### Eclipse Workspace Patch 1.0 #P PDFBox Index: pdfbox/src/main/java/org/apache/pdfbox/pdfparser/PDFStreamParser.java =================================================================== --- pdfbox/src/main/java/org/apache/pdfbox/pdfparser/PDFStreamParser.java (revision 1033990) +++ pdfbox/src/main/java/org/apache/pdfbox/pdfparser/PDFStreamParser.java (working copy) @@ -24,6 +24,8 @@ import java.util.List; import java.util.NoSuchElementException; +import org.apache.commons.logging.Log; +import org.apache.commons.logging.LogFactory; import org.apache.pdfbox.cos.COSBase; import org.apache.pdfbox.cos.COSBoolean; import org.apache.pdfbox.cos.COSDictionary; @@ -45,6 +47,11 @@ */ public class PDFStreamParser extends BaseParser { + /** + * Log instance. + */ + private static final Log log = LogFactory.getLog(PDFStreamParser.class); + private List streamObjects = new ArrayList( 100 ); private RandomAccess file; private PDFOperator lastBIToken = null; @@ -174,8 +181,15 @@ { token = parseNextToken(); } - } catch (IOException e) { - throw new RuntimeException(e); + } + catch (IOException e) + { + if (!forceParsing) { + throw new RuntimeException(e); + } + log.warn("Error parsing next token.", e); + + token = null; } } @@ -397,7 +411,11 @@ //(UnderstandingWebSphereClassLoaders.pdf) which has EI as part //of the image data and will stop parsing prematurely if there is //not a check for EI. - while( !(isWhitespace( twoBytesAgo ) && + // + //the PDF attached to https://issues.apache.org/jira/browse/PDFBOX-789 (pdf_euba.pdf) + //contains the sequence <0x0C>EI<0x0A>. According to 7.3.8 (which is referenced by 8.9.7) + //says that the data portion should end with end-of-line marker. + while( !(isEOL( twoBytesAgo ) && lastByte == 'E' && currentByte == 'I' && isWhitespace() //&& Index: pdfbox/src/main/java/org/apache/pdfbox/pdfparser/BaseParser.java =================================================================== --- pdfbox/src/main/java/org/apache/pdfbox/pdfparser/BaseParser.java (revision 1033990) +++ pdfbox/src/main/java/org/apache/pdfbox/pdfparser/BaseParser.java (working copy) @@ -186,7 +186,13 @@ char r = (char)pdfSource.read(); if( r != 'R' ) { - throw new IOException( "expected='R' actual='" + r + "' " + pdfSource ); + // some real world pdfs contain faulty data. if forceParsing assume it's R. + String error = "expected='R' actual='" + r + "' " + pdfSource; + + if (!forceParsing) { + throw new IOException( error ); + } + log.warn( error ); } COSObjectKey key = new COSObjectKey(((COSInteger) number).intValue(), ((COSInteger) generationNumber).intValue()); @@ -485,6 +491,9 @@ pdfSource.unread( ENDSTREAM ); return; } + else { + out.write(ENDSTREAM, 0, 8); + } }else{ out.write(ENDSTREAM, 0, 7); } @@ -506,6 +515,9 @@ pdfSource.unread( ENDOBJ ); return; } + else { + out.write(ENDOBJ, 0, 5); + } }else{ out.write(ENDOBJ, 0, 4); } @@ -1376,7 +1388,10 @@ } catch( NumberFormatException e ) { - pdfSource.unread(intBuffer.toString().getBytes()); + // In case of an exception it's best not to unread the data because you run the risk + // of getting into an unlimited loop (especially so when forceParsing is true). + // For example with test-integer-too-large.pdf and forceParsing true, the parser gets + // stuck in an unlimited loop when the data is unread throw new IOException( "Error: Expected an integer type, actual='" + intBuffer + "'" ); } return retval; Index: pdfbox/src/main/java/org/apache/pdfbox/pdfparser/PDFParser.java =================================================================== --- pdfbox/src/main/java/org/apache/pdfbox/pdfparser/PDFParser.java (revision 1033990) +++ pdfbox/src/main/java/org/apache/pdfbox/pdfparser/PDFParser.java (working copy) @@ -17,9 +17,8 @@ package org.apache.pdfbox.pdfparser; import java.io.File; -import java.io.InputStream; import java.io.IOException; - +import java.io.InputStream; import java.util.ArrayList; import java.util.Iterator; import java.util.List; @@ -34,11 +33,8 @@ import org.apache.pdfbox.cos.COSObject; import org.apache.pdfbox.exceptions.WrappedIOException; import org.apache.pdfbox.io.RandomAccess; - import org.apache.pdfbox.pdmodel.PDDocument; - import org.apache.pdfbox.pdmodel.fdf.FDFDocument; - import org.apache.pdfbox.persistence.util.COSObjectKey; /** @@ -125,6 +121,18 @@ } /** + * Returns true if parsing should be continued. By default, forceParsing is returned. + * This can be overridden to add application specific handling (for example to stop + * parsing when the number of exceptions thrown exceed a certain number). + * + * @param e The exception if vailable. Can be null if there is no exception available + */ + protected boolean isContinueOnError(Exception e) + { + return forceParsing; + } + + /** * This will parse the stream and populate the COSDocument object. This will close * the stream when it is done parsing. * @@ -173,7 +181,7 @@ } catch(IOException e) { - if(forceParsing) + if(isContinueOnError(e)) { /* * Warning is sent to the PDFBox.log and to the Console that @@ -508,7 +516,11 @@ //" genNumber=" + genNum + " key='" + objectKey + "'" ); if( !objectKey.equals( "obj" ) ) { - throw new IOException("expected='obj' actual='" + objectKey + "' " + pdfSource); + if (!isContinueOnError(null) || !objectKey.equals("o")) { + throw new IOException("expected='obj' actual='" + objectKey + "' " + pdfSource); + } + //assume that "o" was meant to be "obj" (this is a workaround for + // PDFBOX-773 attached PDF Andersens_Fairy_Tales.pdf). } } else @@ -577,38 +589,10 @@ } else 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() ); - } - 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" ) ) - { - if( isClosing() ) - { - //found a case with 17506.pdf object 41 that was like this - //41 0 obj [/Pattern /DeviceGray] ] endobj - //notice the second array close, here we are reading it - //and ignoring and attempting to continue - pdfSource.read(); - } - skipSpaces(); - String thirdPossibleEndObj = readString(); - if( !thirdPossibleEndObj.equals( "endobj" ) ) - { - throw new IOException("expected='endobj' firstReadAttempt='" + endObjectKey + "' " + - "secondReadAttempt='" + secondEndObjectKey + "' " + pdfSource); - } - } - } + //It is possible that the endobj is missing, there + //are several PDFs out there that do that so. Unread + //and assume that endobj was missing + pdfSource.unread( endObjectKey.getBytes() ); } } skipSpaces(); Index: pdfbox/src/main/java/org/apache/pdfbox/util/PDFTextStripper.java =================================================================== --- pdfbox/src/main/java/org/apache/pdfbox/util/PDFTextStripper.java (revision 1033990) +++ pdfbox/src/main/java/org/apache/pdfbox/util/PDFTextStripper.java (working copy) @@ -30,6 +30,8 @@ import java.util.Vector; import java.util.regex.Pattern; +import org.apache.commons.logging.Log; +import org.apache.commons.logging.LogFactory; import org.apache.pdfbox.cos.COSDocument; import org.apache.pdfbox.cos.COSStream; import org.apache.pdfbox.exceptions.CryptographyException; @@ -59,6 +61,10 @@ */ public class PDFTextStripper extends PDFStreamEngine { + /** + * Log instance. + */ + private static final Log log = LogFactory.getLog(PDFTextStripper.class); private static final String thisClassName = PDFTextStripper.class.getSimpleName().toLowerCase(); @@ -362,7 +368,16 @@ if( contentStream != null ) { COSStream contents = contentStream.getStream(); - processPage( nextPage, contents ); + try { + processPage( nextPage, contents ); + } + catch(IOException e) + { + if (!isForceParsing()) { + throw e; + } + log.warn("Error processing page.", e); + } } } } Index: pdfbox/src/main/java/org/apache/pdfbox/pdmodel/common/PDStream.java =================================================================== --- pdfbox/src/main/java/org/apache/pdfbox/pdmodel/common/PDStream.java (revision 1037894) +++ pdfbox/src/main/java/org/apache/pdfbox/pdmodel/common/PDStream.java (working copy) @@ -164,7 +164,8 @@ } else if( base instanceof COSArray ) { - if (((COSArray)base).size() > 0) { + // with certain PDFs, the stream contains no items or only null. + if (((COSArray)base).size() > 0 && ((COSArray)base).getObject(0) != null) { retval = new PDStream( new COSStreamArray( (COSArray)base ) ); } } Index: pdfbox/src/main/java/org/apache/pdfbox/pdmodel/common/COSStreamArray.java =================================================================== --- pdfbox/src/main/java/org/apache/pdfbox/pdmodel/common/COSStreamArray.java (revision 1033990) +++ pdfbox/src/main/java/org/apache/pdfbox/pdmodel/common/COSStreamArray.java (working copy) @@ -205,12 +205,17 @@ for( int i=0;i