2

I have a ZipFile and I'd like to create an array with the offsets/sizes of each entry in it. This is so that a C++ layer (JNI) can read these subfiles directly without having to extract them.

I can get the file's compressed size using entry.getCompressedSize() but I don't see anything to find out the offset of the file within the zip.

Enumeration<? extends ZipEntry> zipEntries = zipFile.entries(); while (zipEntries.hasMoreElements()) { ZipEntry zip = (ZipEntry)zipEntries.nextElement(); //long offset = ? long fileSize = zip.getCompressedSize(); }

Any easy way to get this?

Improve this question
3
  • What problem are you really trying to solve? Overall this sounds like a weird technical approach... Else I have no idea how to do this. Commented Aug 12, 2011 at 22:52
  • I have a project that was built using Android NDK r4 which didn't allow direct access for the native code to the project assets. I worked around it by passing the offset/filesizes to the native code and overrode the file operations to load the values directly from the asset file using the starting/ending positions in the file. For reasons I can't get into (nda) I now need to read an additional .zip file and I don't have the AssetFileDescriptor to get this information. Commented Aug 12, 2011 at 23:16
  • If you cannot find a solution you could still uncompress with Java and serve the file content from temp or (if you dare not store the data anywhere) memory. Commented Aug 12, 2011 at 23:23

1 Answer 1

Reset to default
2

I read that entries are returned by .entries() in the exact order they occur in the file (unsure if this refers to the actual file or the header entry.) Assuming this to be the case the following solves the problem:

Enumeration<? extends ZipEntry> zipEntries = zipFile.entries(); long offset = 0; while (zipEntries.hasMoreElements()) { ZipEntry entry = (ZipEntry)zipEntries.nextElement(); long fileSize = 0; long extra = entry.getExtra() == null ? 0 : entry.getExtra().length; offset += 30 + entry.getName().length() + extra; if(!entry.isDirectory()) { fileSize = entry.getCompressedSize(); // Do stuff here with fileSize & offset } offset += fileSize; }

This is working for my case. The only thing to note though is that according to the ZIP compression spec, it's not a given that in all zip files the files be stored in exactly the same order as in the directory. Since in my case I have control of the construction of the zip file this isn't a problem but this approach may not work if you work with zips from an uncontrolled source.

Improve this answer
Sign up to request clarification or add additional context in comments.

1 Comment

To elaborate, the offset is comprised of 30 bytes of standard zip header info + the file name length + some possible extra OS-specific flags + the data itself. To note, I ran into a bug where Java was incorrectly returning .getExtra() with a byte array not actually in the file. Since I had control over my zipping process I knew the correct value for it to be zero, so I removed the term. Something worth being aware of should you need to do something similar.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.