Output same of the data would make it easier to actually test our results.
On Tue, Jul 7, 2015 at 4:09 PM, Larry Brigman <[email protected]>
wrote:
> You mean something like this?
> awk '{sub(/^\./, "", $1); printf(".%s", $1)}'
>
> This takes the first arg and looks for . at the beginning of the line and
> removes it and prints.
> The print happens regardless of the modification. In this example the
> input to this was the output of
> 'tar tf' being converted to an absolute location.
>
> I haven't use sub without a target recently. So off to the man page.....
> It operates on $0 which is the whole line and sub will operate on the
> longest matched string once per invocation meaning per line.
>
> On Tue, Jul 7, 2015 at 3:39 PM, Rich Shepard <[email protected]>
> wrote:
>
>> I understand sub(), gsub(), and substr() but have difficulty figuring
>> out
>> how to use any or all to modify a field's contents.
>>
>> Context: data files have variable numbers of fields per record; most of
>> these fields represent measured values as integers or (more often) as
>> floating point numbers. When a value is below the laboratory's method
>> detection limit they report the value with '<', '< ', or '-' preceeding
>> the
>> floating point number. I want to strip off the leading symbol (with any
>> following white space) and retain the floating point number as the field's
>> content.
>>
>> It seems that sub() should do the job when I understand how to present
>> the
>> arguments to the function. The syntax is sub(regex, replacement [,
>> target]).
>> Not knowing if I can have if/elseif/elseif/else in the regex position my
>> initial approach is to use four calls to the function; e.g.,
>>
>> sub(/\</,"") or sub(/\< /,"")
>>
>> Would these leave the remainder of the field's contents intact? gsub()
>> probably adds no more capabilities to solving this problem than does
>> sub().
>> Not sure if substr() is really appropriate here.
>>
>> I've read the two post-"The AWK Book" books I have and understand the
>> syntax but not how to apply the functions to achieve what needs to be
>> done.
>>
>> This is just one issue that I need to grok while developing this
>> generic
>> data cleaning program.
>>
>> Rich
>> _______________________________________________
>> PLUG mailing list
>> [email protected]
>> http://lists.pdxlinux.org/mailman/listinfo/plug
>>
>
>
_______________________________________________
PLUG mailing list
[email protected]
http://lists.pdxlinux.org/mailman/listinfo/plug
