diff options
| author | tknall <tknall@7b5415b0-85f9-ee4d-85bd-d5d0c3b42d1c> | 2006-12-01 12:20:24 +0000 |
|---|---|---|
| committer | tknall <tknall@7b5415b0-85f9-ee4d-85bd-d5d0c3b42d1c> | 2006-12-01 12:20:24 +0000 |
| commit | 6025b6016517c6d898d8957d1d7e03ba71431912 (patch) | |
| tree | b15bd6fa5ffe9588a9bca3f2b8a7e358f83b6eba /src/main/java/org/pdfbox/pdfparser | |
| parent | d2c77e820ab4aba8235d71275755021347b3ad10 (diff) | |
| download | pdf-as-3-6025b6016517c6d898d8957d1d7e03ba71431912.tar.gz pdf-as-3-6025b6016517c6d898d8957d1d7e03ba71431912.tar.bz2 pdf-as-3-6025b6016517c6d898d8957d1d7e03ba71431912.zip | |
Initial import of release 2.2.REL-2.2@923
git-svn-id: https://joinup.ec.europa.eu/svn/pdf-as/trunk@4 7b5415b0-85f9-ee4d-85bd-d5d0c3b42d1c
Diffstat (limited to 'src/main/java/org/pdfbox/pdfparser')
| -rw-r--r-- | src/main/java/org/pdfbox/pdfparser/BaseParser.java | 1369 | ||||
| -rw-r--r-- | src/main/java/org/pdfbox/pdfparser/PDFObjectStreamParser.java | 137 | ||||
| -rw-r--r-- | src/main/java/org/pdfbox/pdfparser/PDFParser.java | 557 | ||||
| -rw-r--r-- | src/main/java/org/pdfbox/pdfparser/PDFStreamParser.java | 403 | ||||
| -rw-r--r-- | src/main/java/org/pdfbox/pdfparser/PDFXref.java | 96 | ||||
| -rw-r--r-- | src/main/java/org/pdfbox/pdfparser/package.html | 9 |
6 files changed, 2571 insertions, 0 deletions
diff --git a/src/main/java/org/pdfbox/pdfparser/BaseParser.java b/src/main/java/org/pdfbox/pdfparser/BaseParser.java new file mode 100644 index 0000000..3937025 --- /dev/null +++ b/src/main/java/org/pdfbox/pdfparser/BaseParser.java @@ -0,0 +1,1369 @@ +/**
+ * Copyright (c) 2003-2005, www.pdfbox.org
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright notice,
+ * this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright notice,
+ * this list of conditions and the following disclaimer in the documentation
+ * and/or other materials provided with the distribution.
+ * 3. Neither the name of pdfbox; nor the names of its
+ * contributors may be used to endorse or promote products derived from this
+ * software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+ * DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE FOR ANY
+ * DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+ * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON
+ * ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+ * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ * http://www.pdfbox.org
+ *
+ */
+package org.pdfbox.pdfparser;
+
+import java.io.BufferedInputStream;
+import java.io.InputStream;
+import java.io.IOException;
+import java.io.OutputStream;
+import java.io.RandomAccessFile;
+
+import java.util.ArrayList;
+import java.util.List;
+
+import org.pdfbox.io.ByteArrayPushBackInputStream;
+import org.pdfbox.io.PushBackInputStream;
+
+import org.pdfbox.cos.COSArray;
+import org.pdfbox.cos.COSBase;
+import org.pdfbox.cos.COSBoolean;
+import org.pdfbox.cos.COSDictionary;
+import org.pdfbox.cos.COSDocument;
+import org.pdfbox.cos.COSInteger;
+import org.pdfbox.cos.COSName;
+import org.pdfbox.cos.COSNull;
+import org.pdfbox.cos.COSNumber;
+import org.pdfbox.cos.COSObject;
+import org.pdfbox.cos.COSStream;
+import org.pdfbox.cos.COSString;
+
+import org.pdfbox.persistence.util.COSObjectKey;
+import org.apache.log4j.Logger;
+
+/**
+ * This class is used to contain parsing logic that will be used by both the
+ * PDFParser and the COSStreamParser.
+ *
+ * @author Ben Litchfield (ben@benlitchfield.com)
+ * @version $Revision: 1.57 $
+ */
+public abstract class BaseParser
+{
+ private static Logger log = Logger.getLogger(BaseParser.class);
+
+ /**
+ * This is a byte array that will be used for comparisons.
+ */
+ public static final byte[] ENDSTREAM =
+ new byte[] {101,110,100,115,116,114,101,97,109};//"endstream".getBytes( "ISO-8859-1" );
+
+ /**
+ * This is a byte array that will be used for comparisons.
+ */
+ public static final String DEF = "def";
+
+ /**
+ * This is the stream that will be read from.
+ */
+ //protected PushBackByteArrayStream pdfSource;
+ protected PushBackInputStream pdfSource;
+
+ /**
+ * moved xref here, is a persistence construct
+ * maybe not needed anyway when not read from behind with delayed
+ * access to objects.
+ */
+ private List xrefs = new ArrayList();
+
+ private COSDocument document;
+
+ /**
+ * Constructor.
+ *
+ * @param input The input stream to read the data from.
+ *
+ * @throws IOException If there is an error reading the input stream.
+ */
+ public BaseParser( InputStream input) throws IOException
+ {
+ //pdfSource = new PushBackByteArrayStream( input );
+ pdfSource = new PushBackInputStream( new BufferedInputStream( input, 16384 ), 4096 );
+ }
+
+ /**
+ * Constructor.
+ *
+ * @param input The array to read the data from.
+ *
+ * @throws IOException If there is an error reading the byte data.
+ */
+ protected BaseParser(byte[] input) throws IOException
+ {
+ pdfSource = new ByteArrayPushBackInputStream(input);
+ }
+
+ /**
+ * Set the document for this stream.
+ *
+ * @param doc The current document.
+ */
+ public void setDocument( COSDocument doc )
+ {
+ document = doc;
+ }
+
+ private static boolean isHexDigit(char ch)
+ {
+ return (ch >= '0' && ch <= '9') ||
+ (ch >= 'a' && ch <= 'f') ||
+ (ch >= 'A' && ch <= 'F');
+ // the line below can lead to problems with certain versions of the IBM JIT compiler
+ // (and is slower anyway)
+ //return (HEXDIGITS.indexOf(ch) != -1);
+ }
+
+ /**
+ * This will parse a PDF dictionary value.
+ *
+ * @return The parsed Dictionary object.
+ *
+ * @throws IOException If there is an error parsing the dictionary object.
+ */
+ private COSBase parseCOSDictionaryValue() throws IOException
+ {
+
+ if( log.isDebugEnabled() )
+ {
+ log.debug("parseCOSDictionaryValue() " + pdfSource );
+ }
+ COSBase retval = null;
+ COSBase number = parseDirObject();
+ skipSpaces();
+ char next = (char)pdfSource.peek();
+ if( next >= '0' && next <= '9' )
+ {
+ COSBase generationNumber = parseDirObject();
+ skipSpaces();
+ char r = (char)pdfSource.read();
+ if( r != 'R' )
+ {
+ throw new IOException( "expected='R' actual='" + r + "' " + pdfSource );
+ }
+ COSObjectKey key = new COSObjectKey(((COSInteger) number).intValue(),
+ ((COSInteger) generationNumber).intValue());
+ retval = document.getObjectFromPool(key);
+ }
+ else
+ {
+ retval = number;
+ }
+ return retval;
+ }
+
+ /**
+ * This will parse a PDF dictionary.
+ *
+ * @return The parsed dictionary.
+ *
+ * @throws IOException IF there is an error reading the stream.
+ */
+ protected COSDictionary parseCOSDictionary() throws IOException
+ {
+ if( log.isDebugEnabled() )
+ {
+ log.debug("parseCOSDictionary() " + pdfSource );
+ }
+ char c = (char)pdfSource.read();
+ if( c != '<')
+ {
+ throw new IOException( "expected='<' actual='" + c + "'" );
+ }
+ c = (char)pdfSource.read();
+ if( c != '<')
+ {
+ throw new IOException( "expected='<' actual='" + c + "' " + pdfSource );
+ }
+ skipSpaces();
+ COSDictionary obj = new COSDictionary();
+ boolean done = false;
+ while( !done )
+ {
+ skipSpaces();
+ c = (char)pdfSource.peek();
+ if( c == '>')
+ {
+ done = true;
+ }
+ else
+ {
+ COSName key = parseCOSName();
+ COSBase value = parseCOSDictionaryValue();
+ skipSpaces();
+ if( ((char)pdfSource.peek()) == 'd' )
+ {
+ //if the next string is 'def' then we are parsing a cmap stream
+ //and want to ignore it, otherwise throw an exception.
+ String potentialDEF = readString();
+ if( !potentialDEF.equals( DEF ) )
+ {
+ pdfSource.unread( potentialDEF.getBytes() );
+ }
+ else
+ {
+ skipSpaces();
+ }
+ }
+
+ if( value == null )
+ {
+ throw new IOException("Bad Dictionary Declaration " + pdfSource );
+ }
+ obj.setItem( key, value );
+ }
+ }
+ char ch = (char)pdfSource.read();
+ if( ch != '>' )
+ {
+ throw new IOException( "expected='>' actual='" + ch + "'" );
+ }
+ ch = (char)pdfSource.read();
+ if( ch != '>' )
+ {
+ throw new IOException( "expected='>' actual='" + ch + "'" );
+ }
+ if( log.isDebugEnabled() )
+ {
+ log.debug("parseCOSDictionary() done peek='" + pdfSource.peek() + "'" );
+ }
+ return obj;
+ }
+
+ /**
+ * This will read a COSStream from the input stream.
+ *
+ * @param file The file to write the stream to when reading.
+ * @param dic The dictionary that goes with this stream.
+ *
+ * @return The parsed pdf stream.
+ *
+ * @throws IOException If there is an error reading the stream.
+ */
+ protected COSStream parseCOSStream( COSDictionary dic, RandomAccessFile file ) throws IOException
+ {
+ if( log.isDebugEnabled() )
+ {
+ log.debug("parseCOSStream() " + pdfSource );
+ }
+ COSStream stream = new COSStream( dic, file );
+ OutputStream out = null;
+ try
+ {
+ String streamString = readString();
+ //long streamLength;
+
+ if (!streamString.equals("stream"))
+ {
+ throw new IOException("expected='stream' actual='" + streamString + "'");
+ }
+
+ //PDF Ref 3.2.7 A stream must be followed by either
+ //a CRLF or LF but nothing else.
+
+ int whitespace = pdfSource.read();
+
+ //see brother_scan_cover.pdf, it adds whitespaces
+ //after the stream but before the start of the
+ //data, so just read those first
+ while (whitespace == 0x20)
+ {
+ whitespace = pdfSource.read();
+ }
+
+ if( whitespace == 0x0D )
+ {
+ whitespace = pdfSource.read();
+ if( whitespace != 0x0A )
+ {
+ pdfSource.unread( whitespace );
+ //The spec says this is invalid but it happens in the real
+ //world so we must support it.
+ //throw new IOException("expected='0x0A' actual='0x" +
+ // Integer.toHexString(whitespace) + "' " + pdfSource);
+ }
+ }
+ else if (whitespace == 0x0A)
+ {
+ //that is fine
+ }
+ else
+ {
+ //we are in an error.
+ //but again we will do a lenient parsing and just assume that everything
+ //is fine
+ pdfSource.unread( whitespace );
+ //throw new IOException("expected='0x0D or 0x0A' actual='0x" +
+ //Integer.toHexString(whitespace) + "' " + pdfSource);
+
+ }
+
+
+ COSBase streamLength = dic.getDictionaryObject(COSName.LENGTH);
+ long length = -1;
+ if( streamLength instanceof COSNumber )
+ {
+ length = ((COSNumber)streamLength).intValue();
+ }
+ else if( streamLength instanceof COSObject &&
+ ((COSObject)streamLength).getObject() instanceof COSNumber )
+ {
+ length = ((COSNumber)((COSObject)streamLength).getObject()).intValue();
+ }
+
+ //length = -1;
+ //streamLength = null;
+
+ //Need to keep track of the
+ out = stream.createFilteredStream( streamLength );
+ String endStream = null;
+ //the length is wrong in some pdf documents which means
+ //that PDFBox must basically ignore it in order to be able to read
+ //the most number of PDF documents. This of course is a penalty hit,
+ //maybe I could implement a faster parser.
+ /**if( length != -1 )
+ {
+ byte[] buffer = new byte[1024];
+ int amountRead = 0;
+ int totalAmountRead = 0;
+ while( amountRead != -1 && totalAmountRead < length )
+ {
+ int maxAmountToRead = Math.min(buffer.length, (int)(length-totalAmountRead));
+ amountRead = pdfSource.read(buffer,0,maxAmountToRead);
+ totalAmountRead += amountRead;
+ if( amountRead != -1 )
+ {
+ out.write( buffer, 0, amountRead );
+ }
+ }
+ }
+ else
+ {**/
+ readUntilEndStream( out );
+ /**}*/
+ skipSpaces();
+ endStream = readString();
+
+ if (!endStream.equals("endstream"))
+ {
+ readUntilEndStream( out );
+ endStream = readString();
+ if( !endStream.equals( "endstream" ) )
+ {
+ throw new IOException("expected='endstream' actual='" + endStream + "' " + pdfSource);
+ }
+ }
+ }
+ finally
+ {
+ if( out != null )
+ {
+ out.close();
+ }
+ }
+ if( log.isDebugEnabled() )
+ {
+ log.debug("parseCOSStream() done" );
+ }
+ return stream;
+ }
+
+ private void readUntilEndStream( OutputStream out ) throws IOException
+ {
+ int currentIndex = 0;
+ int byteRead = 0;
+ //this is the additional bytes buffered but not written
+ int additionalBytes=0;
+ byte[] buffer = new byte[ENDSTREAM.length+additionalBytes];
+ int writeIndex = 0;
+ while(!cmpCircularBuffer( buffer, currentIndex, ENDSTREAM ) && byteRead != -1 )
+ {
+ writeIndex = currentIndex - buffer.length;
+ if( writeIndex >= 0 )
+ {
+ out.write( buffer[writeIndex%buffer.length] );
+ }
+ byteRead = pdfSource.read();
+ buffer[currentIndex%buffer.length] = (byte)byteRead;
+ currentIndex++;
+ }
+
+ //we want to ignore the end of the line data when reading a stream
+ //so will make an attempt to ignore it.
+ /*writeIndex = currentIndex - buffer.length;
+ if( buffer[writeIndex%buffer.length] == 13 &&
+ buffer[(writeIndex+1)%buffer.length] == 10 )
+ {
+ //then ignore the newline before the endstream
+ }
+ else if( buffer[(writeIndex+1)%buffer.length] == 10 )
+ {
+ //Then first byte is data, second byte is newline
+ out.write( buffer[writeIndex%buffer.length] );
+ }
+ else
+ {
+ out.write( buffer[writeIndex%buffer.length] );
+ out.write( buffer[(writeIndex+1)%buffer.length] );
+ }*/
+
+ /**
+ * Old way of handling newlines before endstream
+ for( int i=0; i<additionalBytes; i++ )
+ {
+ writeIndex = currentIndex - buffer.length;
+ if( writeIndex >=0 &&
+ //buffer[writeIndex%buffer.length] != 10 &&
+ buffer[writeIndex%buffer.length] != 13 )
+ {
+ out.write( buffer[writeIndex%buffer.length] );
+ }
+ currentIndex++;
+ }
+ */
+ pdfSource.unread( ENDSTREAM );
+
+ }
+
+ /**
+ * This basically checks to see if the next compareTo.length bytes of the
+ * buffer match the compareTo byte array.
+ */
+ private boolean cmpCircularBuffer( byte[] buffer, int currentIndex, byte[] compareTo )
+ {
+ int cmpLen = compareTo.length;
+ int buflen = buffer.length;
+ boolean match = true;
+ int off = currentIndex-cmpLen;
+ if( off < 0 )
+ {
+ match = false;
+ }
+ for( int i=0; match && i<cmpLen; ++i )
+ {
+ match = buffer[(off+i)%buflen] == compareTo[i];
+ }
+ return match;
+ }
+
+ /**
+ * This will parse a PDF string.
+ *
+ * @return The parsed PDF string.
+ *
+ * @throws IOException If there is an error reading from the stream.
+ */
+ protected COSString parseCOSString() throws IOException
+ {
+ if( log.isDebugEnabled() )
+ {
+ log.debug("parseCOSString() " + pdfSource );
+ }
+ char nextChar = (char)pdfSource.read();
+ COSString retval = new COSString();
+ char openBrace;
+ char closeBrace;
+ if( nextChar == '(' )
+ {
+ openBrace = '(';
+ closeBrace = ')';
+ }
+ else if( nextChar == '<' )
+ {
+ openBrace = '<';
+ closeBrace = '>';
+ }
+ else
+ {
+ throw new IOException( "parseCOSString string should start with '(' or '<' and not '" +
+ nextChar + "' " + pdfSource );
+ }
+
+ //This is the number of braces read
+ //
+ int braces = 1;
+ int c = pdfSource.read();
+ while( braces > 0 && c != -1)
+ {
+ char ch = (char)c;
+ int nextc = -2; // not yet read
+ //if( log.isDebugEnabled() )
+ //{
+ // log.debug( "Parsing COSString character '" + c + "' code=" + (int)c );
+ //}
+
+ if(ch == closeBrace)
+ {
+ braces--;
+ byte[] nextThreeBytes = new byte[3];
+ int amountRead = pdfSource.read(nextThreeBytes);
+
+ //lets handle the special case seen in Bull River Rules and Regulations.pdf
+ //The dictionary looks like this
+ // 2 0 obj
+ // <<
+ // /Type /Info
+ // /Creator (PaperPort http://www.scansoft.com)
+ // /Producer (sspdflib 1.0 http://www.scansoft.com)
+ // /Title ( (5)
+ // /Author ()
+ // /Subject ()
+ //
+ // Notice the /Title, the braces are not even but they should
+ // be. So lets assume that if we encounter an this scenario
+ // <end_brace><new_line><opening_slash> then that
+ // means that there is an error in the pdf and assume that
+ // was the end of the document.
+ if( amountRead == 3 )
+ {
+ if( nextThreeBytes[0] == 0x0d &&
+ nextThreeBytes[1] == 0x0a &&
+ nextThreeBytes[2] == 0x2f )
+ {
+ braces = 0;
+ }
+ }
+ pdfSource.unread( nextThreeBytes, 0, amountRead );
+ if( braces != 0 )
+ {
+ retval.append( ch );
+ }
+ }
+ else if( ch == openBrace )
+ {
+ braces++;
+ retval.append( ch );
+ }
+ else if( ch == '\\' )
+ {
+ //patched by ram
+ char next = (char)pdfSource.read();
+ switch(next)
+ {
+ case 'n':
+ retval.append( '\n' );
+ break;
+ case 'r':
+ retval.append( '\r' );
+ break;
+ case 't':
+ retval.append( '\t' );
+ break;
+ case 'b':
+ retval.append( '\b' );
+ break;
+ case 'f':
+ retval.append( '\f' );
+ break;
+ case '(':
+ case ')':
+ case '\\':
+ retval.append( next );
+ break;
+ case 10:
+ case 13:
+ //this is a break in the line so ignore it and the newline and continue
+ c = pdfSource.read();
+ while( isEOL(c) && c != -1)
+ {
+ c = pdfSource.read();
+ }
+ nextc = c;
+ break;
+ case '0':
+ case '1':
+ case '2':
+ case '3':
+ case '4':
+ case '5':
+ case '6':
+ case '7':
+ {
+ StringBuffer octal = new StringBuffer();
+ octal.append( next );
+ c = pdfSource.read();
+ char digit = (char)c;
+ if( digit >= '0' && digit <= '7' )
+ {
+ octal.append( digit );
+ c = pdfSource.read();
+ digit = (char)c;
+ if( digit >= '0' && digit <= '7' )
+ {
+ octal.append( digit );
+ }
+ else
+ {
+ nextc = c;
+ }
+ }
+ else
+ {
+ nextc = c;
+ }
+
+ int character = 0;
+ try
+ {
+ character = Integer.parseInt( octal.toString(), 8 );
+ }
+ catch( NumberFormatException e )
+ {
+ throw new IOException( "Error: Expected octal character, actual='" + octal + "'" );
+ }
+ retval.append( character );
+ break;
+ }
+ default:
+ {
+ retval.append( '\\' );
+ retval.append( next );
+ //another ficken problem with PDF's, sometimes the \ doesn't really
+ //mean escape like the PDF spec says it does, sometimes is should be literal
+ //which is what we will assume here.
+ //throw new IOException( "Unexpected break sequence '" + next + "' " + pdfSource );
+ }
+ }
+ }
+ else
+ {
+ if( openBrace == '<' )
+ {
+ if( isHexDigit(ch) )
+ {
+ retval.append( ch );
+ }
+ }
+ else
+ {
+ retval.append( ch );
+ }
+ }
+ if (nextc != -2)
+ {
+ c = nextc;
+ }
+ else
+ {
+ c = pdfSource.read();
+ }
+ }
+ if (c != -1)
+ {
+ pdfSource.unread(c);
+ }
+ if( openBrace == '<' )
+ {
+ retval = COSString.createFromHexString( retval.getString() );
+ }
+ if( log.isDebugEnabled() )
+ {
+ log.debug("parseCOSString() done parsed=" + retval );
+ }
+ return retval;
+ }
+
+ /**
+ * This will parse a PDF array object.
+ *
+ * @return The parsed PDF array.
+ *
+ * @throws IOException If there is an error parsing the stream.
+ */
+ protected COSArray parseCOSArray() throws IOException
+ {
+ if( log.isDebugEnabled() )
+ {
+ log.debug("parseCOSArray() " + pdfSource );
+ }
+ char ch = (char)pdfSource.read();
+ if( ch != '[')
+ {
+ throw new IOException( "expected='[' actual='" + ch + "'" );
+ }
+ COSArray po = new COSArray();
+ COSBase pbo = null;
+ skipSpaces();
+ int i = 0;
+ while( ((i = pdfSource.peek()) > 0) && ((char)i != ']') )
+ {
+ pbo = parseDirObject();
+ if( pbo instanceof COSObject )
+ {
+ COSInteger genNumber = (COSInteger)po.remove( po.size() -1 );
+ COSInteger number = (COSInteger)po.remove( po.size() -1 );
+ COSObjectKey key = new COSObjectKey(number.intValue(), genNumber.intValue());
+ pbo = document.getObjectFromPool(key);
+ }
+ if( pbo != null )
+ {
+ po.add( pbo );
+ }
+ else
+ {
+ //it could be a bad object in the array which is just skipped
+ }
+ skipSpaces();
+ }
+ pdfSource.read(); //read ']'
+ skipSpaces();
+ if( log.isDebugEnabled() )
+ {
+ log.debug("parseCOSArray() done peek='" + (char)pdfSource.peek() + "'" );
+ }
+ return po;
+ }
+
+ /**
+ * Determine if a character terminates a PDF name.
+ *
+ * @param ch The character
+ * @return <code>true</code> if the character terminates a PDF name, otherwise <code>false</code>.
+ */
+ protected boolean isEndOfName(char ch)
+ {
+ return (ch == ' ' || ch == 13 || ch == 10 || ch == 9 || ch == '>' || ch == '<'
+ || ch == '[' || ch =='/' || ch ==']' || ch ==')' || ch =='(' ||
+ ch == -1 //EOF
+ );
+ }
+
+ /**
+ * This will parse a PDF name from the stream.
+ *
+ * @return The parsed PDF name.
+ *
+ * @throws IOException If there is an error reading from the stream.
+ */
+ protected COSName parseCOSName() throws IOException
+ {
+ if( log.isDebugEnabled() )
+ {
+ log.debug("parseCOSName() " + pdfSource );
+ }
+ COSName retval = null;
+ int c = pdfSource.read();
+ if( (char)c != '/')
+ {
+ throw new IOException("expected='/' actual='" + (char)c + "'-" + c + " " + pdfSource );
+ }
+ // costruisce il nome
+ StringBuffer buffer = new StringBuffer();
+ c = pdfSource.read();
+ while( c != -1 )
+ {
+ char ch = (char)c;
+ if(ch == '#')
+ {
+ char ch1 = (char)pdfSource.read();
+ char ch2 = (char)pdfSource.read();
+
+ // Prior to PDF v1.2, the # was not a special character. Also,
+ // it has been observed that various PDF tools do not follow the
+ // spec with respect to the # escape, even though they report
+ // PDF versions of 1.2 or later. The solution here is that we
+ // interpret the # as an escape only when it is followed by two
+ // valid hex digits.
+ //
+ if (isHexDigit(ch1) && isHexDigit(ch2))
+ {
+ String hex = "" + ch1 + ch2;
+ try
+ {
+ buffer.append( (char) Integer.parseInt(hex, 16));
+ }
+ catch (NumberFormatException e)
+ {
+ if( log.isDebugEnabled() )
+ {
+ log.debug("isHexDigit(ch1)=" + isHexDigit(ch1) + ", isHexDigit(ch2)=" + isHexDigit(ch2));
+ }
+ throw new IOException("Error: expected hex number, actual='" + hex + "'");
+ }
+ c = pdfSource.read();
+ }
+ else
+ {
+ pdfSource.unread(ch2);
+ c = ch1;
+ buffer.append( ch );
+ }
+ }
+ else if (isEndOfName(ch))
+ {
+ break;
+ }
+ else
+ {
+ buffer.append( ch );
+ c = pdfSource.read();
+ }
+ }
+ if (c != -1)
+ {
+ pdfSource.unread(c);
+ }
+ retval = COSName.getPDFName( buffer.toString() );
+ return retval;
+ }
+
+ /**
+ * This will parse a boolean object from the stream.
+ *
+ * @return The parsed boolean object.
+ *
+ * @throws IOException If an IO error occurs during parsing.
+ */
+ protected COSBoolean parseBoolean() throws IOException
+ {
+ COSBoolean retval = null;
+ char c = (char)pdfSource.peek();
+ if( c == 't' )
+ {
+ byte[] trueArray = new byte[ 4 ];
+ int amountRead = pdfSource.read( trueArray, 0, 4 );
+ String trueString = new String( trueArray, 0, amountRead );
+ if( !trueString.equals( "true" ) )
+ {
+ throw new IOException( "Error parsing boolean: expected='true' actual='" + trueString + "'" );
+ }
+ else
+ {
+ retval = COSBoolean.TRUE;
+ }
+ }
+ else if( c == 'f' )
+ {
+ byte[] falseArray = new byte[ 5 ];
+ int amountRead = pdfSource.read( falseArray, 0, 5 );
+ String falseString = new String( falseArray, 0, amountRead );
+ if( !falseString.equals( "false" ) )
+ {
+ throw new IOException( "Error parsing boolean: expected='true' actual='" + falseString + "'" );
+ }
+ else
+ {
+ retval = COSBoolean.FALSE;
+ }
+ }
+ else
+ {
+ throw new IOException( "Error parsing boolean expected='t or f' actual='" + c + "'" );
+ }
+ return retval;
+ }
+
+ /**
+ * This will parse a directory object from the stream.
+ *
+ * @return The parsed object.
+ *
+ * @throws IOException If there is an error during parsing.
+ */
+ protected COSBase parseDirObject() throws IOException
+ {
+ if( log.isDebugEnabled() )
+ {
+ log.debug("parseDirObject() " + pdfSource );
+ }
+ COSBase retval = null;
+
+ skipSpaces();
+ int nextByte = pdfSource.peek();
+ char c = (char)nextByte;
+ switch(c)
+ {
+ case '<':
+ {
+ int leftBracket = pdfSource.read();//pull off first left bracket
+ c = (char)pdfSource.peek(); //check for second left bracket
+ pdfSource.unread( leftBracket );
+ if(c == '<')
+ {
+
+ retval = parseCOSDictionary();
+ skipSpaces();
+ }
+ else
+ {
+ retval = parseCOSString();
+ }
+ break;
+ }
+ case '[': // array
+ {
+ retval = parseCOSArray();
+ break;
+ }
+ case '(':
+ retval = parseCOSString();
+ break;
+ case '/': // name
+ retval = parseCOSName();
+ break;
+ case 'n': // null
+ {
+ String nullString = readString();
+ if( !nullString.equals( "null") )
+ {
+ throw new IOException("Expected='null' actual='" + nullString + "'");
+ }
+ retval = COSNull.NULL;
+ break;
+ }
+ case 't':
+ {
+ byte[] trueBytes = new byte[4];
+ int amountRead = pdfSource.read( trueBytes, 0, 4 );
+ String trueString = new String( trueBytes, 0, amountRead );
+ if( trueString.equals( "true" ) )
+ {
+ retval = COSBoolean.TRUE;
+ }
+ else
+ {
+ throw new IOException( "expected true actual='" + trueString + "' " + pdfSource );
+ }
+ break;
+ }
+ case 'f':
+ {
+ byte[] falseBytes = new byte[5];
+ int amountRead = pdfSource.read( falseBytes, 0, 5 );
+ String falseString = new String( falseBytes, 0, amountRead );
+ if( falseString.equals( "false" ) )
+ {
+ retval = COSBoolean.FALSE;
+ }
+ else
+ {
+ throw new IOException( "expected false actual='" + falseString + "' " + pdfSource );
+ }
+ break;
+ }
+ case 'R':
+ pdfSource.read();
+ retval = new COSObject(null);
+ break;
+ case (char)-1:
+ return null;
+ default:
+ {
+ if( Character.isDigit(c) || c == '-' || c == '+' || c == '.')
+ {
+ StringBuffer buf = new StringBuffer();
+ int ic = pdfSource.read();
+ c = (char)ic;
+ while( Character.isDigit( c )||
+ c == '-' ||
+ c == '+' ||
+ c == '.' ||
+ c == 'E' ||
+ c == 'e' )
+ {
+ buf.append( c );
+ ic = pdfSource.read();
+ c = (char)ic;
+ }
+ if( ic != -1 )
+ {
+ pdfSource.unread( ic );
+ }
+ retval = COSNumber.get( buf.toString() );
+ }
+ else
+ {
+ //This is not suppose to happen, but we will allow for it
+ //so we are more compatible with POS writers that don't
+ //follow the spec
+ String badString = readString();
+ //throw new IOException( "Unknown dir object c='" + c +
+ //"' peek='" + (char)pdfSource.peek() + "' " + pdfSource );
+ if( log.isDebugEnabled() )
+ {
+ log.debug("parseDirObject() bad DIR object found. ignoring: '" + badString + "'");
+ }
+ if( badString == null || badString.length() == 0 )
+ {
+ int peek = pdfSource.peek();
+ // we can end up in an infinite loop otherwise
+ throw new IOException( "Unknown dir object c='" + c +
+ "' cInt=" + (int)c + " peek='" + (char)peek + "' peekInt=" + peek + " " + pdfSource );
+ }
+
+ }
+ }
+ }
+ if( log.isDebugEnabled() )
+ {
+ log.debug("parseDirObject() done retval=" +retval );
+ }
+ return retval;
+ }
+
+ /**
+ * This will read the next string from the stream.
+ *
+ * @return The string that was read from the stream.
+ *
+ * @throws IOException If there is an error reading from the stream.
+ */
+ protected String readString() throws IOException
+ {
+ skipSpaces();
+ StringBuffer buffer = new StringBuffer();
+ int c = pdfSource.read();
+ while( !isEndOfName((char)c) && !isClosing(c) && c != -1 )
+ {
+ buffer.append( (char)c );
+ c = pdfSource.read();
+ }
+ if (c != -1)
+ {
+ pdfSource.unread(c);
+ }
+ return buffer.toString();
+ }
+
+ /**
+ * This will read bytes until the end of line marker occurs.
+ *
+ * @param theString The next expected string in the stream.
+ *
+ * @return The characters between the current position and the end of the line.
+ *
+ * @throws IOException If there is an error reading from the stream or theString does not match what was read.
+ */
+ protected String readExpectedString( String theString ) throws IOException
+ {
+ int c = pdfSource.read();
+ while( isWhitespace(c) && c != -1)
+ {
+ c = pdfSource.read();
+ }
+ StringBuffer buffer = new StringBuffer( theString.length() );
+ int charsRead = 0;
+ while( !isEOL(c) && c != -1 && charsRead < theString.length() )
+ {
+ char next = (char)c;
+ buffer.append( next );
+ if( theString.charAt( charsRead ) == next )
+ {
+ charsRead++;
+ }
+ else
+ {
+ throw new IOException( "Error: Expected to read '" + theString +
+ "' instead started reading '" +buffer.toString() + "'" );
+ }
+ c = pdfSource.read();
+ }
+ while( isEOL(c) && c != -1 )
+ {
+ c = pdfSource.read();
+ }
+ if (c != -1)
+ {
+ pdfSource.unread(c);
+ }
+ return buffer.toString();
+ }
+
+ /**
+ * This will read the next string from the stream up to a certain length.
+ *
+ * @param length The length to stop reading at.
+ *
+ * @return The string that was read from the stream of length 0 to length.
+ *
+ * @throws IOException If there is an error reading from the stream.
+ */
+ protected String readString( int length ) throws IOException
+ {
+ skipSpaces();
+
+ int c = pdfSource.read();
+
+ //average string size is around 2 and the normal string buffer size is
+ //about 16 so lets save some space.
+ StringBuffer buffer = new StringBuffer(length);
+ while( !isWhitespace(c) && !isClosing(c) && c != -1 && buffer.length() < length &&
+ c != '[' &&
+ c != '<' &&
+ c != '(' &&
+ c != '/' )
+ {
+ buffer.append( (char)c );
+ c = pdfSource.read();
+ }
+ if (c != -1)
+ {
+ pdfSource.unread(c);
+ }
+ return buffer.toString();
+ }
+
+ /**
+ * This will tell if the next character is a closing brace( close of PDF array ).
+ *
+ * @return true if the next byte is ']', false otherwise.
+ *
+ * @throws IOException If an IO error occurs.
+ */
+ protected boolean isClosing() throws IOException
+ {
+ return isClosing(pdfSource.peek());
+ }
+
+ /**
+ * This will tell if the next character is a closing brace( close of PDF array ).
+ *
+ * @param c The character to check against end of line
+ * @return true if the next byte is ']', false otherwise.
+ */
+ protected boolean isClosing(int c)
+ {
+ return c == ']';
+ }
+
+ /**
+ * This will read bytes until the end of line marker occurs.
+ *
+ * @return The characters between the current position and the end of the line.
+ *
+ * @throws IOException If there is an error reading from the stream.
+ */
+ protected String readLine() throws IOException
+ {
+ int c = pdfSource.read();
+ while(isWhitespace(c) && c != -1)
+ {
+ c = pdfSource.read();
+ }
+ StringBuffer buffer = new StringBuffer( 11 );
+
+ while( !isEOL(c) && c != -1 )
+ {
+ buffer.append( (char)c );
+ c = pdfSource.read();
+ }
+ while( isEOL(c) && c != -1 )
+ {
+ c = pdfSource.read();
+ }
+ if (c != -1)
+ {
+ pdfSource.unread(c);
+ }
+ return buffer.toString();
+ }
+
+ /**
+ * This will tell if the next byte to be read is an end of line byte.
+ *
+ * @return true if the next byte is 0x0A or 0x0D.
+ *
+ * @throws IOException If there is an error reading from the stream.
+ */
+ protected boolean isEOL() throws IOException
+ {
+ return isEOL(pdfSource.peek());
+ }
+
+ /**
+ * This will tell if the next byte to be read is an end of line byte.
+ *
+ * @param c The character to check against end of line
+ * @return true if the next byte is 0x0A or 0x0D.
+ */
+ protected boolean isEOL(int c)
+ {
+ return c == 10 || c == 13;
+ }
+
+ /**
+ * This will tell if the next byte is whitespace or not.
+ *
+ * @return true if the next byte in the stream is a whitespace character.
+ *
+ * @throws IOException If there is an error reading from the stream.
+ */
+ protected boolean isWhitespace() throws IOException
+ {
+ return isWhitespace( pdfSource.peek() );
+ }
+
+ /**
+ * This will tell if the next byte is whitespace or not.
+ *
+ * @param c The character to check against whitespace
+ *
+ * @return true if the next byte in the stream is a whitespace character.
+ */
+ protected boolean isWhitespace( int c )
+ {
+ return c == 0 || c == 9 || c == 12 || c == 10
+ || c == 13 || c == 32;
+ }
+
+ /**
+ * This will skip all spaces and comments that are present.
+ *
+ * @throws IOException If there is an error reading from the stream.
+ */
+ protected void skipSpaces() throws IOException
+ {
+ //log( "skipSpaces() " + pdfSource );
+ int c = pdfSource.read();
+ // identical to, but faster as: isWhiteSpace(c) || c == 37
+ while(c == 0 || c == 9 || c == 12 || c == 10
+ || c == 13 || c == 32 || c == 37)//37 is the % character, a comment
+ {
+ if ( c == 37 )
+ {
+ // skip past the comment section
+ c = pdfSource.read();
+ while(!isEOL(c) && c != -1)
+ {
+ c = pdfSource.read();
+ }
+ }
+ else
+ {
+ c = pdfSource.read();
+ }
+ }
+ if (c != -1)
+ {
+ pdfSource.unread(c);
+ }
+ //log( "skipSpaces() done peek='" + (char)pdfSource.peek() + "'" );
+ }
+
+ /**
+ * this will compare two byte arrays.
+ *
+ * @param first The first byte array to compare.
+ * @param second The second byte array to compare.
+ *
+ * @return true if both arrays are the same AND forall i : first[i] = second[i]
+ */
+ private boolean cmpArray( byte[] first, byte[] second )
+ {
+ return cmpArray( first, 0, second );
+ }
+
+ /**
+ * This will compare two arrays for equality.
+ *
+ * @param first The first array to compare.
+ * @param firstOffset The first byte to start comparing.
+ * @param second The second array to compare.
+ */
+ private boolean cmpArray( byte[] first, int firstOffset, byte[] second )
+ {
+ boolean retval = true;
+ if( first.length-firstOffset >= second.length )
+ {
+ int arrayLength = second.length;
+ for( int i =0; i<arrayLength && retval; i++ )
+ {
+ retval = retval && first[ firstOffset + i ] == second[ i ];
+ }
+ }
+ else
+ {
+ retval = false;
+ }
+ return retval;
+ }
+
+ /**
+ * This will read an integer from the stream.
+ *
+ * @return The integer that was read from the stream.
+ *
+ * @throws IOException If there is an error reading from the stream.
+ */
+ protected int readInt() throws IOException
+ {
+ skipSpaces();
+ int retval = 0;
+
+ int lastByte = 0;
+ StringBuffer intBuffer = new StringBuffer();
+ while( (lastByte = pdfSource.read() ) != 32 &&
+ lastByte != 10 &&
+ lastByte != 13 &&
+ lastByte != 0 && //See sourceforge bug 853328
+ lastByte != -1 )
+ {
+ intBuffer.append( (char)lastByte );
+ }
+ try
+ {
+ retval = Integer.parseInt( intBuffer.toString() );
+ }
+ catch( NumberFormatException e )
+ {
+ throw new IOException( "Error: Expected an integer type, actual='" + intBuffer + "'" );
+ }
+ return retval;
+ }
+
+ /**
+ * This will add an xref.
+ *
+ * @param xref The xref to add.
+ */
+ public void addXref( PDFXref xref )
+ {
+ xrefs.add(xref);
+ }
+
+ /**
+ * This will get all of the xrefs.
+ *
+ * @return A list of all xrefs.
+ */
+ public List getXrefs()
+ {
+ return xrefs;
+ }
+
+ /**
+ * This will set the xrefs for this parser.
+ *
+ * @param newXrefs The xrefs for this parser.
+ */
+ private void setXrefs( List newXrefs )
+ {
+ xrefs = newXrefs;
+ }
+}
\ No newline at end of file diff --git a/src/main/java/org/pdfbox/pdfparser/PDFObjectStreamParser.java b/src/main/java/org/pdfbox/pdfparser/PDFObjectStreamParser.java new file mode 100644 index 0000000..6fb7563 --- /dev/null +++ b/src/main/java/org/pdfbox/pdfparser/PDFObjectStreamParser.java @@ -0,0 +1,137 @@ +/**
+ * Copyright (c) 2003-2004, www.pdfbox.org
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright notice,
+ * this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright notice,
+ * this list of conditions and the following disclaimer in the documentation
+ * and/or other materials provided with the distribution.
+ * 3. Neither the name of pdfbox; nor the names of its
+ * contributors may be used to endorse or promote products derived from this
+ * software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+ * DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE FOR ANY
+ * DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+ * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON
+ * ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+ * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ * http://www.pdfbox.org
+ *
+ */
+package org.pdfbox.pdfparser;
+
+import java.io.IOException;
+
+import java.util.ArrayList;
+import java.util.List;
+
+import org.pdfbox.cos.COSBase;
+import org.pdfbox.cos.COSDocument;
+import org.pdfbox.cos.COSInteger;
+import org.pdfbox.cos.COSObject;
+import org.pdfbox.cos.COSStream;
+
+
+import org.apache.log4j.Logger;
+
+/**
+ * This will parse a PDF 1.5 object stream and extract all of the objects from the stream.
+ *
+ * @author Ben Litchfield (ben@csh.rit.edu)
+ * @version $Revision: 1.4 $
+ */
+public class PDFObjectStreamParser extends BaseParser
+{
+ private static Logger log = Logger.getLogger(PDFObjectStreamParser.class);
+ private List streamObjects = null;
+ private List objectNumbers = null;
+ private COSStream stream;
+
+ /**
+ * Constructor.
+ *
+ * @param strm The stream to parse.
+ * @param doc The document for the current parsing.
+ *
+ * @throws IOException If there is an error initializing the stream.
+ */
+ public PDFObjectStreamParser( COSStream strm, COSDocument doc ) throws IOException
+ {
+ super( strm.getUnfilteredStream() );
+ setDocument( doc );
+ stream = strm;
+ }
+
+ /**
+ * This will parse the tokens in the stream. This will close the
+ * stream when it is finished parsing.
+ *
+ * @throws IOException If there is an error while parsing the stream.
+ */
+ public void parse() throws IOException
+ {
+ if( log.isDebugEnabled() )
+ {
+ log.debug( "parse() start" );
+ }
+
+ try
+ {
+ //need to first parse the header.
+ int numberOfObjects = stream.getInt( "N" );
+ objectNumbers = new ArrayList( numberOfObjects );
+ streamObjects = new ArrayList( numberOfObjects );
+ for( int i=0; i<numberOfObjects; i++ )
+ {
+ int objectNumber = readInt();
+ int offset = readInt();
+ if( log.isDebugEnabled() )
+ {
+ log.debug( "objNum:" + objectNumber + " offset:" + offset );
+ }
+ objectNumbers.add( new Integer( objectNumber ) );
+ }
+ COSObject object = null;
+ COSBase cosObject = null;
+ int objectCounter = 0;
+ while( (cosObject = parseDirObject()) != null )
+ {
+ object = new COSObject(cosObject);
+ object.setGenerationNumber( COSInteger.ZERO );
+ COSInteger objNum =
+ new COSInteger( ((Integer)objectNumbers.get( objectCounter)).intValue() );
+ object.setObjectNumber( objNum );
+ streamObjects.add( object );
+ objectCounter++;
+ }
+ }
+ finally
+ {
+ pdfSource.close();
+ }
+ if( log.isDebugEnabled() )
+ {
+ log.debug( "parse() end" );
+ }
+ }
+
+ /**
+ * This will get the objects that were parsed from the stream.
+ *
+ * @return All of the objects in the stream.
+ */
+ public List getObjects()
+ {
+ return streamObjects;
+ }
+}
\ No newline at end of file diff --git a/src/main/java/org/pdfbox/pdfparser/PDFParser.java b/src/main/java/org/pdfbox/pdfparser/PDFParser.java new file mode 100644 index 0000000..d655ef1 --- /dev/null +++ b/src/main/java/org/pdfbox/pdfparser/PDFParser.java @@ -0,0 +1,557 @@ +/**
+ * Copyright (c) 2003-2005, www.pdfbox.org
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright notice,
+ * this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright notice,
+ * this list of conditions and the following disclaimer in the documentation
+ * and/or other materials provided with the distribution.
+ * 3. Neither the name of pdfbox; nor the names of its
+ * contributors may be used to endorse or promote products derived from this
+ * software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+ * DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE FOR ANY
+ * DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+ * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON
+ * ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+ * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ * http://www.pdfbox.org
+ *
+ */
+package org.pdfbox.pdfparser;
+
+import java.io.File;
+import java.io.RandomAccessFile;
+import java.io.InputStream;
+import java.io.IOException;
+
+import java.util.Iterator;
+
+import org.pdfbox.cos.COSBase;
+import org.pdfbox.cos.COSDictionary;
+import org.pdfbox.cos.COSDocument;
+import org.pdfbox.cos.COSObject;
+import org.pdfbox.cos.COSStream;
+import org.pdfbox.exceptions.WrappedIOException;
+
+import org.pdfbox.pdmodel.PDDocument;
+
+import org.pdfbox.pdmodel.fdf.FDFDocument;
+
+import org.pdfbox.persistence.util.COSObjectKey;
+
+import org.apache.log4j.Logger;
+
+/**
+ * This class will handle the parsing of the PDF document.
+ *
+ * @author Ben Litchfield (ben@benlitchfield.com)
+ * @version $Revision: 1.47 $
+ */
+public class PDFParser extends BaseParser
+{
+ private static Logger log = Logger.getLogger( PDFParser.class );
+ private static final int SPACE_BYTE = 32;
+
+ private static final String PDF_HEADER = "%PDF-";
+ private COSDocument document;
+
+ /**
+ * Temp file directory.
+ */
+ private File tempDirectory = new File( System.getProperty( "java.io.tmpdir" ) );
+
+ private RandomAccessFile raf = null;
+
+ /**
+ * Constructor.
+ *
+ * @param input The input stream that contains the PDF document.
+ *
+ * @throws IOException If there is an error initializing the stream.
+ */
+ public PDFParser( InputStream input ) throws IOException
+ {
+ this(input, null);
+ }
+
+ /**
+ * Constructor to allow control over RandomAccessFile.
+ * @param input The input stream that contains the PDF document.
+ * @param rafi The RandomAccessFile to be used in internal COSDocument
+ *
+ * @throws IOException If there is an error initializing the stream.
+ */
+ public PDFParser(InputStream input, RandomAccessFile rafi)
+ throws IOException
+ {
+ super(input);
+ this.raf = rafi;
+ }
+
+ /**
+ * This is the directory where pdfbox will create a temporary file
+ * for storing pdf document stream in. By default this directory will
+ * be the value of the system property java.io.tmpdir.
+ *
+ * @param tmpDir The directory to create scratch files needed to store
+ * pdf document streams.
+ */
+ public void setTempDirectory( File tmpDir )
+ {
+ tempDirectory = tmpDir;
+ }
+
+ /**
+ * This will prase the stream and create the PDF document. This will close
+ * the stream when it is done parsing.
+ *
+ * @throws IOException If there is an error reading from the stream.
+ */
+ public void parse() throws IOException
+ {
+ try
+ {
+ if ( raf == null )
+ {
+ document = new COSDocument( tempDirectory );
+ }
+ else
+ {
+ document = new COSDocument( raf );
+ }
+ setDocument( document );
+ String header = readLine();
+ if( log.isDebugEnabled() )
+ {
+ log.debug( "Header=" + header );
+ }
+ document.setHeaderString( header );
+
+ if( header.length() < PDF_HEADER.length()+1 )
+ {
+ throw new IOException( "Error: Header is corrupt '" + header + "'" );
+ }
+
+ //sometimes there are some garbage bytes in the header before the header
+ //actually starts, so lets try to find the header first.
+ int headerStart = header.indexOf( PDF_HEADER );
+
+ //greater than zero because if it is zero then
+ //there is no point of trimming
+ if( headerStart > 0 )
+ {
+ //trim off any leading characters
+ header = header.substring( headerStart, header.length() );
+ }
+
+ try
+ {
+ float pdfVersion = Float.parseFloat(
+ header.substring( PDF_HEADER.length(), Math.min( header.length(), PDF_HEADER.length()+3) ) );
+ document.setVersion( pdfVersion );
+ }
+ catch( NumberFormatException e )
+ {
+ throw new IOException( "Error getting pdf version:" + e );
+ }
+
+ skipHeaderFillBytes();
+
+
+ Object nextObject;
+ boolean wasLastParsedObjectAnXref = false;
+ try
+ {
+ while( (nextObject = parseObject()) != null )
+ {
+ if( nextObject instanceof PDFXref )
+ {
+ PDFXref xref = (PDFXref)nextObject;
+ addXref(xref);
+ wasLastParsedObjectAnXref = true;
+ }
+ else
+ {
+ wasLastParsedObjectAnXref = false;
+ }
+ skipSpaces();
+ }
+ if( document.getTrailer() == null )
+ {
+ COSDictionary trailer = new COSDictionary();
+ Iterator xrefIter = document.getObjectsByType( "XRef" ).iterator();
+ while( xrefIter.hasNext() )
+ {
+ COSStream next = (COSStream)((COSObject)xrefIter.next()).getObject();
+ trailer.addAll( next );
+ }
+ document.setTrailer( trailer );
+ }
+ if( !document.isEncrypted() )
+ {
+ document.dereferenceObjectStreams();
+ }
+ }
+ catch( IOException e )
+ {
+ if( wasLastParsedObjectAnXref )
+ {
+ log.debug( "Skipping some garbage", e );
+ //Then we assume that there is just random garbage after
+ //the xref, not sure why the PDF spec allows this but it does.
+ }
+ else
+ {
+ //some other error so just pass it along
+ throw e;
+ }
+ }
+ }
+ catch( Throwable t )
+ {
+ //so if the PDF is corrupt then close the document and clear
+ //all resources to it
+ if( document != null )
+ {
+ document.close();
+ }
+ if( t instanceof IOException )
+ {
+ throw (IOException)t;
+ }
+ else
+ {
+ throw new WrappedIOException( t );
+ }
+ }
+ finally
+ {
+ pdfSource.close();
+ }
+ }
+
+ /**
+ * This will skip a header's binary fill bytes. This is in accordance to
+ * PDF Specification 1.5 pg 68 section 3.4.1 "Syntax.File Structure.File Header"
+ *
+ * @throws IOException If there is an error reading from the stream.
+ */
+ protected void skipHeaderFillBytes() throws IOException
+ {
+ skipSpaces();
+ int c = pdfSource.peek();
+
+ if( !Character.isDigit( (char)c ) )
+ {
+ // Fill bytes conform with PDF reference (but without comment sign)
+ // => skip until EOL
+ readLine();
+ }
+ // else: no fill bytes
+ }
+
+ /**
+ * This will get the document that was parsed. parse() must be called before this is called.
+ * When you are done with this document you must call close() on it to release
+ * resources.
+ *
+ * @return The document that was parsed.
+ *
+ * @throws IOException If there is an error getting the document.
+ */
+ public COSDocument getDocument() throws IOException
+ {
+ if( document == null )
+ {
+ throw new IOException( "You must call parse() before calling getDocument()" );
+ }
+ return document;
+ }
+
+ /**
+ * This will get the PD document that was parsed. When you are done with
+ * this document you must call close() on it to release resources.
+ *
+ * @return The document at the PD layer.
+ *
+ * @throws IOException If there is an error getting the document.
+ */
+ public PDDocument getPDDocument() throws IOException
+ {
+ return new PDDocument( getDocument() );
+ }
+
+ /**
+ * This will get the FDF document that was parsed. When you are done with
+ * this document you must call close() on it to release resources.
+ *
+ * @return The document at the PD layer.
+ *
+ * @throws IOException If there is an error getting the document.
+ */
+ public FDFDocument getFDFDocument() throws IOException
+ {
+ return new FDFDocument( getDocument() );
+ }
+
+ /**
+ * This will parse a document object from the stream.
+ *
+ * @return The parsed object.
+ *
+ * @throws IOException If an IO error occurs.
+ */
+ private Object parseObject() throws IOException
+ {
+ Object object = null;
+ char peekedChar = (char)pdfSource.peek();
+ if( log.isDebugEnabled() )
+ {
+ log.debug( "PDFParser.parseObject() peek='" + peekedChar + "'" );
+ }
+ if( pdfSource.isEOF() )
+ {
+ if( log.isDebugEnabled() )
+ {
+ log.debug( "Skipping because of EOF" );
+ //end of file we will return a null object and call it a day.
+ }
+ }
+ else if( peekedChar == 'x' ||
+ peekedChar == 't' ||
+ peekedChar == 's')
+ {
+ //System.out.println( "parseObject() parsing xref" );
+
+ //FDF documents do not always have the xref
+ if( peekedChar == 'x' || peekedChar == 't' )
+ {
+ object = parseXrefSection();
+ }
+
+ //if peeked char is xref or startxref
+ if( peekedChar == 'x' || peekedChar == 's')
+ {
+ skipSpaces();
+ while( pdfSource.peek() == 'x' )
+ {
+ parseXrefSection();
+ }
+ String startxref = readString();
+ if( !startxref.equals( "startxref" ) )
+ {
+ throw new IOException( "expected='startxref' actual='" + startxref + "' " + pdfSource );
+ }
+ skipSpaces();
+ //read some integer that is in the stream but PDFBox doesn't use
+ readInt();
+ }
+
+ //This MUST be readLine because readString strips out comments
+ //and it will think that %% is a comment in from of the EOF
+ String eof = readExpectedString( "%%EOF" );
+ if( eof.indexOf( "%%EOF" )== -1 && !pdfSource.isEOF() )
+ {
+ throw new IOException( "expected='%%EOF' actual='" + eof + "' next=" + readString() +
+ " next=" +readString() );
+ }
+ else if( !pdfSource.isEOF() )
+ {
+ //we might really be at the end of the file, there might just be some crap at the
+ //end of the file.
+ if( pdfSource.available() < 1000 )
+ {
+ //We need to determine if we are at the end of the file.
+ byte[] data = new byte[ 1000 ];
+
+ int amountRead = pdfSource.read( data );
+ if( amountRead != -1 )
+ {
+ pdfSource.unread( data, 0, amountRead );
+ }
+ boolean atEndOfFile = true;//we assume yes unless we find another.
+ for( int i=0; i<amountRead-3 && atEndOfFile; i++ )
+ {
+ atEndOfFile = !(data[i] == 'E' &&
+ data[i+1] == 'O' &&
+ data[i+2] == 'F' );
+ }
+ if( atEndOfFile )
+ {
+ while( pdfSource.read( data, 0, data.length ) != -1 )
+ {
+ //read until done.
+ }
+ }
+ }
+ }
+ }
+ else
+ {
+ int number;
+ int genNum;
+ String objectKey = null;
+ try
+ {
+ number = readInt();
+ }
+ catch( IOException e )
+ {
+ //ok for some reason "GNU Ghostscript 5.10" puts two endobj
+ //statements after an object, of course this is nonsense
+ //but because we want to support as many PDFs as possible
+ //we will simply try again
+ number = readInt();
+ }
+ skipSpaces();
+ genNum = readInt();
+ if( log.isDebugEnabled() )
+ {
+ log.debug( "Parsing object (" + number + "," + genNum + ")" );
+ }
+
+ objectKey = readString( 3 );
+ //System.out.println( "parseObject() num=" + number + " genNumber=" + genNum + " key='" + objectKey + "'" );
+ if( !objectKey.equals( "obj" ) )
+ {
+ throw new IOException("expected='obj' actual='" + objectKey + "' " + pdfSource );
+ }
+
+ skipSpaces();
+ COSBase pb = parseDirObject();
+ String endObjectKey = readString();
+ if( endObjectKey.equals( "stream" ) )
+ {
+ pdfSource.unread( endObjectKey.getBytes() );
+ pdfSource.unread( ' ' );
+ if( pb instanceof COSDictionary )
+ {
+ pb = parseCOSStream( (COSDictionary)pb, getDocument().getScratchFile() );
+ }
+ else
+ {
+ // this is not legal
+ // the combination of a dict and the stream/endstream forms a complete stream object
+ throw new IOException("stream not preceded by dictionary");
+ }
+ endObjectKey = readString();
+ }
+ COSObjectKey key = new COSObjectKey( number, genNum );
+ COSObject pdfObject = document.getObjectFromPool( key );
+ object = pdfObject;
+ pdfObject.setObject(pb);
+
+ if( !endObjectKey.equals( "endobj" ) )
+ {
+ if( !pdfSource.isEOF() )
+ {
+ try
+ {
+ //It is possible that the endobj is missing, there
+ //are several PDFs out there that do that so skip it and move on.
+ Float.parseFloat( endObjectKey );
+ pdfSource.unread( SPACE_BYTE );
+ pdfSource.unread( endObjectKey.getBytes() );
+ if( log.isDebugEnabled() )
+ {
+ log.debug( "Missing endobj, found '" + endObjectKey +
+ "' instead, assuming that endobj is not present and will continue parsing." );
+ }
+ }
+ catch( NumberFormatException e )
+ {
+ //we will try again incase there was some garbage which
+ //some writers will leave behind.
+ String secondEndObjectKey = readString();
+ if( !secondEndObjectKey.equals( "endobj" ) )
+ {
+ throw new IOException("expected='endobj' firstReadAttempt='" + endObjectKey + "' " +
+ "secondReadAttempt='" + secondEndObjectKey + "' " + pdfSource);
+ }
+ }
+ }
+ }
+ skipSpaces();
+
+ }
+ //System.out.println( "parsed=" + object );
+ return object;
+ }
+
+
+ /**
+ * This will parse the xref table and trailers from the stream.
+ *
+ * @return a new PDFXref
+ *
+ * @throws IOException If an IO error occurs.
+ */
+ protected PDFXref parseXrefSection() throws IOException
+ {
+ int[] params = new int[2];
+ parseXrefTable(params);
+ parseTrailer();
+
+ return new PDFXref(params[0], params[1]);
+ }
+
+ /**
+ * This will parse the xref table from the stream.
+ *
+ * It stores the starting object number and the count
+ *
+ * @param params The start and count parameters
+ *
+ * @throws IOException If an IO error occurs.
+ */
+ protected void parseXrefTable(int[] params) throws IOException
+ {
+ String nextLine = null;
+
+ nextLine = readLine();
+ if( nextLine.equals( "xref" ) )
+ {
+ params[0] = readInt();
+ params[1] = readInt();
+ nextLine = readString();
+ }
+ skipSpaces();
+ while( !nextLine.equals( "trailer" ) && !pdfSource.isEOF() && !isEndOfName((char)pdfSource.peek()))
+ {
+ //skip past all the xref entries.
+ nextLine = readString();
+ skipSpaces();
+ }
+ skipSpaces();
+ }
+
+ private void parseTrailer() throws IOException
+ {
+ COSDictionary parsedTrailer = parseCOSDictionary();
+ COSDictionary docTrailer = document.getTrailer();
+ if( log.isDebugEnabled() )
+ {
+ log.debug( "parsedTrailer=" + parsedTrailer );
+ log.debug( "docTrailer=" + docTrailer );
+ }
+ if( docTrailer == null )
+ {
+ document.setTrailer( parsedTrailer );
+ }
+ else
+ {
+ docTrailer.addAll( parsedTrailer );
+ }
+ }
+}
diff --git a/src/main/java/org/pdfbox/pdfparser/PDFStreamParser.java b/src/main/java/org/pdfbox/pdfparser/PDFStreamParser.java new file mode 100644 index 0000000..d59c5a4 --- /dev/null +++ b/src/main/java/org/pdfbox/pdfparser/PDFStreamParser.java @@ -0,0 +1,403 @@ +/**
+ * Copyright (c) 2003-2004, www.pdfbox.org
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright notice,
+ * this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright notice,
+ * this list of conditions and the following disclaimer in the documentation
+ * and/or other materials provided with the distribution.
+ * 3. Neither the name of pdfbox; nor the names of its
+ * contributors may be used to endorse or promote products derived from this
+ * software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+ * DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE FOR ANY
+ * DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+ * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON
+ * ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+ * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ * http://www.pdfbox.org
+ *
+ */
+package org.pdfbox.pdfparser;
+
+import java.io.ByteArrayOutputStream;
+import java.io.InputStream;
+import java.io.IOException;
+import java.io.RandomAccessFile;
+
+import java.util.ArrayList;
+import java.util.List;
+
+import org.pdfbox.cos.COSBase;
+import org.pdfbox.cos.COSBoolean;
+import org.pdfbox.cos.COSDictionary;
+import org.pdfbox.cos.COSName;
+import org.pdfbox.cos.COSNull;
+import org.pdfbox.cos.COSNumber;
+import org.pdfbox.cos.COSObject;
+import org.pdfbox.cos.COSStream;
+
+import org.pdfbox.util.PDFOperator;
+import org.pdfbox.util.ImageParameters;
+
+import org.apache.log4j.Logger;
+
+/**
+ * This will parse a PDF byte stream and extract operands and such.
+ *
+ * @author Ben Litchfield (ben@csh.rit.edu)
+ * @version $Revision: 1.29 $
+ */
+public class PDFStreamParser extends BaseParser
+{
+ private static Logger log = Logger.getLogger(PDFStreamParser.class);
+ private List streamObjects = new ArrayList( 100 );
+ private RandomAccessFile file;
+ private PDFOperator lastBIToken = null;
+
+ /**
+ * Constructor that takes a stream to parse.
+ *
+ * @param stream The stream to read data from.
+ * @param raf The random access file.
+ *
+ * @throws IOException If there is an error reading from the stream.
+ */
+ public PDFStreamParser( InputStream stream, RandomAccessFile raf ) throws IOException
+ {
+ super( stream );
+ file = raf;
+ }
+
+ /**
+ * Constructor.
+ *
+ * @param stream The stream to parse.
+ *
+ * @throws IOException If there is an error initializing the stream.
+ */
+ public PDFStreamParser( COSStream stream ) throws IOException
+ {
+ this( stream.getUnfilteredStream(), stream.getScratchFile() );
+ }
+
+ /**
+ * This will parse the tokens in the stream. This will close the
+ * stream when it is finished parsing.
+ *
+ * @throws IOException If there is an error while parsing the stream.
+ */
+ public void parse() throws IOException
+ {
+ if( log.isDebugEnabled() )
+ {
+ log.debug( "parse() start" );
+ }
+
+ try
+ {
+ Object token = null;
+ while( (token = parseNextToken()) != null )
+ {
+ streamObjects.add( token );
+ }
+ }
+ finally
+ {
+ pdfSource.close();
+ }
+ if( log.isDebugEnabled() )
+ {
+ log.debug( "parse() end" );
+ }
+ }
+
+ /**
+ * This will get the tokens that were parsed from the stream.
+ *
+ * @return All of the tokens in the stream.
+ */
+ public List getTokens()
+ {
+ return streamObjects;
+ }
+
+ /**
+ * This will parse the next token in the stream.
+ *
+ * @return The next token in the stream or null if there are no more tokens in the stream.
+ *
+ * @throws IOException If an io error occurs while parsing the stream.
+ */
+ private Object parseNextToken() throws IOException
+ {
+ if( log.isDebugEnabled() )
+ {
+ log.debug( "parseNextToken() start" );
+ }
+ Object retval = null;
+
+ skipSpaces();
+ int nextByte = pdfSource.peek();
+ if( ((byte)nextByte) == -1 )
+ {
+ return null;
+ }
+ char c = (char)nextByte;
+ switch(c)
+ {
+ case '<':
+ {
+ int leftBracket = pdfSource.read();//pull off first left bracket
+ c = (char)pdfSource.peek(); //check for second left bracket
+ pdfSource.unread( leftBracket ); //put back first bracket
+ if(c == '<')
+ {
+
+ COSDictionary pod = parseCOSDictionary();
+ skipSpaces();
+ if((char)pdfSource.peek() == 's')
+ {
+ retval = parseCOSStream( pod, file );
+ }
+ else
+ {
+ retval = pod;
+ }
+ }
+ else
+ {
+ retval = parseCOSString();
+ }
+ break;
+ }
+ case '[': // array
+ {
+ retval = parseCOSArray();
+ break;
+ }
+ case '(': // string
+ retval = parseCOSString();
+ break;
+ case '/': // name
+ retval = parseCOSName();
+ break;
+ case 'n': // null
+ {
+ String nullString = readString();
+ if( nullString.equals( "null") )
+ {
+ retval = COSNull.NULL;
+ }
+ else
+ {
+ retval = PDFOperator.getOperator( nullString );
+ }
+ break;
+ }
+ case 't':
+ case 'f':
+ {
+ String next = readString();
+ if( next.equals( "true" ) )
+ {
+ retval = COSBoolean.TRUE;
+ break;
+ }
+ else if( next.equals( "false" ) )
+ {
+ retval = COSBoolean.FALSE;
+ }
+ else
+ {
+ retval = PDFOperator.getOperator( next );
+ }
+ break;
+ }
+ case 'R':
+ {
+ String line = readString();
+ if( line.equals( "R" ) )
+ {
+ retval = new COSObject( null );
+ }
+ else
+ {
+ retval = PDFOperator.getOperator( line );
+ }
+ break;
+ }
+ case '0':
+ case '1':
+ case '2':
+ case '3':
+ case '4':
+ case '5':
+ case '6':
+ case '7':
+ case '8':
+ case '9':
+ case '-':
+ case '+':
+ case '.':
+ {
+ if( Character.isDigit(c) || c == '-' || c == '+' || c == '.')
+ {
+ StringBuffer buf = new StringBuffer();
+ while( Character.isDigit(( c = (char)pdfSource.peek()) )|| c== '-' || c== '+' || c =='.' )
+ {
+ buf.append( c );
+ pdfSource.read();
+ }
+ retval = COSNumber.get( buf.toString() );
+ }
+ else
+ {
+ throw new IOException( "Unknown dir object c='" + c +
+ "' peek='" + (char)pdfSource.peek() + "' " + pdfSource );
+ }
+ break;
+ }
+ case 'B':
+ {
+ String next = readString();
+ retval = PDFOperator.getOperator( next );
+
+ if( next.equals( "BI" ) )
+ {
+ lastBIToken = (PDFOperator)retval;
+ COSDictionary imageParams = new COSDictionary();
+ lastBIToken.setImageParameters( new ImageParameters( imageParams ) );
+ Object nextToken = null;
+ while( (nextToken = parseNextToken()) instanceof COSName )
+ {
+ Object value = parseNextToken();
+ imageParams.setItem( (COSName)nextToken, (COSBase)value );
+ }
+ //final token will be the image data, maybe??
+ PDFOperator imageData = (PDFOperator)nextToken;
+ lastBIToken.setImageData( imageData.getImageData() );
+ }
+ break;
+ }
+ case 'I':
+ {
+ ImageParameters imageParams = lastBIToken.getImageParameters();
+ int expectedBytes = (int)Math.ceil(imageParams.getHeight() * imageParams.getWidth() *
+ (imageParams.getBitsPerComponent()/8) );
+ //Special case for ID operator
+ String id = "" + (char)pdfSource.read() + (char)pdfSource.read();
+ if( !id.equals( "ID" ) )
+ {
+ throw new IOException( "Error: Expected operator 'ID' actual='" + id + "'" );
+ }
+ ByteArrayOutputStream imageData = new ByteArrayOutputStream();
+ boolean foundEnd = false;
+ if( this.isWhitespace() )
+ {
+ //pull off the whitespace character
+ pdfSource.read();
+ }
+ int twoBytesAgo = 0;
+ int lastByte = pdfSource.read();
+ int currentByte = pdfSource.read();
+ int count = 0;
+ //PDF spec is kinda unclear about this. Should a whitespace
+ //always appear before EI? Not sure, I found a PDF
+ //(UnderstandingWebSphereClassLoaders.pdf) which has EI as part
+ //of the image data and will stop parsing prematurely if there is
+ //not a check for <whitespace>EI<whitespace>.
+ while( !(isWhitespace( twoBytesAgo ) &&
+ lastByte == 'E' &&
+ currentByte == 'I' &&
+ isWhitespace() //&&
+ //amyuni2_05d__pdf1_3_acro4x.pdf has image data that
+ //is compressed, so expectedBytes is useless here.
+ //count >= expectedBytes
+ ) &&
+ !pdfSource.isEOF() )
+ {
+ imageData.write( lastByte );
+ twoBytesAgo = lastByte;
+ lastByte = currentByte;
+ currentByte = pdfSource.read();
+ count++;
+ }
+ pdfSource.unread( 'I' ); //unread the EI operator
+ pdfSource.unread( 'E' );
+ retval = PDFOperator.getOperator( "ID" );
+ ((PDFOperator)retval).setImageData( imageData.toByteArray() );
+ break;
+ }
+ case ']':
+ {
+ // some ']' around without its previous '['
+ // this means a PDF is somewhat corrupt but we will continue to parse.
+ pdfSource.read();
+ retval = COSNull.NULL; // must be a better solution than null...
+ break;
+ }
+ default:
+ {
+ //we must be an operator
+ String operator = readOperator();
+ if( operator.trim().length() == 0 )
+ {
+ //we have a corrupt stream, stop reading here
+ retval = null;
+ }
+ else
+ {
+ retval = PDFOperator.getOperator( operator );
+ }
+ }
+
+ }
+ if( log.isDebugEnabled() )
+ {
+ log.debug( "parseNextToken() retval=" + retval + " peek=" + (char)pdfSource.peek() + " end" );
+ }
+
+ return retval;
+ }
+
+ /**
+ * This will read an operator from the stream.
+ *
+ * @return The operator that was read from the stream.
+ *
+ * @throws IOException If there is an error reading from the stream.
+ */
+ protected String readOperator() throws IOException
+ {
+ skipSpaces();
+
+ //average string size is around 2 and the normal string buffer size is
+ //about 16 so lets save some space.
+ StringBuffer buffer = new StringBuffer(4);
+ while(
+ !isWhitespace() &&
+ !isClosing() &&
+ !pdfSource.isEOF() &&
+ pdfSource.peek() != (int)'[' &&
+ pdfSource.peek() != (int)'<' &&
+ pdfSource.peek() != (int)'(' &&
+ pdfSource.peek() != (int)'/' &&
+ (pdfSource.peek() < (int)'0' ||
+ pdfSource.peek() > (int)'9' ) )
+ {
+ buffer.append( (char)pdfSource.read() );
+ }
+ return buffer.toString();
+ }
+}
\ No newline at end of file diff --git a/src/main/java/org/pdfbox/pdfparser/PDFXref.java b/src/main/java/org/pdfbox/pdfparser/PDFXref.java new file mode 100644 index 0000000..abe0f35 --- /dev/null +++ b/src/main/java/org/pdfbox/pdfparser/PDFXref.java @@ -0,0 +1,96 @@ +/**
+ * Copyright (c) 2003, www.pdfbox.org
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright notice,
+ * this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright notice,
+ * this list of conditions and the following disclaimer in the documentation
+ * and/or other materials provided with the distribution.
+ * 3. Neither the name of pdfbox; nor the names of its
+ * contributors may be used to endorse or promote products derived from this
+ * software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+ * DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE FOR ANY
+ * DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+ * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON
+ * ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+ * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ * http://www.pdfbox.org
+ *
+ */
+package org.pdfbox.pdfparser;
+
+/**
+ * This class represents a PDF xref.
+ *
+ * @author Ben Litchfield (ben@csh.rit.edu)
+ * @version $Revision: 1.3 $
+ */
+public class PDFXref
+{
+
+ private long count;
+ private long start;
+
+ /**
+ * constructor.
+ *
+ * @param startValue The start attribute.
+ * @param countValue The count attribute.
+ */
+ public PDFXref( long startValue, long countValue )
+ {
+ setStart( startValue );
+ setCount( countValue );
+ }
+
+ /**
+ * This will get the count attribute.
+ *
+ * @return The count.
+ */
+ public long getCount()
+ {
+ return count;
+ }
+
+ /**
+ * This will get the start attribute.
+ *
+ * @return The start.
+ */
+ public long getStart()
+ {
+ return start;
+ }
+
+ /**
+ * This will set the count attribute.
+ *
+ * @param newCount The new count.
+ */
+ private void setCount(long newCount)
+ {
+ count = newCount;
+ }
+
+ /**
+ * This will set the start attribute.
+ *
+ * @param newStart The new start attribute.
+ */
+ private void setStart(long newStart)
+ {
+ start = newStart;
+ }
+}
\ No newline at end of file diff --git a/src/main/java/org/pdfbox/pdfparser/package.html b/src/main/java/org/pdfbox/pdfparser/package.html new file mode 100644 index 0000000..fe012c1 --- /dev/null +++ b/src/main/java/org/pdfbox/pdfparser/package.html @@ -0,0 +1,9 @@ +<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2 Final//EN">
+<html>
+<head>
+
+</head>
+<body>
+The pdfparser package contains classes to parse PDF documents and objects within the document.
+</body>
+</html>
|