jestermax Member Posts: 1064 From: Ontario, Canada Registered: 06-21-2006 |
Has anyone ever used the Deflator/Inflator ZLIB classes in the java API? I'm thinking of playing with it. Just curious if anyone else has done anything with it..... man, we need more Java people here  ------------------ |
|
jestermax Member Posts: 1064 From: Ontario, Canada Registered: 06-21-2006 |
Update: I've gotten ZIP format compression working with java, so that's kind of neat but my search for Java ZLIB resources continues... ------------------ |
samw3 Member Posts: 542 From: Toccoa, GA, USA Registered: 08-15-2006 |
Have you seen jzlib? http://www.jcraft.com/jzlib/index.html Its a pure java implement of zlib. The download has a few examples of how to deflate/inflate streams. ------------------ |
|
jestermax Member Posts: 1064 From: Ontario, Canada Registered: 06-21-2006 |
i have indeed looked at it but i want to use the native api, since it has classes for it anyways. I've almost figured it out by looking at the api documentation though.  btw, did you mean that they gave examples using their library or the api? ------------------ |
|
jestermax Member Posts: 1064 From: Ontario, Canada Registered: 06-21-2006 |
So i figured out how the ZLIB classes of the java api work somewhat; the only example code i found was in the actual API javadocs. So i'm going to take it and convert it to input zip files and load objects (such as BufferedImages) from memory. The issue i have now is that in the example they give where they take a 14 character-long string and compress and decompress it; in the data buffer that holds the compressed data, they set it to a 100 byte long array.... is that actually meaningful? Is there a way of calculating how much space compressed data will take? or should i just set the buffer to the size of the original data and hope for the best? ------------------ |
|
samw3 Member Posts: 542 From: Toccoa, GA, USA Registered: 08-15-2006 |
I'd go for 120% of the original size. Worse case would be an uncompressable file which would have the compression dictionary at the head. The extra 20% in the buffer should cover that. Its just a ballpark guess not an exact amount. ------------------ |
|
jestermax Member Posts: 1064 From: Ontario, Canada Registered: 06-21-2006 |
So i got the ZLIB stuff working, and the Zip archive stuff too. My plan is to create a manager that unzips an archive into memory and gives you access to the files. Basically a resource manager similar to Ogre or Irrlicht. Yes i know i should be reading the source code for it to get an idea, but....well i have no excuse. Anyways, I was telling my dad about my plans for this. This is the same guy that told me he wanted me to write his spreadsheets in Java.... yeah. So he out of nowhere, asks me "does the computer have enough memory for all those files?". It made me think... is it actually useful to load large amounts of files into memory? Does anyone know how Ogre/Irrlicht does this? Ogre's resource manager searches a directory for stuff like images to use as textures and scripts for stuff such as materials. Does it just keep track of the filenames? or does it precache everything so it's fast to use?
[This message has been edited by jestermax (edited March 25, 2007).] |