Owen McGovern created PDFBOX-5485:
-------------------------------------
Summary: Stackoverflow writing out a subset of PDF pages -
COSWriterObjectStream
Key: PDFBOX-5485
URL: https://issues.apache.org/jira/browse/PDFBOX-5485
Project: PDFBox
Issue Type: Bug
Components: Writing
Affects Versions: 3.0.0 JBIG2
Environment: MacOS, but likely not OS specific.
Reporter: Owen McGovern
Version: org.apache.pdfbox:pdfbox:3.0.0-alpha3
In a subset of PDFs I process, I cannot extract a range of PDF pages and write
them out to a new PDF. ( As part of test code )
Here's the Kotlin code I use
{code:java}
fun extractPages(documentName: String, fromPage: Int, toPage: Int) : Path {
val pdfFile = Paths.get("data", "input", "PDFS", "${documentName}.pdf")
val pdfPagesFile = Paths.get("data", "input", "PDFS",
"${documentName}_Page_$fromPage-$toPage.pdf")
val pdfDoc = org.apache.pdfbox.Loader.loadPDF(pdfFile.toFile())
val pageExtractor = PageExtractor(pdfDoc, fromPage, toPage)
val pdfPages = pageExtractor.extract()
pdfPages.save(pdfPagesFile.toFile())
return pdfPagesFile
}{code}
It doesn't occur in all PDFS... maybe 10-20% of the PDFs I use.
The a slice of the stack trace is
{code:java}
java.lang.StackOverflowError
at java.base/java.util.HashMap.tableSizeFor(HashMap.java:380)
at java.base/java.util.HashMap.<init>(HashMap.java:453)
at java.base/java.util.LinkedHashMap.<init>(LinkedHashMap.java:347)
at java.base/java.util.HashSet.<init>(HashSet.java:162)
at java.base/java.util.LinkedHashSet.<init>(LinkedHashSet.java:154)
at org.apache.pdfbox.util.SmallMap.entrySet(SmallMap.java:380)
at org.apache.pdfbox.cos.COSDictionary.entrySet(COSDictionary.java:1225)
at
org.apache.pdfbox.pdfwriter.compress.COSWriterObjectStream.writeCOSDictionary(COSWriterObjectStream.java:336)
at
org.apache.pdfbox.pdfwriter.compress.COSWriterObjectStream.writeObject(COSWriterObjectStream.java:230)
at
org.apache.pdfbox.pdfwriter.compress.COSWriterObjectStream.writeCOSArray(COSWriterObjectStream.java:319)
at
org.apache.pdfbox.pdfwriter.compress.COSWriterObjectStream.writeObject(COSWriterObjectStream.java:226)
at
org.apache.pdfbox.pdfwriter.compress.COSWriterObjectStream.writeCOSDictionary(COSWriterObjectStream.java:341)
at
org.apache.pdfbox.pdfwriter.compress.COSWriterObjectStream.writeObject(COSWriterObjectStream.java:230)
at
org.apache.pdfbox.pdfwriter.compress.COSWriterObjectStream.writeCOSDictionary(COSWriterObjectStream.java:341)
at
org.apache.pdfbox.pdfwriter.compress.COSWriterObjectStream.writeObject(COSWriterObjectStream.java:230)
at
org.apache.pdfbox.pdfwriter.compress.COSWriterObjectStream.writeCOSArray(COSWriterObjectStream.java:319)
at
org.apache.pdfbox.pdfwriter.compress.COSWriterObjectStream.writeObject(COSWriterObjectStream.java:226)
at
org.apache.pdfbox.pdfwriter.compress.COSWriterObjectStream.writeCOSDictionary(COSWriterObjectStream.java:341)
at
org.apache.pdfbox.pdfwriter.compress.COSWriterObjectStream.writeObject(COSWriterObjectStream.java:230)
at
org.apache.pdfbox.pdfwriter.compress.COSWriterObjectStream.writeCOSDictionary(COSWriterObjectStream.java:341)
at
org.apache.pdfbox.pdfwriter.compress.COSWriterObjectStream.writeObject(COSWriterObjectStream.java:230)
at
org.apache.pdfbox.pdfwriter.compress.COSWriterObjectStream.writeCOSArray(COSWriterObjectStream.java:319)
at
org.apache.pdfbox.pdfwriter.compress.COSWriterObjectStream.writeObject(COSWriterObjectStream.java:226)
at
org.apache.pdfbox.pdfwriter.compress.COSWriterObjectStream.writeCOSDictionary(COSWriterObjectStream.java:341)
at
org.apache.pdfbox.pdfwriter.compress.COSWriterObjectStream.writeObject(COSWriterObjectStream.java:230)
at
org.apache.pdfbox.pdfwriter.compress.COSWriterObjectStream.writeCOSDictionary(COSWriterObjectStream.java:341)
{code}
As I mentioned, hits some PDFs, not all.
I legally cannot share the original source PDFs but it looks like a recursive
loop in writeCOSDictionary and writeObject in COSWriterObjectStream.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]