[COFF] TECO variants

[Mike Markowski via COFF]

On 9/22/25 10:28 PM, Joseph Holsten via COFF wrote:
> ...I’ve wondered how unusual extensions like Control-R were on other systems. Were different implementations as different as some authors make it sound? Or were they roughly as similar as the QED implementations?
>
> Also, do you buy the modern argument that TECO vs emacs proved the command language and the extension language should be entirely different?

The 1980 DEC TECO manual has a section on some differences, but they 
don't seem that big.  Their list is likely not all-inclusive, only 
talking about their versions of teco on various OSes.  I learned TECO at 
age 16 in 1978 on a PDP-11/70 RSTS/E system and thought I was 
all-powerful :-D, such an incredible program.  For fun I recently 
compiled Paul Cantrell's Video TECO on my ubuntu box (must add -lncurses 
to makefile).  It's aaalmost as nice as vi but of course not as fully 
visually oriented.  I won't be giving up on vim...

Many TECO commands can be preceded by an integer or followed by a 
string.  The output of a command can be used as the input to another, 
making TECO a quasi-functional language.  I don't have the background to 
comment on the pros and cons of the command language, but for something 
begun in 1962 it's impressive.

"It has been observed that a TECO command sequence more closely 
resembles transmission line noise than readable text."  -Ed Post

Here is a macro of Cantrell's that converts the word at current pointer 
position to lower case:

  .UU<0A-32"E0;'0A-9"E0;'0A-10"E0;'C>.UL$QUJQL-QU<0A-64"G0A-91"L0A+32IDR''C>$

What do you mean, it's unreadable? :-)  Curious what it was doing, I 
rewrote with comments:

.UU                 ! Put dot in Q-reg U.                 !
<                   ! Loop till finding white space.      !
     0A-32"E0;'      !   Ascii code at dot == space? break !
     0A-9"E0;'       !   Ascii code at dot == \t?    break !
     0A-10"E0;'      !   Ascii code at dot == \n?    break !
     C               !   Fwd 1 char.                       !
 >
.UL$                ! Put loc of 1st white space in L.    !

QUJ                 ! Return to original dot.             !
QL-QU<              ! Loop from U (orig dot) to L (1st white space). !
     0A-64"G         !   If ascii code at dot >= 'A':      !
         0A-91"L     !     If also <= 'Z':                 !
             0A+32I  !       Insert lower case letter.     !
             D       !       Delete upper case letter.     !
             R       !       Back up 1 char to prepare for 'C' below. !
         '
     '
     C               ! Fwd 1 char !
 >$

Mike Markowski