On Dec 11, 2003, at 10:52 AM, Eric Walker wrote:


Well, the problem is that this is just one section
of a file the other sections actualy have different format.
[..]
every know and then you get something
like list( "one" "two").
[..]
The data i am making up as I didn't think it
was important.  I figured the form is what was needed.
[..]
Once I get the
key value pairs it doesn't matter because I am going to read another
file like this and compare the data to make sure nothing has changed
between the two files.
[..]

p0: the 'data' that you made up, was good enough to
get us some direction to start looking at solutions.

p1: You might ultimately want to look at a re-write
of how those 'files' are being created, so that the
sub-sequent post parsing is simpler. IF that is
not an option - then you might want to see if the
application that created the format has a library
that can be 'ported' and hence build out a perl module
of your own that would share it's parser structure.

        p1.1 - either go with say an xml-ish model
                hence use the LibXml module for parsing

        p1.2 - assume that the applications library is
                lib_fung_parser.so - then you want to get
                your hands on the header files and use h2xs to
                generate your Fung::Parser module

p2: The problem of multiple section readings from a
file should lead you towards a simpler process where
you have keys in the file that indicate which 'parser'
to be using at which point you might want to look at
a 'dispatcher'... where each 'function' knows how\
to return stuff related to what it knows how to read,
and specifically how it's 'end of section' will tell it
to stop parsing and go back.



ciao
drieux

---


-- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] <http://learn.perl.org/> <http://learn.perl.org/first-response>




Reply via email to