Hi Michael -
At 12:42 PM 4/11/00 -0700, you wrote:
>I'm currently defining a 'print element of an object as a function. I'm
>assuming that the REBOL definition of 'print will be hidden in the
>definition of this object (and definitions of functions local to the >object)
Actually .....
print would write to a port, which could be standard
output OR a physical file. The port would be defined
as a member of the object as well.
let's call this content-object ; will build virtual
; pages OR could be used to build a physical web page
; let me know what you think of the following:
; thanks for your interest :>)
; tim
content-object: make object!
[
write_to_file: func []
[
; is this being run from a server?
either equal? system/options/cgi/server-name none
[return true]
[{else}return false]
]
init_output: func[fname[string!] /local fpl]
[
either write_to_file ; no server so open a write port
; to fname
[
file_name: make file! fname
fp1: open/new/write file_name
]
; yup, we're on the server so keep writing to stdout
[{else} fp: system/ports/output]
return fpl
]
; what the hay!! We can call this anything, but
; print would be great if it didn't screw up implicit
; output to stdout for the "original print"
fp: init_output "hello.htm"
print: func [fp[port!] value]
[append fp value]
]
; If this process works with out conflict, then I would
; create a debug object which would write to a file that
; would be created every time the application runs
; and an errorlog object that would be appended every
; time the application ran, given an error or warning
; condition appeared. Standard parts of the my C/C++/CGI
; toolkit and expected by sysops that my cgi programs
; run on. They could all have a print element (method)
; OR I could call it something else.
>- which is ok. Outside of the element 'print, the REBOL definition works
>just fine. Following is the relevant piece of my object:
>
>player-def: make object! [
> name: none
> connection: none
> ; METHODS
> ; - - - - - - - - - - - - - - - - - - - - - -
> print: function [
> "Print a message to this player, with trailing new-line"
> msg [string! block!]
> ][new-msg][
> ; Cleanup the message suitable for telnet display
> new-msg: player-format-print msg
> append new-msg new-line
> append connection new-msg
> ]
> ; - - - - - - - - - - - - - - - - - - - - - -
> prin: function [
> "Print a message to this player"
> msg [string! block!]
> ][new-msg][
> ; Cleanup the message suitable for telnet display
> new-msg: player-format-print msg
> append connection new-msg
> ]
>]
>
>- Michael Jelinek
>
>-----Original Message-----
>From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]
>Sent: Tuesday, April 11, 2000 11:32 AM
>To: [EMAIL PROTECTED]
>Subject: [REBOL] [REBOL] Redefining functions with objects
>
>
>I'm thinking of developing a class.
>Let's call it
>
>tims-object
>
>Suppose I write a function for this
>class and I call it
>
>print
>
>Will
>tims-object/print
>redefine rebol's own print?
>
>I don't really want to do this, so
>I would welcome comments on this.
>
>thanks
>tim
>
>
>
