有关pdfbox-1.3.1中Identity-H编码为乱码的解决方法
最近用lucene做一个搜索文档的小程序，其中索引pdf文件时使用pdfbox1.3时出现乱码。
索引pdf的函数如下：（使用pdfbox-1.3.1.jar以及fontbox-1.3.1.jar）
package luceneTest;
import java.io.File;
import java.io.Reader;
import java.io.StringReader;
import org.apache.lucene.document.Document;
import org.apache.lucene.document.Field;
import org.apache.pdfbox.pdmodel.PDDocument;
import org.apache.pdfbox.util.PDFTextStripper;
public class LucenePdf {
public static Document getDocument(File pdf){
String pdfpath = pdf.getAbsolutePath();
PDDocument pdDocument = null;
Document document = new Document();
String title = pdf.getName();
try{
pdDocument = PDDocument.load(pdf);
PDFTextStripper stripper = new PDFTextStripper();
String s1 = stripper.getText(pdDocument);
System.out.println(s1);
Reader contents = new StringReader(s1);
document.add(new Field("title", title, Field.Store.YES, Field.Index.ANALYZED));
document.add(new Field("contents",contents));
document.add(new Field("path", pdfpath, Field.Store.YES, Field.Index.NO));
pdDocument.close();
}catch(Exception e){
e.printStackTrace();
}
return document;
}
}
结果出现了乱码，调试时发现pdf文档的编码格式为Identify-H。我又用了pdfbox-1.2.1.jar来替换pdfbox-1.3.1.jar，结果该文档可以正常显示。通过比较两个不同版本中的org.apache.pdfbox.pdmodel.font.PDFont源代码，我发现1.2.1中有一段代码专门用来处理Identity-H编码，而在1.3.1中则没有。于是将这段代码放入1.3.1版本中的PDFont中。
下面是pdfbox-1.3.1.jar中的org.apache.pdfbox.pdmodel.font.PDFont中有关编码的函数
private void determineEncoding() throws IOException
{
String cmapName = null;
COSName encodingName = null;
COSBase toUnicode = font.getDictionaryObject( COSName.TO_UNICODE );
COSBase encoding = getEncodingObject();
if( toUnicode != null )
{
if ( toUnicode instanceof COSStream )
{
try {
parseCmap(null, ((COSStream)toUnicode).getUnfilteredStream(), null);
}
catch(IOException exception)
{
log.error("Error: Could not load embedded CMAP" );
}
}
else if ( toUnicode instanceof COSName)
{
encodingName = (COSName)toUnicode;
cmap = cmapObjects.get( encodingName.getName() );
if (cmap == null)
{
cmapName = encodingName.getName();
}
}
}
if (encoding != null)
{
if (encoding instanceof COSName)
{
if (cmap == null)
{
encodingName = (COSName)encoding;
cmap = cmapObjects.get( encodingName.getName() );
if (cmap == null)
{
cmapName = encodingName.getName();
//其中红色部分为我后加的为解决idefntity-H编码的代码
if (encodingName.getName().equals( COSName.IDENTITY_H.getName() ))
{
COSArray descendantFontArray =
(COSArray)font.getDictionaryObject( COSName.DESCENDANT_FONTS );
if (descendantFontArray != null)
{
COSDictionary descendantFontDictionary =
(COSDictionary)descendantFontArray.getObject( 0 );
PDFont descendentFont = PDFontFactory.createFont( descendantFontDictionary );
COSDictionary cidsysteminfo =
(COSDictionary)descendentFont.font.getDictionaryObject(COSName.CIDSYSTEMINFO);
if (cidsysteminfo != null)
{
String ordering = cidsysteminfo.getString(COSName.ORDERING);
String registry = cidsysteminfo.getString(COSName.REGISTRY);
cmapName = registry + "-" + ordering+"-UCS2";
}
}
}
}
}
if (cmap == null && cmapName != null)
{
try
{
fontEncoding =
EncodingManager.INSTANCE.getEncoding(encodingName);
}
catch(IOException exception)
{
log.debug("Debug: Could not find encoding for " + encodingName );
}
}
}
else if (encoding instanceof COSDictionary)
{
try
{
fontEncoding = new DictionaryEncoding((COSDictionary)encoding);
}
catch(IOException exception)
{
log.error("Error: Could not create the DictionaryEncoding" );
}
}
else if(encoding instanceof COSStream )
{
if (cmap == null)
{
COSStream encodingStream = (COSStream)encoding;
try
{
parseCmap( null, encodingStream.getUnfilteredStream(), null );
}
catch(IOException exception)
{
log.error("Error: Could not parse the embedded CMAP" );
}
}
}
}
COSDictionary cidsysteminfo = (COSDictionary)font.getDictionaryObject(COSName.CIDSYSTEMINFO);
if (cidsysteminfo != null)
{
String ordering = cidsysteminfo.getString(COSName.ORDERING);
String registry = cidsysteminfo.getString(COSName.REGISTRY);
int supplement = cidsysteminfo.getInt(COSName.SUPPLEMENT);
cmapName = registry + "-" + ordering+ "-" + supplement;
cmapName = CMapSubstitution.substituteCMap( cmapName );
cmap = cmapObjects.get( cmapName );
}
FontMetric metric = getAFM();
if( metric != null )
{
fontEncoding = new AFMEncoding( metric );
}
if (cmap == null && cmapName != null)
{
String resourceName = resourceRootCMAP + cmapName;
try {
parseCmap( resourceRootCMAP, ResourceLoader.loadResource( resourceName ), encodingName );
if( cmap == null && encodingName == null)
{
log.error("Error: Could not parse predefined CMAP file for '" + cmapName + "'" );
}
}
catch(IOException exception)
{
log.error("Error: Could not find predefined CMAP file for '" + cmapName + "'" );
}
}
getEncodingFromFont();
}
这样还是不行，原因是1.3.1版本jar包的cmap中没有adobe-gb1-us2的转码表，于是我又将pdfbox-1.2.1.jar解压缩，将其中\org\apache\pdfbox\resources\cmap目录下的Adobe-GB1-US2表复制到pdfbox-1.3.1.jar中相同目录下的cmap中，这样在将修改后的pdfbox-1.3.1.jar放入工程中，则Identity-H编码问题解决了。