From: fbkotler@nospicedham.myfairpoint.net   
      
   Kevin Mardoux wrote:   
   >   I try this it is OK   
   >   
   >  may be somebody have a better solution ?   
      
   Hi Kevin,   
      
   I'm sorry your post was delayed again. You changed your address   
   slightly, and that was enough to get you off the "whitelist" - you're   
   back on it now.   
      
   I'm still not entirely clear on what you're doing. Simply writing A 1 B   
   2 C 3... is simple, but probably not what you want.   
      
   > ; the file R1 containts bytes value A,a,B,b,C,c,D,d....   
      
   If your code works as shown, this file must not have the commas in it,   
   It must be "AaBbCc...", right? Does it go all the way to "Z"? Or don't   
   we know that?   
      
   > ; the file S2 contain 1,2,3,4,5,6...   it is my table   
      
   Similarly, this "must" not actually have the commas in it. Since we need   
   to write this on the command line, perhaps we sometimes have a different   
   "table" - 2468... or whatever? Are these "binary numbers" or "text   
   representing the numbers"? If text, what do you do about "10" and above?   
      
   > ; The aim to have il file R1 with  bytes A,1,B,2,C,3,D,4....   
      
   Okay. A more "usual" way would be to read your data file - "AaBbCcDd..."   
   into a buffer, and then intersperse the 1 2 3 or whatever into the same,   
   or a different buffer, and then write the whole buffer to an output file   
   with the same or different name. That might be "better" than fooling   
   around with "seek"ing around and writing it one byte at a time. Or it   
   might not be "better", depending on what the point is. It would almost   
   certainly be faster! Tell us how you're trying to "improve" it.   
      
   Best,   
   Frank   
      
      
   > ; i use A86 8-16 bytes compilator   
   > ; R2.COM is the name of this file   
   >   
   > mov si,0080H  ;  how many Char after ?   
   > lodsb   
   >  cmp al,0   
   >  jne suite   
   >  erreur:  mov dx,offset usage   
   >   mov ah,9   
   >   int 21h  ; indicate the syntax : to run type " R2 S2.DAT "   
   >   ret    ; this file will open automatiquely the file R1.DAT   
   >   
   > suite:   cbw   
   > xchg ax,bx   
   > mov byte [bx+si],00    ; 0 Terminate   
   > inc si ; mov si,0082H   
   >   
   > mov ax,3d12H   
   > mov dx,0082H         ; adresse name S2.DAT  i MUST tape it   
   > int 21H   
   > jb err82     ; can i open it ?   jnc   
   > ; the user MUST type de name of the file with data 1,2,3,4...   
   >   
   > xchg ax,bx   ; Handle   
   > mov ah,3fH   
   > mov cx,0FFFFh  ;             ; size max  buffer   
   > mov dx,offset table          ; position  file  of S2.DAT   
   > int 21H    ; cmp ax,0000  ; JBE fermer   
   > ; i suppose more than 0 byte to read   
   > ; i suppose that datas of S2.DAT are on address Table   
   >   
   > mov word VS2,ax ;  just for information Verify S2   
   > mov word ntest,ax ;  True number of bytes to read in the TABLE  S2.DAT   
   > ; ntest is usefull : will count number of cycles   
   > push bx ; i qave the handle of file S2.DAT   
   >   
   > mov ax,3d12H        ; i will write in this file   
   > mov dx,offset S1R1              ; file R1.DAT   
   > int 21H   
   > jb err1      ; i verify : i have NO error : counter stay at 0   
   >   
   > xchg ax,bx  ; handle of R1.DAT in BX   
   > mov ah,3fH   ; read  file R1   
   > mov cx,0FFFFh              ; test a large buffer   
   > mov dx,offset table+30000   ; i suppose file S2 is less than 30k   
   > int 21H        ;             ; S2.DAT begin at offset TABLE   
   > jb err2        ; same no error   
   >   
   > dec byte testok  ; if no error before, i comme here   
   >   
   > jmp short continue   
   >                         ; this part count errors   
   > err82: inc byte cerr82   
   > jmp short continue   
   > err1:  inc byte cerr1   
   > jmp short continue   
   > err2:  inc byte cerr2   
   > continue:   
   >   
   > tesbou:  mov ax,4200H         ; position  file   
   > inc word verify         ; i come here    number   ?   
   > ; word verify just a test  > the counter indicate a good answer   
   >   
   > xor cx,cx             ; begin byte 0   
   > confil:  mov dx,1   
   >             ;  i suppose i read the byte 1 of the file R1.DAT   
   >   
   > ; the value of confil is   BA 01 01  next BA 03 01    next BA 0501   
   > ; if i do  mov dx,offset table+30000 ; the size of the file grow   
   >   
   > int 21H   
   > ; i use a counter : BA 0101  BA 0301  BA 0501 "BA" is "mov dx"   
   > mov ax, word confil+1   ; next step read the byte 1 +2 = 3   
   > add ax,2   
   > mov word confil+1,ax ;  I prepare for the next test   
   >   
   > mov ah,40H   
   > mov cx,1                     ; write 1 byte   
   >   
   >   conta: mov dx,offset table   ; + 1 by test   
   > int 21H   
   >   mov ax, word conta+1   ; i suppose i read the 1st byte   
   >   add ax,1               ; then i read the 2nd   
   >   mov word conta+1,ax    ;  then the 3rd   
   >                         ; i use it like a counter   
   >   
   > mov ax,word ntest   ; i count the number of bytes in file S2.DAT   
   > dec ax   
   > zontest: mov word ntest,ax   
   > cmp ax,0          ; byte to read ?  yes do again   
   > jnz tesbou        ; no byte to read ; go away   
   >   
   > mov dx,82H   
   > fermer: mov ah,3eH   ; i close  S2.DAT   
   > int 21H   
   > pop bx   
   > mov dx,offset S1R1   ; AH stay at value 3E   
   > int 21H   
   >   
   > ; part below only to register informations   
   >   
   > jmp short openfil  ; this part to verify   
   >   
   > crefil:           ; i register in a file errors a number of byte read   
   > mov ah,3ch   
   > mov dx,offset fildat   
   > xor cx,cx              ; attribute   
   > int 21h    ; if FIL.DAT do not exist, i have an error, but it don't care   
   >           ;  the i have FIL.DAT size 0 ; second try fileexist, it is ok   
   >   
   > openfil:   
   > mov ax,3d12h   
   > mov dx,offset fildat   
   > int 21h   
   > jb crefil   
   > xchg ax,bx  ; save the handle   
   >   
   > mov ax,4202h   ;  handle in BX   
   > xor cx,cx         ; begin byte 0   
   > xor dx,dx  ; no offset   
   > int 21h   
   > jc erreur1   
   >   
   > mov dx,ax   
   > mov ax,4200h  ;   
   > xor cx,cx   
   > int 21h   
   >   
   > mov ah,40h     ; write file  ;  BX keep same value   
   > mov cx,16     ; write 16 bytes   1 line is easy to read   
   > mov dx,offset verify   
   > int 21h   
   > jb erreur1   
   >   
   > mov ah,3eh     ; close   ; mov bx,[offset] handle   
   > int 21h   
   > mov ah,09   
   > mov dx,offset txt   
   > int 21h   
   > int 20h   
   >   
   > txt db:   
   > verify dw: ,0  ; count number of test   "tesbou"   ; 2 bytes   
   > VS2  dw:   ,0  ; number of byte read in S2.DAT   
   > db:  ,9        ; separator  again 1 byte   
   >   
   > cerr82 db: ,0   ;  1 B    is the inkey correct ?   
   > cerr1 db:  ,0   ;  1 b    about main file R1.DAT   
   > cerr2 db:  ,0   ;  1 b   
   > testok db: ,0   ; if ok i read value FF   
   > db:        ,9   
   > fildat db: "FIL.DAT",0,36   ; used for TXT error and History File   
   > erreur1:   
   > mov dx,offset fildat   
   > mov ah,09   
   > int 21h   
   > int 20H   
   >   
   > S1R1 db: "R1.DAT" ,0   
   > ntest dw: ,0   
   > usage db: "R2 S2.DAT"  ,24H   
   > table db:   
   > ret   
   >   
   > Best regards   
   > Kevin   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)