Subj : Were i can get source cod
To   : Jasen Betts
From : mark lewis
Date : Mon Apr 01 2002 03:57 am

 JB> Hi mark.

 JB> 22-Mar-02 21:21:41, mark lewis wrote to John H. Guillory

 ml> witness the following (that i removed from a library 
 ml> because of my belief that we should pass mail as 
 ml> received and not modified [...]

 JB> so you don't claim ownership of this code, prolly a good thing :) 

hehehe, yeah, kinda...

 ml> Function CleanChar(InChar: Char): Char;
 ml>   Const
 ml>     CtlChars: String[32] = 'oooooooooXoollo><|!Pg*|^v><-^v';
 ml>     HiChars: String[64] =
 ml>  'CueaaaageeeiiiAAEaaooouuyOUcLYPfarounNao?--//!<>***|||||||||||||'
 ml>     HiChars2: String[64]=
 ml>  '|--|-+||||=+|=++-=--==-||||*****abcnEduto0nd80En=+><fj/~oo.vn2* '
 ml>   Begin
 ml>   Case InChar of
 ml>     #0..#31: CleanChar := CtlChars[Ord(InChar) + 1];
 ml>     #128..#191: CleanChar := HiChars[Ord(InChar) - 127];
 ml>     #192..#255: CleanChar := HiChars2[Ord(InChar) - 191];
 ml>     Else
 ml>       CleanChar := InChar;
 ml>     End;
 ml>   End;

 JB> what a mess!

to a point but it does work for the desired operation... that operation being 
to strip out certain characters and replace them with the ones in the constant 
strings/arrays...

 JB> most of the problem is using the wrong data type for the constants
 JB> - use array [char] of char

i agree that maybe using the string structure was not efficient but i don't 
recall there being any real problem for the job that the code does...

 JB> someone well respected said that chosing the right data structures
 JB> can make the programming task much easier.

 JB>  Function CleanChar(InChar: Char): Char;
 JB>    Const
 JB>      CtlChars: array[#0..#31] of char = '
 JB> oooooooooXoollo><|!Pg*|^v><-=^v';      HiChars: array[#128..#255]
 JB> of char =
 JB>       
 JB> 'CueaaaageeeiiiAAEaaooouuyOUcLYPfarounNao?--//!<>***|||||||||||||'+
 JB>       
 JB> '|--|-+||||=+|=++-=--==-||||*****abcnEduto0nd80En=+><fj/~oo.vn2* ';
 JB>    Begin
 JB>    Case InChar of
 JB>      #0  ..#31:  CleanChar := CtlChars[InChar];
 JB>      #128..#255: CleanChar := HiChars[InChar];
 JB>      Else        CleanChar := InChar;
 JB>      End;
 JB>    End;

 JB> Not that that'll be any help except as an example.

hey... it works and does speed up the operation... cycle counters would like it 
as there's no math functions to have to be done <<GG>>

believe it or not, i've a similar operation in some REXX code that is/was used 
for "converting" ISO-8859-1 to PC850... problem is that when one is gathering 
web pages with wget, cleaning them and posting them to echos, one can't tell 
what character set is used unless there is something actually in the page data 
that states... i'm about to the point of either just replacing that routine 
with a 'null' routine or just altering the existing replacement table to be the 
'normal' (standard?) US set...

 JB> P.S. I never saw any messages on this subject before yours.

it was a year or so back... i had locked the message in my message base and 
forgotten to unlock it... saw it the other day and figured that it was better 
late than never...

 JB> did the originator find the source code they wanted?

dunno... ;-(

)\/(ark

 
* Origin:  (1:3634/12)

.