Hello Julia,
On 22-05-2012 18:58, Julia Medvedeva wrote:
My current task is a implementation of UDF 1.02. Level of restrictions
- 1. Read-only. I am doing my project step by step. It means that
after I have finished read some part of docs I try to implement it in
code. Udf has very «lovely docs» and sometimes I have to check it in
code to be sure of my opinion about format file system. It is why I
can't separate the reading docs and coding. My timeline is a little
different than Google's schedule.
As part of my proposal I have done reading volume recognition sequence
and anchor volume descriptor pointer for implementation of mount
function. It is my point of start.
My first big work it is a OSTA functions fo udf. Papers contain the
code but it is was far from HelenOS style. I have done it compliant
with my udf server in HelenOS code style. Also I added all descriptors
which needed for working with volume.
The second part it is the reading of volume descriptor sequence. It
could contain primary volume descriptors, logical volume descriptors
and partition descriptors. As each of these descriptors could be
rewrited, on this step is necessary to use prevailing rules for saving
only actual descriptors (according to ecma 167 3/8.4.3).
Now I have information about volumes, logical volumes and partitions.
So I can start search for root directory. Each logical volume
descriptor can direct to fileset descriptor which contain direction to
root directory. My next commit will be provide of reading root
directory. When I can do it I will be able to read file of names.
Great, thanks for the detailed heads up! Just don't forget the regular
mainline merges like I have already suggested. It should not be a major
problem for your branch as your work is mostly isolated to the UDF driver.
M.D.
_______________________________________________
HelenOS-devel mailing list
[email protected]
http://lists.modry.cz/cgi-bin/listinfo/helenos-devel
