Text file processing utilities

Topic

New Reply

I wonder whether anyone here is an expert on GSAR or SED?

GSAR: I can’t get the @responsefile command to work, and the minimal documentation loses any mention of a “response file” between versions 1.20 and 1.21

SED: I can’t see how to add control characters (specifically CR and LF).

Let me describe the problem.

I have a text file which contains records of the form
[DoW, date, time] up to 40 characters of message text with a blank character at the end
real sample:
[Thu, 2008-06-12 13:37:53] – Log emailed [Thu, 2008-06-12 13:38:01] – Administrator logout – IP:192.168.11.2
BUT these are in one long string, without a carriage return/line feed pair after each training blank and before the [ character.

Effectively I wish to replace
” [Mon,” with “crlf[Mon,”
” [Tue,” with “crlf[Tue,”
” [Wed,” with “crlf[Wed,”
… down to
” [Sun,” with “crlf[Sun,”
where crlf represents the hex characters 0D0A, so that the text is split into lines starting each time with the [ character before the day-of-the-week abbreviation.

If I could get the response file to work, I would use GSAR, otherwise I would need to make seven passes, each processing a specific day-of-the-week.
I’ve looked at various SED tutorials, but can’t find how to do what I want.

Any ideas? Thanks

BATcher

Plethora means a lot to me.

Reply | Quote

Replies

From your description, couldn’t you just do a global replace of ” [” with a “[“? Or am I missing something? And does it have to be SED or GSAR?

Alan

Reply | Quote

Inadequate requirements definition – some records have a second [ in their text.

No, it doesn’t have to be GSAR or SED (I don’t know any others), but is does have to be command-line!

BATcher

Plethora means a lot to me.

Reply | Quote

(Edited by jonWallace on 26-Jun-08 23:48. expanded it a bit…)

Have you looked at awk http://www.vectorsite.net/tsawk.html%5B/url%5D ? I used this recently for a similar job at work. I’ll get more information at work tomorrow (today really, it’s late here…) and post back.

John

Reply | Quote

Thanks, John – I’ll look into this – but I always thought that “AWK” was just a shortened form of “awkward”! (Much like the IBM mainframe data selection utility “EasyTrieve” was known by us as “DiffiTrieve”…)

BATcher

Plethora means a lot to me.

Reply | Quote

You could also follow up the leads from the WikipediA entry for AWK. It’s not nearly as awkward as it may appear at first glance. And it has a power and elegance that should also appeal to a BATch file afficianado

Reply | Quote

OK. Get ahold of awk95.exe. I’m lazy so I put everything in the same directory so I don’t have to worry about filepaths.
Create a textfile (I called it new2.awk) that looks like

{gsub(/[/,”n[“);print > “new3.txt”}

I made a file with your specimen data repeated twice on one big line and changed the second set to [Fri…] and saved that as new1.txt.
Then from a commandline (in the same directory — see above) I typed

awk95 -f new2.awk new1.txt

This calls new2.awk which reads new1.txt and parses it and pipes the output to a file called new3.txt with one record per line.

Hope this is some help

Reply | Quote

I hate to sound nosey or pushy here, but did you find a solution to your problem?

Reply | Quote

Hello John

As quite often happens, more pressing matters made me put this matter on the back burner.

I spent about five minutes writing a BATch file to make seven passes of the same file using GSAR, once for each day-of-the week, rewriting it each time to the same filename. Spectacularly inefficient, but will do for the moment until I have time to learn AWK…

Now if I could afford the originally-mainframe utility SELCOPY I could have done the command coding in about the same five minutes!

BATcher

Plethora means a lot to me.

Reply | Quote

[indent]

---

As quite often happens, more pressing matters made me put this matter on the back burner.

---

[/indent]

Yeah, I know how that one goes! There are always more pressing matters. Still good to see you’ve found a solution.

Reply | Quote

Sometimes the “sledge hammer” is the most expedient, if not elegant solution! (Guilty of similar offences )

I’d be interested your opinion of AWK, once you’ve had time to have a proper play with it.

Reply | Quote

> As quite often happens, more pressing matters made me put this matter on the back burner.

Haven’t you sorted out your ironing regimen yet!?

Alan

Reply | Quote