temisu / whdloadarchivefs Goto Github PK

Hi temisu,
I would like to use this issue for related discussions.

For archivefs_dirCache() it is important, that the order in which entries are provided makes sure that for files in a subdirectory the directory entry above is provided before. This is required to build the structure in WHDLoad.
Is this the case?
I wonder also if archives are possible which do not contain the subdirectory above. Can you tell me?

The functions getFileSize and fileRead must handle filenames not case sensitive as the Amiga filesystems do.
As WHDLoad currently only handles case of US-ASCII this would also be enough for WHDLoadArchiveFS.

Hi, may specs were unclear...
In the function registerFunc called by archivefs_dirCache I like to get the full path including the name of the actual object. Currently it seems to only return the path without the object. I know the object name is in the fib but having it in the path too makes things simpler.
Could you change it that way?

To let WHDLoad more precise react on errors from WHDLoadArchiveFS it would be needed to define some error codes.
I would like to distinguish on getFileSize from simply that file isn't in the archive from any other error.
For that I would propose to define an VFS_ERROR_OBJECT_NOT_FOUND=-2 in whdvfs.i and an starting point of individual errors defined by WHDLoadArchiveFS privately itself, e.g. VFS_ERROR_CUSTOM=-100.
What do you think?

Hi again,
can you please change archivefs_fileCache/archivefs_dirCache to return an error code if the the callback function fileFunc/registerFunc has returned 0 for stop (these only return 0 on an error condition).
At the moment it is not really needed to provide a error string for this (but maybe better for completeness)