Index: src/main/java/org/apache/pdfbox/exceptions/LoggingObject.java =================================================================== --- src/main/java/org/apache/pdfbox/exceptions/LoggingObject.java (revision 763250) +++ src/main/java/org/apache/pdfbox/exceptions/LoggingObject.java (working copy) @@ -55,7 +55,7 @@ I recommend INFO for debug builds and either SEVERE or OFF for production builds. */ - logger_.setLevel(Level.OFF); + logger_.setLevel(Level.WARNING); } return logger_; Index: src/main/java/org/apache/pdfbox/pdmodel/PDDocument.java =================================================================== --- src/main/java/org/apache/pdfbox/pdmodel/PDDocument.java (revision 763250) +++ src/main/java/org/apache/pdfbox/pdmodel/PDDocument.java (working copy) @@ -585,6 +585,21 @@ { return load( url.openStream() ); } + /** + * This will load a document from a url. Used for skipping corrupt + * pdf objects + * + * @param url The url to load the PDF from. + * @param force When true, the parser will skip corrupt pdf objects and + * will continue parsing at the next object in the file + * + * @return The document that was loaded. + * + * @throws IOException If there is an error reading from the stream. + */ + public static PDDocument load(URL url, boolean force) throws IOException{ + return load(url.openStream(), force); + } /** * This will load a document from a url. @@ -614,6 +629,22 @@ { return load( new FileInputStream( filename ) ); } + + /** + * This will load a document from a file. Allows for skipping corrupt pdf + * objects + * + * @param filename The name of the file to load. + * @param force When true, the parser will skip corrupt pdf objects and + * will continue parsing at the next object in the file + * + * @return The document that was loaded. + * + * @throws IOException If there is an error reading from the stream. + */ + public static PDDocument load(String filename, boolean force) throws IOException{ + return load(new FileInputStream( filename ), force); + } /** * This will load a document from a file. @@ -675,8 +706,24 @@ /** * This will load a document from an input stream. + * Allows for skipping corrupt pdf objects * * @param input The stream that contains the document. + * @param force When true, the parser will skip corrupt pdf objects and + * will continue parsing at the next object in the file + * + * @return The document that was loaded. + * + * @throws IOException If there is an error reading from the stream. + */ + public static PDDocument load(InputStream input, boolean force) throws IOException{ + return load(input, null, force); + } + + /** + * This will load a document from an input stream. + * + * @param input The stream that contains the document. * @param scratchFile A location to store temp PDFBox data for this document. * * @return The document that was loaded. @@ -689,6 +736,24 @@ parser.parse(); return parser.getPDDocument(); } + + /** + * This will load a document from an input stream. Allows for skipping corrupt pdf objects + * + * @param input The stream that contains the document. + * @param scratchFile A location to store temp PDFBox data for this document. + * @param force When true, the parser will skip corrupt pdf objects and + * will continue parsing at the next object in the file + * + * @return The document that was loaded. + * + * @throws IOException If there is an error reading from the stream. + */ + public static PDDocument load(InputStream input, RandomAccess scratchFile, boolean force) throws IOException{ + PDFParser parser = new PDFParser( new BufferedInputStream( input ), scratchFile, force); + parser.parse(); + return parser.getPDDocument(); + } /** * This will save this document to the filesystem. Index: src/main/java/org/apache/pdfbox/pdfparser/PDFParser.java =================================================================== --- src/main/java/org/apache/pdfbox/pdfparser/PDFParser.java (revision 763250) +++ src/main/java/org/apache/pdfbox/pdfparser/PDFParser.java (working copy) @@ -20,13 +20,16 @@ import java.io.InputStream; import java.io.IOException; +import java.rmi.server.LogStream; import java.util.Iterator; +import java.util.logging.Level; import org.apache.pdfbox.cos.COSBase; import org.apache.pdfbox.cos.COSDictionary; import org.apache.pdfbox.cos.COSDocument; import org.apache.pdfbox.cos.COSObject; import org.apache.pdfbox.cos.COSStream; +import org.apache.pdfbox.exceptions.LoggingObject; import org.apache.pdfbox.exceptions.WrappedIOException; import org.apache.pdfbox.io.RandomAccess; @@ -49,6 +52,7 @@ private static final String PDF_HEADER = "%PDF-"; private static final String FDF_HEADER = "%FDF-"; private COSDocument document; + private boolean forceParsing = false; /** * Temp file directory. @@ -82,6 +86,23 @@ super(input); this.raf = rafi; } + + /** + * Constructor to allow control over RandomAccessFile. + * Also enables parser to skip corrupt objects to try and force parsing + * @param input The input stream that contains the PDF document. + * @param rafi The RandomAccessFile to be used in internal COSDocument + * @param force When true, the parser will skip corrupt pdf objects and + * will continue parsing at the next object in the file + * + * @throws IOException If there is an error initializing the stream. + */ + public PDFParser(InputStream input, RandomAccess rafi, boolean force) + throws IOException{ + super(input); + this.raf = rafi; + this.forceParsing = force; + } /** * This is the directory where pdfbox will create a temporary file @@ -134,8 +155,22 @@ if(pdfSource.isEOF()){ break; } - wasLastParsedObjectEOF = parseObject(); - + try{ + wasLastParsedObjectEOF = parseObject(); + } + catch(IOException e){ + if(forceParsing){ + /* + * Warning is sent to the PDFBox.log and to the Console that + * we skipped over an object + */ + logger().log(Level.WARNING ,"Parsing Error, Skipping Object", e); + skipToNextObj(); + } + else{ + throw e; + } + } skipSpaces(); } //Test if we saw a trailer section. If not, look for an XRef Stream (Cross-Reference Stream) @@ -187,23 +222,49 @@ pdfSource.close(); } } + /** + * This code ensures that we are at the right point in the stream + * if we find an IOException when parsing. We're assuming that most of the + * lines in the file are formatted properly so that eventually we will get to + * a properly formatted object. + * @throws IOException + */ + private void skipToNextObj() throws IOException { + String s = ""; + while(!pdfSource.isEOF()){ + s = readLine(); + if(s.startsWith("trailer") || + s.matches("\\d* *\\d* *obj") || s.startsWith("xref") || + s.startsWith("startxref") || + s.startsWith("stream")){ + byte[] array = s.getBytes(); + /* Add back the newLine char to the pdfSource since readLine + * removed it. + */ + int i = '\n'; + pdfSource.unread(i); + pdfSource.unread(array); + break; + } + } + } private void parseHeader() throws IOException { - // read first line - String header = readLine(); - // some pdf-documents are broken and the pdf-version is in one of the following lines - if ((header.indexOf( PDF_HEADER ) == -1) && (header.indexOf( FDF_HEADER ) == -1)) - { - header = readLine(); - while ((header.indexOf( PDF_HEADER ) == -1) && (header.indexOf( FDF_HEADER ) == -1)) - { - // if a line starts with a digit, it has to be the first one with data in it - if (Character.isDigit (header.charAt(0))) - break ; - header = readLine(); - } - } + // read first line + String header = readLine(); + // some pdf-documents are broken and the pdf-version is in one of the following lines + if ((header.indexOf( PDF_HEADER ) == -1) && (header.indexOf( FDF_HEADER ) == -1)) + { + header = readLine(); + while ((header.indexOf( PDF_HEADER ) == -1) && (header.indexOf( FDF_HEADER ) == -1)) + { + // if a line starts with a digit, it has to be the first one with data in it + if (Character.isDigit (header.charAt(0))) + break ; + header = readLine(); + } + } // nothing found if ((header.indexOf( PDF_HEADER ) == -1) && (header.indexOf( FDF_HEADER ) == -1)) Index: src/main/java/org/apache/pdfbox/ExtractText.java =================================================================== --- src/main/java/org/apache/pdfbox/ExtractText.java (revision 763250) +++ src/main/java/org/apache/pdfbox/ExtractText.java (working copy) @@ -46,6 +46,7 @@ private static final String END_PAGE = "-endPage"; private static final String SORT = "-sort"; private static final String HTML = "-html"; // jjb - added simple HTML output + private static final String FORCE = "-force"; //enables pdfbox to skip corrupt objects /** * private constructor. @@ -66,6 +67,7 @@ { boolean toConsole = false; boolean toHTML = false; + boolean force = false; boolean sort = false; String password = ""; String encoding = null; @@ -126,6 +128,10 @@ { toConsole = true; } + else if( args[i].equals( FORCE ) ) + { + force = true; + } else { if( pdfFile == null ) @@ -155,7 +161,7 @@ //basically try to load it from a url first and if the URL //is not recognized then try to load it from the file system. URL url = new URL( pdfFile ); - document = PDDocument.load( url ); + document = PDDocument.load(url, force); String fileName = url.getFile(); if( outputFile == null && fileName.length() >4 ) { @@ -164,7 +170,7 @@ } catch( MalformedURLException e ) { - document = PDDocument.load( pdfFile ); + document = PDDocument.load(pdfFile, force); if( outputFile == null && pdfFile.length() >4 ) { outputFile = pdfFile.substring( 0, pdfFile.length() -4 ) + ext; @@ -245,6 +251,7 @@ " -console Send text to console instead of file\n" + " -html Output in HTML format instead of raw text\n" + " -sort Sort the text before writing\n" + + " -force Enables pdfbox to ignore corrupt objects\n" + " -startPage The first page to start extraction(1 based)\n" + " -endPage The last page to extract(inclusive)\n" + " The PDF document to use\n" +