Re: Format of DOS 3.3 Random-Access-Files?

[rubywand@aol.com (RUBYWAND)]

In article <anders.137.000E8CE9@cipserv1.physik.uni-ulm.de>,
anders@cipserv1.physik.uni-ulm.de (STEFAN ANDERS) writes:

>
>we have transferred several random-access text files from an Apple ][
with
>DOS 
>3.3 to a PC. These contain many records. Is it possible to distinguish
the 
>several records also on the PC? Without the record-number information or 
>somthing compatible we cannot use the transferred data :(.
>
>

     DOS 3.3 does not automatically include record Length or Number of
Records information when creating a random-access Text file. In fact, a
standard sequential Text file can be Read-from or Written-to using L and R
parms to set a starting point in the file. DOS 3.3 would not complain. It
has no way of knowing whether a Text file has information organized into
records or not.

     This explains why it is usually a good idea to maintain random-access
Length and Number or records info in Record 0, where it can be read via
simple Input commands without worrying about record Length. An alternative
is to include record Length in the file name (e.g. MAIL.LIST.90)  or to
maintain a separate file which remembers at least record Length for files
on a diskette.

     If your files do not include record Length information in Rec 0 or
this info is not saved elsewhere, you can figure out record Length by
looking at file contents on the PC. Often, you will be able to detect
groupings of information and discern where one record ends and another
begins. XTREE is handy for viewing file contents. A text editor or word
processor could also be used.


Rubywand