>    =20
>       }
>   }
> We then need to change FlateFilter.decode(InputStream compressedData, Out=
putStream result,
> COSDictionary options, int filterIndex )
> to look like ...
>  if (compressedData.available() > 0)
>           {
>               try
>               {
>                       baos =3D  decompress(compressedData);
>               }
> if (predictor=3D=3D-1 || predictor =3D=3D 1 )
>               {
>                  result.write(baos.toByteArray());
>               }
> else
> {
>  use the bytearrayoutput stream as before ...
> }

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrato=
rs: https://issues.apache.org/jira/secure/ContactAdministrators!default.jsp=
a
For more information on JIRA, see: http://www.atlassian.com/software/jira