[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [bug-gawk] Debugging Script

From: david kerns
Subject: Re: [bug-gawk] Debugging Script
Date: Sun, 14 Apr 2019 07:15:53 -0700

There is a feature request to be had here though...
I often have a wrapper shell script around my awk programs, and when
there's a syntax error in the awk program I get a fatal error with a line
number. But the line number is relative to start of the awk program, not
the entire script (which is what I edit). bash's LINENO variable is not
export by default and is not set unless evaluated (per experimentation, not
a read of the source code) so you'd have to -v LINENO=$LINENO on the awk
cmd line... but it would be handy if LINENO was set, to add it to the
reported value. Since the developer had do something to manipulate LINENO
you could get away without reporting both values.
... mull it over ;)

On Sun, Apr 14, 2019 at 1:16 AM <address@hidden> wrote:

> Holger Klene <address@hidden> wrote:
> > So if a normal file is a glass of beer, you say these strange files
> > cannot be put on the table, you have to drink them in one go.
> A normal file is a piece of paper you can read as many times as you
> want.  The other files self-destruct after reading them once. :-)
> > Yes indeed, it is a feature request! The author of the document foresaw
> > it coming [1] ... and ever since whenever people stumbled over it, it
> > gives the impression: this is a known topic and you better not waste
> > anybody's time by asking a question from the FAQ again.
> I think you're reading more into it than what's there.
> > Meanwhile a kind soul taught me how easy it can be to write a temporary
> > file (I better not go into the details of what I tried to accomplish the
> > same):
> >
> > > cat <<'TheEnd' >/tmp/$USER$$
> > >    .
> > >    . 20 lines of awk code, nicely indented
> > >    .
> > > TheEnd
> > > gawk -f /tmp/$USER$$
> > > rm /tmp/$USER$$
> This is indeed the way to go.
> > Maybe you could add something like this as a "known workaround" to the
> > known limitations [1]?
> I will try to add something.
> > In the end I could solve my problem and with your explanations, I
> > somewhat understand it now.
> Great. Enjoy.
> Arnold

reply via email to

[Prev in Thread] Current Thread [Next in Thread]