epubcheck icon indicating copy to clipboard operation
epubcheck copied to clipboard

Support for ZIP64 archives?

Open sandersk opened this issue 11 years ago • 8 comments

Does epubcheck support ZIP64 archives (for EPUBs > 4 GB in size)? I ran a test on a ~14 GB EPUB compressed as what I believe is a valid ZIP64 archive, and got the following error:

ERROR: test.epub: I/O error: invalid CEN header (bad signature) 

Check finished with warnings or errors

If I'm reading the EPUB 3.0 spec correctly, ZIP64 archives are supported (http://www.idpf.org/epub/30/spec/epub30-ocf.html#sec-zip-container-zipreqs)

"OCF ZIP Containers may use the ZIP64 extensions defined as 'Version 1' in section V, subsection G of the application note at [ZIP APPNOTE] and should use only those extensions when the content requires them. OCF Processors must support the ZIP64 extensions defined as 'Version 1'."

I'm not an expert on the ZIP format, so it's quite possible I've just made a mistake here in how I compressed the EPUB. But figured I'd report in case further investigation was merited.

Thanks, Sanders

sandersk avatar Jan 30 '14 17:01 sandersk

From this stack overflow entry, it seems this is caused by Java (version 6 or earlier) not supporting ZIP 64bit.

Would you be able to try running EpubCheck with Java 7 ? Alternatively, could you provide a sample ZIP (ideally under 14GB :wink: ) ?

rdeltour avatar Jan 30 '14 22:01 rdeltour

Thanks very much for the quick response, @rdeltour. I installed Java 7 and then tried running epubcheck again against my large ZIP 64 file, and it succeeded as expected.

So this is indeed a Java 6 compatibility issue, not an epubcheck issue. I'll close the ticket.

Thanks again for your help!

sandersk avatar Feb 04 '14 01:02 sandersk

I am still getting same issue.

java -jar  /opt/apache-tomcat-8.0.0/webapps/epubcheck3-0b2/epubcheck-3.0b2.jar '/data/formats/output/bookstore/20160614053934669237/9781471521430.epub' >> /tmp/12_big_epub.txt
ERROR: /data/formats/output/bookstore/20160614053934669237/9781471521430.epub: I/O error: invalid CEN header (bad signature) 

Check finished with warnings or errors!

Java version:

[iv2user@develpoment-24 reader]$ java -version
openjdk version "1.8.0_91"
OpenJDK Runtime Environment (build 1.8.0_91-b14)
OpenJDK 64-Bit Server VM (build 25.91-b14, mixed mode)

File Size is 2.1 GB

[iv2user@develpoment-24 reader]$ du -h /data/formats/output/bookstore/20160614053934669237/9781471521430.epub
2.1G    /data/formats/output/bookstore/20160614053934669237/9781471521430.epub

vivekbsable avatar Jun 14 '16 11:06 vivekbsable

Hi, This is due to invalid package/zip file.

Regards,

Arulselvan R Pallavarayar Direct: +91 (0)44 4916 2259 | Ext: 259 Mobile: +91 (0)9710416093 | [email protected]

arulselvanrpallavarayar avatar Jun 14 '16 11:06 arulselvanrpallavarayar

No, epub contains video and audio files of 2.0 GB. When I remove video and audio file then it is working.

[iv2user@develpoment-24 epubcheck4.0.1]$ java -jar  /opt/apache-tomcat-8.0.0/webapps/epubcheck3-0b2/epubcheck-3.0b2.jar '/data/formats/output/bookstore/20160614053934669237/NOID_1.epub' >> /tmp/12_big_epub.txt 
ERROR: /data/formats/output/bookstore/20160614053934669237/NOID_1.epub/OPS/css/NOID.css: invalid operator before a function.

Check finished with warnings or errors!

vivekbsable avatar Jun 14 '16 11:06 vivekbsable

Zip type is ZIP 64+

which is doing something. because when I Copy EPub on Window OS and did Unzip and again ZIP, after that I checked with ePub Check then it its working.

More important, this time ZIP Type is Zip and (No 64+)

vivekbsable avatar Jun 17 '16 03:06 vivekbsable

No, epub contains video and audio files of 2.0 GB

This looks like an edge case.

Is this still an issue for anyone? I'm considering closing as wontfix , unless someone upvotes this issue or provide a repro.

rdeltour avatar Feb 26 '19 13:02 rdeltour

@vivekbsable Does your ZIP file correctly announce the use of ZIP 64? @rdeltour MS Office can correctly handle OOXML files containing huge video. Since somebody might want to embed huge video within an EPUB publication, I would like to know more about this issue.

murata2makoto avatar Feb 26 '19 13:02 murata2makoto