Yes, this is totally lame.
All the platforms on the planet offer
random access to a harddrive.
I noticed this the other day when
I was trying to build a very large file
and I couldn't just jump to the end
and write some stuff.
I didn't know that skip works on open/direct,
since I only tried it with open/direct/binary.
Lots of people wanting to do anything at all serious
with files would use this kind of access.
I haven't got the slightest idea why this bug
hasn't been reported yet.
Tim, I suggest you send a message to [EMAIL PROTECTED]
Apparently they haven't been watching the list here.
I would have reported it myself if I had known that
open/direct actually does work with skip.
The suggestion by Larry to use copy/part to skip the
file ahead is not reasonable for any kind of large file.
Plus, Rebol had better be able to skip backwards
on a binary file or else it is brain-dead for file
handling. It's not like were asking for built-in
nuclear cooling tower handling or something.
This is bread and butter. Binary is more basic
than line-oriented handling, unix experiences to
the contrary not withstanding :(-
-Galt
>===== Original Message From [EMAIL PROTECTED] =====
>Hi Tim
>
>Seems that skip does not work with open/direct/binary although it does work
>with open/direct. This may be a bug.
>
>In the meantime, you can advance the file pointer with copy/part:
>
>>> fp: open/direct/binary %test-char.r
>>> copy/part fp 10 ; returned value discarded when run in a script
>== #{6162636465666768696A}
>>> print to-string buffer: copy/part fp 10
>klmnopqrst
>>> close fp
>>>
>
>----- Original Message -----
>From: <[EMAIL PROTECTED]>
>To: <[EMAIL PROTECTED]>
>Sent: Sunday, July 16, 2000 8:56 AM
>Subject: [REBOL] Reading Binary data Re:(2)
>
>
>> Hi Jim:
>> That part worked thanks! However, if have not gotten
>> the results that I had hoped for.
>> I had assumed that skip fp 10 would take me to offset
>> 10. It appears that it has not.
>>
>> TestBin.txt contains the characters:
>> "abcdefghijklmnopqrstuvwxyz"
>>
>> The following code :
>> ;===============================================================
>> fp: open/read/direct/binary %TestBin.txt
>> skip fp 10
>> print to-string buffer: copy/part fp 10
>> close fp
>> ;===============================================================
>> produces:
>>
>> >> do %BinFile.r
>> abcdefghij
>>
>> The result I would be looking for in buffer
>> would be "klmnopqrst"
>>
>> So it appears that skip fp 10 is not advancing
>> the "file pointer" to offset 10.
>>
>> what would I do to make that happen?
>> Thanks
>> Tim
>-------snip-----------
