0.95: Metadata is not valid XML document even if there are no metadata defined in FO source
Hello,
I have a problem with metadada part of PDF created by Apache FOP 0.95 (using
Sun Java SE 1.6.0_10 on MS Windows 2000). Although resulted PDF looks OK in
Acrobat Reader, I am not able to process resulted PDF using iText 2.1.3. It
throws Exception shown below. After several e-mails among iText developers (and
me), it looks like even if I do not specify any metadata in FO, FOP 0.95
produces PDF with compressed not valid XML metadata section according to iText
developers. If I specify metadata according to Your example at
http://xmlgraphics.apache.org/fop/0.95/metadata.html#xmp-example
there is the same Exception. So, could You, please, verify, if the problem is
really caused by FOP and if yes, to repait it? Thank You very much in advance.
Stepan Rybar
Původní zpráva
Od: Paulo Soares
Předmět: Re: [iText-questions] iText 2.1.3: Invalid byte 1 of 1-byte UTF-8
sequence Exception while processing PDF from Apache FOP 0.95
Datum: 05.11.2008 15:44:00
The metadata can be compressed, the problem is that the metadata is not a
valid
xml document (or at least it wasn't in the pdf I loooked at). iText should
ignore the error and carry on but it currently doesn't. The problem starts at
FOP in any case.
Paulo
From: 1T3XT info []
Sent: Wednesday, November 05, 2008 2:08 PM
To: Post all your questions about iText here
Subject: Re: [iText-questions] iText 2.1.3: Invalid byte 1 of 1-byte UTF-8
sequence Exception while processing PDF from Apache FOP 0.95
Stepan RYBAR wrote:
Hello,
thank You answer. But after some tests I guess, that this is not caused by
missing metadata.
Think again.
Even if there are no metadata in source.fo
Look at your PDF document.
In the root dictionary, there's a reference to XMP data.
/Metadata 7 0 R
If we have a look at object 5, we see:
7 0 obj
/Type /Metadata
/Subtype /XML
/Length 12 0 R
/Filter /FlateDecode
stream
xoe?QN~jÂ0?}?ü++?×?I??3¤.a){q??÷\??oem?#?Ú?öIu^...??
W+OE=?îÍ9'ç#?u.Ë=X´...?]Ä?óý?#bü?Íé1/4šÁNß7?VÍA~Z6{9Qxšô{ýž¨y}¨?`s?
?žu`2ÍÂ~ÂQ?]Y]?¨Ýómy^(2)1?-^^(1)}'f?0G?'l8?¤s?X?T'
^(1)s?wEUR^(2)Ç?o`?D++E`ë?Ë(c)HÇü/yÜ/[EMAIL
PROTECTED]'^(1)?ôvA`?EURFk6âa~?ÎFWqJ?é0yé§$W^?B+x
a~?3/4äî?ÁO~ëE`U`s??!gÍ??.}??...kU^?Z?
Oops... What's this?
This is a compressed XMP stream.
That's not allowed in PDF!
# Původní zpráva
# Od: Stepan RYBAR
# Předmět: Re: [iText-questions] iText 2.1.3: Invalid byte 1 of 1-byte UTF-8
# sequence Exception while processing PDF from Apache FOP 0.95
# Datum: 05.11.2008 13:52:34
#
# Hello,
#
# thank You answer. But after some tests I guess, that this is not caused by
# missing metadata. Even if there are no metadata in source.fo, I can use
iText
# to print to stdout total number of pages. So it means, that iText can read
# source.pdf without throwing any exception. See attached source.fo
(commented
# metadata), source.pdf (commented creation of output) and Java code for
iText,
# which passes without problem.
#
# Once I want to create output file and save it without any modification (to
# prevent mistake in font encoding setting), there is an Exception:
#
# ExceptionConverter:
# com.sun.org.apache.xerces.internal.impl.io.MalformedByteSequenceException:
# Invalid byte 1 of 1-byte UTF-8 sequence.
#
# Stepan Rybar
#
# Původní zpráva
# Od: Paulo Soares
# Předmět: Re: [iText-questions] iText 2.1.3: Invalid byte 1 of 1-byte UTF-8
# sequence Exception while processing PDF from Apache FOP 0.95
# Datum: 04.11.2008 17:23:03
#
# The metadata generated by FOP is broken. They have:
#
# x:xmpmeta
#
# instead of:
#
# x:xmpmeta xmlns:x=adobe:ns:meta/ x:xmptk=3.1-702
#
# The namespace is not defined and the parser complains about it.
#
# Paulo
#
# From: Stepan RYBAR []
# Sent: Tuesday, November 04, 2008 4:11 PM
# To: [EMAIL PROTECTED]
# Subject: [iText-questions] iText 2.1.3: Invalid byte 1 of 1-byte UTF-8
# sequence
# Exception while processing PDF from Apache FOP 0.95
#
# Hello,
#
# I am trying to use iText for adding page numbers to the already existed PDF
# using PDFStamper class. I have problem. PDFs, which I create using Apache
FOP
# 0.95, cause Exception as shown below, although they look OK in Adobe Acrobat
# Reader 8.1.2. I guess, that this is problem of missing or wrong encoding
# somewhere (but where?). Attached to this e-mail are: FO source, resulted
PDF,
# Java code for iText. I am using Sun Java SE 1.6.0_10 on MS Windows 2000.
# Please,
# can You point me, where I am making miskate?
#
# Thank You. Stepan
#
# L:\Documents\Capitol\vypisyKv1\_otherL:\RunFiles\jdk\jre\bin\java -cp
# .;iText-
# 2.1.3.jar AddPageNumbersToExistingPageNumberPDF
# ExceptionConverter:
#
Re: 0.95: Metadata is not valid XML document even if there are no metadata defined in FO source
Ok, so we have two problems:
1. Compressed Metadata stream: FOP doesn't compress the metadata stream
by default. The problem is that FOP 0.95 has a demo configuration file
that contains a filter list that causes all stream to be compressed.
Please remove all filterList elements including their children from
the configuration file. I've already committed a fix to FOP Trunk which
makes this unnecessary some time ago:
http://svn.apache.org/viewvc?rev=695491view=rev
2. The XMP is not valid RDF because the namespace declarations are
missing: If that happens you're using a JAXP TraX implementation that
has a buggy serializer. I've just tried with a fresh Sun Java 1.6.0_10
installation on WinXP and didn't have that problem. I assume you've
installed a different JAXP implementation.
On 05.11.2008 16:28:54 Stepan RYBAR wrote:
Hello,
I have a problem with metadada part of PDF created by Apache FOP 0.95
(using Sun Java SE 1.6.0_10 on MS Windows 2000). Although resulted PDF
looks OK in Acrobat Reader, I am not able to process resulted PDF using
iText 2.1.3. It throws Exception shown below. After several e-mails
among iText developers (and me), it looks like even if I do not specify
any metadata in FO, FOP 0.95 produces PDF with compressed not valid XML
metadata section according to iText developers. If I specify metadata
according to Your example at
http://xmlgraphics.apache.org/fop/0.95/metadata.html#xmp-example
there is the same Exception. So, could You, please, verify, if the
problem is really caused by FOP and if yes, to repait it? Thank You
very much in advance.
Stepan Rybar
Pùvodní zpráva
Od: Paulo Soares
Pøedmìt: Re: [iText-questions] iText 2.1.3: Invalid byte 1 of 1-byte UTF-8
sequence Exception while processing PDF from Apache FOP 0.95
Datum: 05.11.2008 15:44:00
The metadata can be compressed, the problem is that the metadata is not a
valid
xml document (or at least it wasn't in the pdf I loooked at). iText should
ignore the error and carry on but it currently doesn't. The problem starts
at
FOP in any case.
Paulo
From: 1T3XT info []
Sent: Wednesday, November 05, 2008 2:08 PM
To: Post all your questions about iText here
Subject: Re: [iText-questions] iText 2.1.3: Invalid byte 1 of 1-byte UTF-8
sequence Exception while processing PDF from Apache FOP 0.95
Stepan RYBAR wrote:
Hello,
thank You answer. But after some tests I guess, that this is not caused by
missing metadata.
Think again.
Even if there are no metadata in source.fo
Look at your PDF document.
In the root dictionary, there's a reference to XMP data.
/Metadata 7 0 R
If we have a look at object 5, we see:
7 0 obj
/Type /Metadata
/Subtype /XML
/Length 12 0 R
/Filter /FlateDecode
stream
xoe?QN~jÂ0?}?ü++?×?I??3¤.a){q??÷\??oem?#?Ú?öIu^...??
W+OE=?îÍ9'ç#?u.Ë=X´...?]Ä?óý?#bü?Íé1/4ÁNß7?VÍA~Z6{9Qxô{ý¨y}¨?`s?
?u`2ÍÂ~ÂQ?]Y]?¨Ýómy^(2)1?-^^(1)}'f?0G?'l8?¤s?X?T'
^(1)s?wEUR^(2)Ç?o`?D++E`ë?Ë(c)HÇü/yÜ/[EMAIL
PROTECTED]'^(1)?ôvA`?EURFk6âa~?ÎFWqJ?é0yé§$W^?B+x
a~?3/4äî?ÁO~ëE`U`s??!gÍ??.}??...kU^?Z?
Oops... What's this?
This is a compressed XMP stream.
That's not allowed in PDF!
# Pùvodní zpráva
# Od: Stepan RYBAR
# Pøedmìt: Re: [iText-questions] iText 2.1.3: Invalid byte 1 of 1-byte UTF-8
# sequence Exception while processing PDF from Apache FOP 0.95
# Datum: 05.11.2008 13:52:34
#
# Hello,
#
# thank You answer. But after some tests I guess, that this is not caused by
# missing metadata. Even if there are no metadata in source.fo, I can use
iText
# to print to stdout total number of pages. So it means, that iText can read
# source.pdf without throwing any exception. See attached source.fo
(commented
# metadata), source.pdf (commented creation of output) and Java code for
iText,
# which passes without problem.
#
# Once I want to create output file and save it without any modification (to
# prevent mistake in font encoding setting), there is an Exception:
#
# ExceptionConverter:
# com.sun.org.apache.xerces.internal.impl.io.MalformedByteSequenceException:
# Invalid byte 1 of 1-byte UTF-8 sequence.
#
# Stepan Rybar
#
# Pùvodní zpráva
# Od: Paulo Soares
# Pøedmìt: Re: [iText-questions] iText 2.1.3: Invalid byte 1 of 1-byte UTF-8
# sequence Exception while processing PDF from Apache FOP 0.95
# Datum: 04.11.2008 17:23:03
#
# The metadata generated by FOP is broken. They have:
#
# x:xmpmeta
#
# instead of:
#
# x:xmpmeta xmlns:x=adobe:ns:meta/ x:xmptk=3.1-702
#
# The namespace is not defined and the parser complains about it.
#
# Paulo
#
#
Re: 0.95: Metadata is not valid XML document even if there are no metadata defined in FO source
ad 1) I did it, it looks like it works.
ad 2) ...you're using a JAXP TraX... You are right. Although I do not know,
that (and even how to) I changed anything about JAXP, when I run this on my
second comp (MS Windows Vista 64-bit Enterprise, Sun Java 64-bit 1.6.0_10-b33),
it works fine.
Thank You. Stepan
Původní zpráva
Od: Jeremias Maerki [EMAIL PROTECTED]
Předmět: Re: 0.95: Metadata is not valid XML document even if there are no
metadata defined in FO source
Datum: 05.11.2008 17:51:56
Ok, so we have two problems:
1. Compressed Metadata stream: FOP doesn't compress the metadata stream
by default. The problem is that FOP 0.95 has a demo configuration file
that contains a filter list that causes all stream to be compressed.
Please remove all filterList elements including their children from
the configuration file. I've already committed a fix to FOP Trunk which
makes this unnecessary some time ago:
http://svn.apache.org/viewvc?rev=695491view=rev
2. The XMP is not valid RDF because the namespace declarations are
missing: If that happens you're using a JAXP TraX implementation that
has a buggy serializer. I've just tried with a fresh Sun Java 1.6.0_10
installation on WinXP and didn't have that problem. I assume you've
installed a different JAXP implementation.
On 05.11.2008 16:28:54 Stepan RYBAR wrote:
Hello,
I have a problem with metadada part of PDF created by Apache FOP 0.95
(using Sun Java SE 1.6.0_10 on MS Windows 2000). Although resulted PDF
looks OK in Acrobat Reader, I am not able to process resulted PDF using
iText 2.1.3. It throws Exception shown below. After several e-mails
among iText developers (and me), it looks like even if I do not specify
any metadata in FO, FOP 0.95 produces PDF with compressed not valid XML
metadata section according to iText developers. If I specify metadata
according to Your example at
http://xmlgraphics.apache.org/fop/0.95/metadata.html#xmp-example
there is the same Exception. So, could You, please, verify, if the
problem is really caused by FOP and if yes, to repait it? Thank You
very much in advance.
Stepan Rybar
Pùvodní zpráva
Od: Paulo Soares
Pøedmìt: Re: [iText-questions] iText 2.1.3: Invalid byte 1 of 1-byte UTF-8
sequence Exception while processing PDF from Apache FOP 0.95
Datum: 05.11.2008 15:44:00
The metadata can be compressed, the problem is that the metadata is not a
valid
xml document (or at least it wasn't in the pdf I loooked at). iText should
ignore the error and carry on but it currently doesn't. The problem starts
at
FOP in any case.
Paulo
From: 1T3XT info []
Sent: Wednesday, November 05, 2008 2:08 PM
To: Post all your questions about iText here
Subject: Re: [iText-questions] iText 2.1.3: Invalid byte 1 of 1-byte UTF-8
sequence Exception while processing PDF from Apache FOP 0.95
Stepan RYBAR wrote:
Hello,
thank You answer. But after some tests I guess, that this is not caused
by
missing metadata.
Think again.
Even if there are no metadata in source.fo
Look at your PDF document.
In the root dictionary, there's a reference to XMP data.
/Metadata 7 0 R
If we have a look at object 5, we see:
7 0 obj
/Type /Metadata
/Subtype /XML
/Length 12 0 R
/Filter /FlateDecode
stream
xoe?QN~jÂ0?}?ü++?×?I??3¤.a){q??÷\??oem?#?Ú?öIu^...??
W+OE=?îÍ9'ç#?u.Ë=X´...?]Ä?óý?#bü?Íé1/4ÁNß7?VÍA~Z6{9Qxô{ý¨y}¨?`s?
?u`2ÍÂ~ÂQ?]Y]?¨Ýómy^(2)1?-^^(1)}'f?0G?'l8?¤s?X?T'
^(1)s?wEUR^(2)Ç?o`?D++E`ë?Ë(c)HÇü/yÜ/[EMAIL
PROTECTED]'^(1)?ôvA`?EURFk6âa~?ÎFWqJ?é0yé§$W^?B+x
a~?3/4äî?ÁO~ëE`U`s??!gÍ??.}??...kU^?Z?
Oops... What's this?
This is a compressed XMP stream.
That's not allowed in PDF!
# Pùvodní zpráva
# Od: Stepan RYBAR
# Pøedmìt: Re: [iText-questions] iText 2.1.3: Invalid byte 1 of 1-byte UTF-8
# sequence Exception while processing PDF from Apache FOP 0.95
# Datum: 05.11.2008 13:52:34
#
# Hello,
#
# thank You answer. But after some tests I guess, that this is not caused by
# missing metadata. Even if there are no metadata in source.fo, I can use
iText
# to print to stdout total number of pages. So it means, that iText can read
# source.pdf without throwing any exception. See attached source.fo
(commented
# metadata), source.pdf (commented creation of output) and Java code for
iText,
# which passes without problem.
#
# Once I want to create output file and save it without any modification (to
# prevent mistake in font encoding setting), there is an Exception:
#
# ExceptionConverter:
# com.sun.org.apache.xerces.internal.impl.io.MalformedByteSequenceException:
#
Re: 0.95: Metadata is not valid XML document even if there are no metadata defined in FO source
On 05.11.2008 18:32:20 Stepan RYBAR wrote:
ad 1) I did it, it looks like it works.
ad 2) ...you're using a JAXP TraX... You are right. Although I do not know,
that (and even how to) I changed anything about JAXP, when I run this
on my second comp (MS Windows Vista 64-bit Enterprise, Sun Java 64-bit
1.6.0_10-b33), it works fine.
There are multiple possibilities:
http://xml.apache.org/xalan-j/faq.html#faq-N100EF
Or you might have specified something like
-Djavax.xml.transform.TransformerFactory=net.sf.saxon.TransformerFactoryImpl
(or some other implementation class)
Thank You. Stepan
Původní zpráva
Od: Jeremias Maerki [EMAIL PROTECTED]
Předmět: Re: 0.95: Metadata is not valid XML document even if there are no
metadata defined in FO source
Datum: 05.11.2008 17:51:56
Ok, so we have two problems:
1. Compressed Metadata stream: FOP doesn't compress the metadata stream
by default. The problem is that FOP 0.95 has a demo configuration file
that contains a filter list that causes all stream to be compressed.
Please remove all filterList elements including their children from
the configuration file. I've already committed a fix to FOP Trunk which
makes this unnecessary some time ago:
http://svn.apache.org/viewvc?rev=695491view=rev
2. The XMP is not valid RDF because the namespace declarations are
missing: If that happens you're using a JAXP TraX implementation that
has a buggy serializer. I've just tried with a fresh Sun Java 1.6.0_10
installation on WinXP and didn't have that problem. I assume you've
installed a different JAXP implementation.
On 05.11.2008 16:28:54 Stepan RYBAR wrote:
Hello,
I have a problem with metadada part of PDF created by Apache FOP 0.95
(using Sun Java SE 1.6.0_10 on MS Windows 2000). Although resulted PDF
looks OK in Acrobat Reader, I am not able to process resulted PDF using
iText 2.1.3. It throws Exception shown below. After several e-mails
among iText developers (and me), it looks like even if I do not specify
any metadata in FO, FOP 0.95 produces PDF with compressed not valid XML
metadata section according to iText developers. If I specify metadata
according to Your example at
http://xmlgraphics.apache.org/fop/0.95/metadata.html#xmp-example
there is the same Exception. So, could You, please, verify, if the
problem is really caused by FOP and if yes, to repait it? Thank You
very much in advance.
Stepan Rybar
Pùvodní zpráva
Od: Paulo Soares
Pøedmìt: Re: [iText-questions] iText 2.1.3: Invalid byte 1 of 1-byte
UTF-8
sequence Exception while processing PDF from Apache FOP 0.95
Datum: 05.11.2008 15:44:00
The metadata can be compressed, the problem is that the metadata is not
a
valid
xml document (or at least it wasn't in the pdf I loooked at). iText
should
ignore the error and carry on but it currently doesn't. The problem
starts
at
FOP in any case.
Paulo
From: 1T3XT info []
Sent: Wednesday, November 05, 2008 2:08 PM
To: Post all your questions about iText here
Subject: Re: [iText-questions] iText 2.1.3: Invalid byte 1 of 1-byte
UTF-8
sequence Exception while processing PDF from Apache FOP 0.95
Stepan RYBAR wrote:
Hello,
thank You answer. But after some tests I guess, that this is not
caused
by
missing metadata.
Think again.
Even if there are no metadata in source.fo
Look at your PDF document.
In the root dictionary, there's a reference to XMP data.
/Metadata 7 0 R
If we have a look at object 5, we see:
7 0 obj
/Type /Metadata
/Subtype /XML
/Length 12 0 R
/Filter /FlateDecode
stream
xoe?QN~jÂ0?}?ü++?×?I??3¤.a){q??÷\??oem?#?Ú?öIu^...??
W+OE=?îÍ9'ç#?u.Ë=X´...?]Ä?óý?#bü?Íé1/4ÁNß7?VÍA~Z6{9Qxô{ý¨y}¨?`s?
?u`2ÍÂ~ÂQ?]Y]?¨Ýómy^(2)1?-^^(1)}'f?0G?'l8?¤s?X?T'
^(1)s?wEUR^(2)Ç?o`?D++E`ë?Ë(c)HÇü/yÜ/[EMAIL
PROTECTED]'^(1)?ôvA`?EURFk6âa~?ÎFWqJ?é0yé§$W^?B+x
a~?3/4äî?ÁO~ëE`U`s??!gÍ??.}??...kU^?Z?
Oops... What's this?
This is a compressed XMP stream.
That's not allowed in PDF!
# Pùvodní zpráva
# Od: Stepan RYBAR
# Pøedmìt: Re: [iText-questions] iText 2.1.3: Invalid byte 1 of 1-byte
UTF-8
# sequence Exception while processing PDF from Apache FOP 0.95
# Datum: 05.11.2008 13:52:34
#
# Hello,
#
# thank You answer. But after some tests I guess, that this is not caused
by
# missing metadata. Even if there are no metadata in source.fo, I can
use
iText
# to print to stdout total number of pages. So it means, that
