PDFBox

Logged In: NO

Hi both, any progress on this?

Thanks,
Ben

I've attached the two CMap streams that prevent text-extract for my PDF.
ExtractFonts didn't find them as they are resources of PDXObjectForm
objects rather than pages.

Perhaps the PDF creation software is at fault? Ben, can you point me to
the relevant specification? It would be good to cope anyway though if
there is a reasonable approach.

There are two issues:
1) CR in seemly incorrect places e.g. <0000\r>
2) beginbfchar<0000\r> - missing whitespace caused misparse.

Not so nice patch to work-around / illustrate these issue attached.
File Added: bug1702313-1.patch

File Added: SWPRNU+Myriad-Bold-Identity-H.cmap

File Added: RWPRNU+Univers-Light.cmap

Hi, I've recently come across exactly the same error when attempting to
extract text from a certain PDF. Has there been any progress fixing it? I'm
using pdfbox 0.7.3 and fontbox 0.1.0.

Hope you can help, thanks.

Ben Kirby
kirby.bm@gmail.com

Sorry for the delay. Updated extract output at

http://www.hillsdon.net/CMapDocument3.pdf

Stack trace for text extract as before:

Exception in thread "main" java.io.IOException: Error: expected the end of
a dictionary.
at
org.fontbox.cmap.CMapParser.parseNextToken(CMapParser.java:269)
at org.fontbox.cmap.CMapParser.parse(CMapParser.java:117)
...

Thanks, Matt.

Hi Matt,

any update?

Ben

ok, I looked at it some more and I'd like to have you get the latest
nightly build and try to run ExtractText on your original PDF again. If it
doesn't work then run the ExtractFonts again(using the nightly build) and
post the results.

The issue is that there is some extra data at the end of the Cmap stream
and tonight I happened to fix an issue with parsing and having extra data
at the end of the stream for a different user. So I don't know if this is
the same issue but I'd rather have you try the nightly build than have me
chasing a ghost.

Ben

Output with the decryption here
http://www.hillsdon.net/CMapDocument2.pdf

Thanks.

shoot, I think your document was encrypted. It needs to be decrypted for
the extraction to work, I should have had that as part of the program. Can
you take the attached program and add the lines after the PDDocument.load
call

if( doc.isEncrypted() )
{
doc.decrypt( "" );
}

and resend the CMapDocument.pdf

Thanks,
Ben

Result too large to attach. Please see
http://www.hillsdon.net/CMapDocument.pdf

Attached is a simple java program that will create a new pseudo PDF
document that contains just the Font information. Please run it on the
problem PDF and upload the resulting CmapDocument.pdf

It is a simple command line program, first compile then run it like this

java ExtractFonts my.pdf

Let me know if you have any questions getting it running.

Ben
File Added: ExtractFonts.java

No change unfortunately - with FontBox-0.2.0-dev-20070424 the stack trace
is identical.
Exception in thread "main" java.io.IOException: Error: expected the end of
a dictionary.
at
org.fontbox.cmap.CMapParser.parseNextToken(CMapParser.java:269)
...

I just update the CMapParser with a bug from

https://sourceforge.net/forum/message.php?msg_id=4269559

please get tonights FontBox build and give it a try

http://www.fontbox.org/fontbox

Hi Ben, thanks for the quick response.

Using the nightly build [1] the stack trace is the same except for line
numbers:

Exception in thread "main" java.io.IOException: Error: expected the end of
a dictionary.
at
org.fontbox.cmap.CMapParser.parseNextToken(CMapParser.java:269)
at org.fontbox.cmap.CMapParser.parse(CMapParser.java:117)
at org.pdfbox.pdmodel.font.PDFont.parseCmap(PDFont.java:509)
at org.pdfbox.pdmodel.font.PDFont.encode(PDFont.java:380)
at
org.pdfbox.util.PDFStreamEngine.showString(PDFStreamEngine.java:325)
at org.pdfbox.util.operator.ShowText.process(ShowText.java:64)
at
org.pdfbox.util.PDFStreamEngine.processOperator(PDFStreamEngine.java:452)
at
org.pdfbox.util.PDFStreamEngine.processSubStream(PDFStreamEngine.java:215)
at
org.pdfbox.util.PDFStreamEngine.processStream(PDFStreamEngine.java:174)
at
org.pdfbox.util.PDFTextStripper.processPage(PDFTextStripper.java:336)
at
org.pdfbox.util.PDFTextStripper.processPages(PDFTextStripper.java:259)
at
org.pdfbox.util.PDFTextStripper.writeText(PDFTextStripper.java:216)
at
org.pdfbox.util.PDFTextStripper.getText(PDFTextStripper.java:149)
...

Extracting the fonts sounds ideal.

[1] http://www.pdfbox.org/dist/PDFBox-0.7.4-dev-20070418.zip

Hi Matt,

Can you try one for me first; upgrade to the latest nightly build of
PDFBox( http://www.pdfbox.org/dist/ ) and see if this is still an issue.
There have been some changes to the CMAPParser.

If it is still an issue I think we can write a simple program to extract
just the fonts from your PDF and that should be enough for me to fix the
bug.

Ben