ePub & PDF decryption on Mac OSX

[Jellby]

Originally Posted by epstewart

I have to assume, then, that ePUB books are typically zipped??? I didn't know that. They don't have to be unzipped to be read, seemingly, but if you want to edit them, then they need to be unzipped ... is that right?

Yes, because ePUB reading software already "knows" the ePUB is zipped and can work with that, but text-editing software doesn't.

(Why would anyone want to edit an ePUB, anyway? In case one doesn't like the ending Dan Brown gave his latest novel?)

No, to change "Brown, Dan" into "Dan Brown" in the author metadata, for instance, or to remove that ugly unreadable font included in the book, or to fix the clumsy absolute font sizes it uses, or delete the blurbs or advertisements at the beginning of the book... If the book is well done, you usually won't want to change anything, but face it, books are often not well done.

[adullday]

Originally Posted by Jellby

No, to change "Brown, Dan" into "Dan Brown" in the author metadata, for instance, or to remove that ugly unreadable font included in the book, or to fix the clumsy absolute font sizes it uses, or delete the blurbs or advertisements at the beginning of the book... If the book is well done, you usually won't want to change anything, but face it, books are often not well done.

... or to fix the R.A. Salvatore novels that have roosters on the cover (no, I'm not kidding), or to fix the Spanish textbook where tables are entered as images that are corrupted when read by Adobe DE, and of course the occasional OCR errors that turn words like "lie" into "he" and totally confuse the reader.

Honestly... don't they have any proofreaders??

[wallcraft]

Originally Posted by epstewart

does anyone understand the need to unzip/zip?

There is some incompatibility between the zip software used by ineptepub.pyw and the ZIPed ePub file. The Mac's built-in unzip works (is compatible), and pdurrant's script then produces a valid ePub that ineptepub.pyw can process.

[pdurrant]

The file format for ePub books is that of a zip archive. An epub file can be decompressed with any unzip program. Zipping it up again requires a bit more care, as the order and compression of the files is important.

Originally Posted by epstewart

I have to assume, then, that ePUB books are typically zipped??? I didn't know that.

[adullday]

I'm pretty sure that both the command line unzip and the python scripts use the zlib library, but the Leopard gui unzip does not.

Cygwin, which also uses zlib, also fails.

Windows Vista built-in fails, but 7-Zip succeeds.

Windows XP (on the one file I'm testing) asks for passwords for several files in the archive.

Linux unzip fails, which is no surprise since it also uses zlib, but the gui unzip succeeds (I'm using gnome... I think it's the Nautilus archive manager doing the work. I don't spend much time in the Linux gui. )

I'm guessing that there's either a bug in the zlib library, or more likely (especially since this is a problem on the Windows boxes too), the archiving software used by either the publisher or Adobe or whoever is responsible for the post-encryption zipping.

Has this ever been a problem with epubs downloaded from Window ADE, or is this particular on the Macs? And if so, why would we be getting different files?

[epstewart]

Originally Posted by adullday

Has this ever been a problem with epubs downloaded from Window ADE, or is this particular on the Macs? And if so, why would we be getting different files?

If you will provide me with a book title and source (either here in this thread, or in a PM) I will try to download it in Windows and also on a Mac. I have a Mac app called DiffMerge that can compare two files and show their differences, if any. It should be able to confirm whether the two downloads are the same.

BTW, I used DiffMerge to compare the adeptkey.der file created by ineptkey.pyw in Windows with that created by the Unix commands on a Mac (and by my AppleScript which embeds those commands). I found that the two files were identical. That should dispel any lingering doubts as to whether the need to unzip and rezip decrypted epubs on a Mac is due to any difference in the Mac adeptkey.der file.

Cheers,
Eric

[pthwaite]

I suspect its the way the files are unzipped. I have 2 unzip programmes, Zipeg and Stuffit. When some packages are downloaded, although one programme may appear to unzip it, it gives an error when trying to run it, whereas the other will unzip it and it will work.
So, it sounds like it is a similar problem here.
Howard

[mlc1]

Hi experts
I have an iliad reader and mac and bought an ebook in epubformat with DRM key. the problem is iliad cant read the book due to epub drm. On a windows computer we tried to decrypt the file but it gave the error

File name in directory "OEBPS/robe_978_oeb" and header "OEBPS/xxx.jpg" differ.

I read in this forum you succeeded to solve this error.
I am not familiar with OperationSystem of Mac I am only Mac user

Question; is there anybody who wants to decrypt my ebook (dutch language)

Thanks in advance

Best regards Ron

[epstewart]

Originally Posted by mlc1

Hi experts
I have an iliad reader and mac and bought an ebook in epubformat with DRM key. the problem is iliad cant read the book due to epub drm. On a windows computer we tried to decrypt the file but it gave the error

File name in directory "OEBPS/robe_978_oeb" and header "OEBPS/xxx.jpg" differ.

I read in this forum you succeeded to solve this error.
I am not familiar with OperationSystem of Mac I am only Mac user

Question; is there anybody who wants to decrypt my ebook (dutch language)

Thanks in advance

Best regards Ron

The problem has to do with the fact that some downloaded Adobe eBook files need to be unzipped and then rezipped again before decrypting them on a Mac. eBooks in the ePub format (even if they are ADEPT-encrypted) are basically ZIP files. They've been zipped into an archive containing a signle folder that holds multiple individual files. The entire archive has been compressed to make it smaller. For some reason, the ineptepub.py decryption software on a Mac can't do the unzipping properly, preparatory to doing the decryption. But if you unzip the archive and then rezip it, you can feed the rezipped version to ineptepub.py and it will work.

The key thing to do seems to be to follow the directions in post #60.

You will work on a copy of your eBook, not the original. I suggest putting the copy on the Desktop. It has to be renamed in Get Info, in the Finder, so that it has a .zip extension replacing .epub. Then, when you double-click the renamed .zip file, the Mac will unzip it into a folder on the Desktop.

Post #60 refers to this post in another thread, in which you will find a downloadable AppleScript "droplet" that you will download and unzip to become an ePub Zip 1.0.2 icon on your Desktop. Then you will drag the unzipped eBook archive, now a folder on your Desktop, to this droplet in order to rezip it. It becomes a file on the Desktop. This file has an .epub extension. You can use this file as the Input File to ineptepub.py.

[pthwaite]

Did it give an output file?

If so, change the file extension from .epub to .zip
Expand (unzip) the zip file and see if you can read the files inside.
If you can, it suggests it's the problem may be in the final compression stages and pdurrant's epub-zip applescript may help.

If no file is created or you cannot read the data in the zip file, it suggests you haven't fully followed the decrypt procedures.
Check you have the correct python and pycrypto environment, then check again you have run the scripts using the appropriate parameters. Python does seem fussy about " ' marks, you may have to do a bit of reading in the forum to see what works in windows as you are doing it in windows.

[pthwaite]

epstewart, you posted just before me, like minds posting at the same time eh

[epstewart]

Originally Posted by pthwaite

epstewart, you posted just before me, like minds posting at the same time eh

Indeed so! Cheers!

I would like to ask one of the participants in this thread to (either here or in a PM) give me a title and source for one of the books that needs unzip/rezip. I'd like to try to see whether it works as is on Windows but fails on the Mac. I suspect it does, but it never hurts to conduct a controlled experiment.

[epstewart]

I obtained An Echo in the Bone by Diana Gabaldon from ShortCovers.com and ran an experiment. I found that ineptepub.py/ineptepub.pyw got an error unless I unzipped and rezipped it as per earlier posts in this thread. It happened on a Mac using Python 2.5, on a Mac using Python 2.6, and also in Windows XP!

It looks as if the problem is platform-independent. From earlier posts in this thread I gather that the source of the problem may well be a bug or incompatibility in the 'zlib' library that the Python script imports. If so, the library has the same bug/incompatibility on different platforms.

Cheers,
Eric

[JSWolf]

I did remove the DRM from An Echo in the Bone. But it did have a problem. It had two invalid directories in the ePub. So what I have to do was unzip all the files and then put it back together so I could then strip the DRM.

[epstewart]

Originally Posted by JSWolf

I did remove the DRM from An Echo in the Bone. But it did have a problem. It had two invalid directories in the ePub. So what I have to do was unzip all the files and then put it back together so I could then strip the DRM.

I have been in touch with someone who knows more than I do about the way an ePub file is supposed to be put together. I asked him to look at An Echo in the Bone. He said:

The file is indeed an invalid epub. Two of the entries in the central directory of the epub (zip) file have had a portion of their name field set to all '\0' bytes. This is probably a bug with the software that generated the epub file.

That is exactly what JSWolf found, so I think we can conclude that some Adobe Digital Editions ePubs are improperly constructed zip files. The result is that ineptepub generates an error with them unless the files are unzipped and rezipped. The Mac's Archive Utility unzips them without error, maybe since it simply doesn't check the incorrect filenames.