achowe@tiger.waterloo.edu (anthony howe) (09/18/89)
I finally tried to use the last version of << GNU AWK >> that came down the pike to do some global text search and delete (since FOLDED 0.9e lacks the ability to search and replace control codes). I quickly review what I needed in the book << The AWK Programming Language >> and tried the following: { gsub( /_\b/, "" ); print } This was meant to globally search and delete all undersorce-backspace combinations (I'm stripping a formatted file of cude underlining and other code that mess up my printer). However AWK keeps giving me an "error near line 1" and quotes just about the whole line. What's wrong? I don't even know the diffs between << GNU AWK >> and the implementation described in the above book. There was no documentation. All I remember was a comment or a small readme that said I should get a copy of my UNIX version's man page. Was there any docs with << GNU AWK >> and if so are not my rights to it being abused by not providing it (as mentioned in every GNU COPYING & LICENSE notice)? - Ant achowe@tiger.waterloo.edu | "Life is not fair. Anyone who tells _ -|-|_ _ | you different is trying to sell you (_\ |\| | | | (_) |\| \/ | something." - The Princess Bride ___/ | disclaimer...
close@cacilj.UUCP (Diane Barlow Close) (09/18/89)
In article <16459@watdragon.waterloo.edu> achowe@tiger.waterloo.edu (anthony howe) writes: >I finally tried to use the last version of << GNU AWK >> that came down > > { gsub( /_\b/, "" ); print } > >However AWK keeps giving me an "error near line 1" and quotes just >about the whole line. What's wrong? As the person who wrote the GNU awk documentation, perhaps I can help. GNU awk was developed to run on much larger machines than ST's and was ported to the ST by non-GNU persons. While this is perfectly acceptable (given the GNU copyleft info), it DOES NOT IMPLY complete compatibility with the Unix version of GNU awk. Several people in my local Atari club also tried to use GNU awk on the ST and ran into the problem you describe. They found that if they stripped out ALL the spaces on the command line, awk worked just fine. This is a changed made by the person(s) who ported GNU awk to the ST. Why they made such a change is beyond me; perhaps they don't even know the change is there? The documentation for GNU awk was developed separately from the code, and is distributed separately. The manual is very complete and has an excellent (if I don't say so myself :-) beginner's section. Some of the documentation will only apply to larger machines; ST user's will have to rewrite the manual to fit ST GNU awk's specifics (making it freely distributable, of course). I hope this helps. You can ftp GNU awk documentation from prep, as is usual with all GNU-made products. -- Diane Barlow Close {uunet, crash, ucsd}!cacilj!close close@cacilj.uucp Free Canada -- Trade Mulroney
mj@myrias.com (Michal Jaegermann) (09/21/89)
From article <1228@cacilj.UUCP>, by close@cacilj.UUCP (Diane Barlow Close): > I hope this helps. You can ftp GNU awk documentation from prep, as is > usual with all GNU-made products. > All awk users on ST should also be aware that mentioned above documentation describes a current version of gawk, which behaves like nawk (more or less), i.e new version of awk described in a book "The AWK Programming Language" by AWK - Aho, Weinberger, Kernighan. A version ported to ST is an older one, it apparently suffers from some bugs in an original code and follows "classic" awk, now partially obsolete. It looks like that one should get back to the drawing board... Michal Jaegermann Myrias Research Corporation Edmonton, Alberta, CANADA mj@myrias.COM ...{uunet,alberta}!myrias!mj