Andras, in Ext2Directory : addFile and addDirectory

Andras, in Ext2Directory : addFile and addDirectory are exactly the same !
I suppose it's normal. What do you think ?


No, they are not really the same.
Most of the stuff involved in adding a directory or a file is the same, so I have put the common stuff in addDirectoryRecord().
If you look at the source, you'll see that what addFile() and addDirectory() do is call addDirectoryRecord() with a different Ext2DirectoryRecord as parameter.

Sorry, what exactly should I think about what? Smiling


Sorry, I thought they was exactly the same

Of course you are right, I made a mistake : I haven't see the little difference.

But, in addDirectory, you have :
return new Ext2Entry(newINode, name, Ext2Constants.EXT2_FT_REG_FILE, fs)
Isn't it Ext2Constants.EXT2_FT_DIR instead of Ext2Constants.EXT2_FT_REG_FILE that you would like to use?


Sure, thanks for spotting it, I'll correct it right away.

Please, don't correct it !

As I am doing abstract classes, I corrected it on my PC (sorry but I commit now). So, there is no necessity for you to commit just for this.

If already committed, that's not a problem.

some changes in CVS

Hi Fabien,
Please be aware that I had to make some changes to classes in the ext2 package (Ext2Entry, Ext2Directory, Ext2File, etc). If you are still working on abstract versions of these classes, please do an update from CVS to merge with the changes.


I have just updated from CVS and got conflicts with Ext2Directory, Ext2Entry and Ext2FileSystem.

That's due to the huge quantity of things (in all fs) I have done : moving attribute/method to abstract class, renaming method, merge of methods ...

I have a big bug (just before the update) : a Stackoverflow error while ant is build the boot disk. This sounds like an infinite loop Sad