[
https://issues.apache.org/jira/browse/PDFBOX-5894?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17894656#comment-17894656
]
Derek Wickern commented on PDFBOX-5894:
---------------------------------------
There is an issue now that {{Loader.loadFDF(String)}} and
{{Loader.loadFDF(File)}} overloads will leak a {{{}FileChannel{}}}, because the
{{RandomAccessReadBufferedFile}} won't be closed.
{{PDDocument}} holds a reference to the {{RandomAccessRead}} and closes it when
the document is closed:
[https://github.com/apache/pdfbox/blob/1ea3ae6fda8558a4a006f982d93a49ff53c89f19/pdfbox/src/main/java/org/apache/pdfbox/pdmodel/PDDocument.java#L1272]
Maybe {{FDFDocument}} should also keep a reference to the {{RandomAccessRead?}}
> Problem loading FDF
> -------------------
>
> Key: PDFBOX-5894
> URL: https://issues.apache.org/jira/browse/PDFBOX-5894
> Project: PDFBox
> Issue Type: Bug
> Affects Versions: 3.0.3 PDFBox
> Reporter: Derek Wickern
> Assignee: Tilman Hausherr
> Priority: Major
> Fix For: 3.0.4 PDFBox, 4.0.0
>
> Attachments: sticky-notes.fdf
>
>
> Loading the attached {{sticky-notes.fdf}} produces incorrect results. PDFBox
> logs a bunch of errors like this:
> {noformat}
> 15:59:55.131 [main] ERROR org.apache.pdfbox.cos.COSObject -- Can't
> dereference COSObject{6 0 R}
> java.io.IOException: org.apache.pdfbox.io.RandomAccessReadBufferedFile
> already closed
> at
> org.apache.pdfbox.io.RandomAccessReadBufferedFile.checkClosed(RandomAccessReadBufferedFile.java:266)
> at
> org.apache.pdfbox.io.RandomAccessReadBufferedFile.getPosition(RandomAccessReadBufferedFile.java:118)
> at
> org.apache.pdfbox.pdfparser.COSParser.dereferenceCOSObject(COSParser.java:623)
> at org.apache.pdfbox.cos.COSObject.getObject(COSObject.java:117)
> at
> org.apache.pdfbox.cos.COSDocument.getObjectsByType(COSDocument.java:273)
> at
> org.apache.pdfbox.cos.COSDocument.getObjectsByType(COSDocument.java:255)
> at
> org.apache.pdfbox.cos.COSDocument.getObjectsByType(COSDocument.java:241)
> at
> PdfboxLoadFDFRegressionTest.main(PdfboxLoadFDFRegressionTest.java:16){noformat}
> The issue seems to be that {{Loader.loadFDF}} closes the
> {{RandomAccessReadBuffer}} immediately after parsing, but the read buffer is
> still being used. A workaround is to make sure the read buffer is kept open
> until you're done with the {{{}FDFDocument{}}}.
> Reproduction code:
> {code:java}
> import org.apache.pdfbox.Loader;
> import org.apache.pdfbox.cos.COSName;
> import org.apache.pdfbox.io.RandomAccessReadBuffer;
> import org.apache.pdfbox.pdfparser.FDFParser;
> import java.nio.file.Files;
> import java.nio.file.Path;
> public class PdfboxLoadFDFRegressionTest {
> public static void main(String[] args) throws Exception {
> var filename = "sticky-notes.fdf";
> // broken:
> try (var fdf = Loader.loadFDF(filename)) {
> var annots = fdf.getDocument().getObjectsByType(COSName.ANNOT);
> System.out.println(annots); // []
> }
> // workaround:
> try (var readBuffer = new
> RandomAccessReadBuffer(Files.readAllBytes(Path.of(filename)));
> var fdf = new FDFParser(readBuffer).parse()) {
> var annots = fdf.getDocument().getObjectsByType(COSName.ANNOT);
> System.out.println(annots); // [COSObject{3 0 R}, COSObject{4 0
> R}, COSObject{5 0 R}, COSObject{6 0 R}]
> }
> }
> }{code}
>
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@pdfbox.apache.org
For additional commands, e-mail: dev-h...@pdfbox.apache.org
