gawk manual missing index entry for "do" statement, etc.

[Dan Jacobson]

gawk manual missing index entry for "do" statement.  It has got it own
personal Info node, but no personal index entry.
Also please check if the do keyword gets the proper color treatment in
awk mode in all circumstances.

By the way there is a table in the manual
        +----------------------------------------------
        |       STRING          NUMERIC         STRNUM
     --------+----------------------------------------------
        |
     STRING     |       string          string          string
        |
     NUMERIC    |       string          numeric         numeric
        |
     STRNUM     |       string          numeric         numeric
     --------+----------------------------------------------
that looks weird on my terminal, probably due to tabs.  How about you
don't use tabs so that it won't get messed up in various situations.
Also check the rest of the manual for tabs in tables... might as well
check all of Info.


By the way, in the manual:
Dark Corners
------------
   Until the POSIX standard (and `The Gawk Manual'), many features of
`awk' were either poorly documented, or not documented at all.
Descriptions of such features (often called "dark corners") are noted
in this Info file with "(d.c.)".  They also appear in the index under
the heading "dark corner."

OK, but how about putting a Info-follow-reference link or something so
we can get there in a hurry.  And mmmm, I think little "(d.c.)"'s
scattered about the manual will only lead to heads full of question
marks from international audiences who access the manual
randomly... how about you say "(this is a Dark Corner of [g]awk)" or
something, and make it a Info-follow-reference to the main dark
corners chapter.

By the way, in the manual
     $ awk 'BEGIN { print "line one\nline two\nline three" }'
     -| line one
     -| line two
     -| line three
I know you are well intentioned, but wouldn't
     $ awk 'BEGIN { print "line one\nline two\nline three" }'
     line one
     line two
     line three
be a little more, um... OK, each has their merits, never mind.

By the [extra] way, I just spent an hour wondering over this, and now
know the logic, but I think your manual section discussing 'print'
needs an example, like

x=235454.387195; y=x*10
print x, x+1, x+1.0, x+1.1, 987654.321, y
giving
235454 235455 235455 235455 987654 2.35454e+06
to drive home the point about how the default state is not necessarily
what the user expects.

This example could then point to the "Controlling Numeric Output with
`print'" section for more information.

Many users probably think after looking at other parts of awk, that
awk surely will accommodate to their data, and therefore wouldn't
expect it coming out funny, so they need my example to remind them
that in this area, awk doesn't "go with the flow", and they need to
adjust OFMT to get out what they put in.  I don't the manual has such
an example yet in this area.

By the way, I didn't check, but the manual probably needs an
print/precedence example of why
print "w=" w=2
fails, while
print  w=2 "=w"
is OK.  Therefore use
print "w=" (w=2)
print "w=", w=2
etc.
-- 
http://www.geocities.com/jidanni Tel886-4-25854780 e-mail:restore .com.