[Perlfect-search] Indexing Problem (DB_File)

[Christoph Reithmair]

Hello everybody,

I've installed Perlfect Search on a Linux-Server (Version 3.20) 
and it works great.

But I have a problem with indexing my page, because the
hoster only allows cgi-processes for short 10 seconds.
I prefer indexing by HTTP!

Now I've tried to index the page local on Win2k and on Linux.
But after upload, the search.pl creates an error.
It seemed, that the DB-File is not the same on each platform.

Now I tried to convert the DB-Files in ASCII-Files local by
a little Perl-Script.
Then i want to upload them and convert them back into DB-Files.

This worked almost for all files. But the file "inv_index" 
in Ascii looks like this ("#" is the field seperator):
1528"#"Q �& �h �+ �
1529"#"Q c& ch c+ c
153"#" �
1530"#"Q c& ch c+ c

These are some lines of this file. It was not possible to convert them
back to a DB-File.

Has anybody an idea how i can make an index local (Win) and then
upload it to an server with another operating system.

Is it possible to index Word-Documents too. I think there is a
command under linux: strings
It extracts the words from a Word-Document.

Thanks for your help
Christoph Reithmair

______________________________________________________________________________
Die Fachpresse ist sich einig: WEB.DE 20mal Testsieger! Kostenlos E-Mail, 
Fax, SMS, Verschl�sselung, POP3, WAP....testen Sie uns! http://freemail.web.de