Shell Programming and Scripting

Hi All,

Seeking for your assistance to get the records once the $2 met the condition.

Code:

Ex. file 1.txt
123455,10-Aug-2020 07:33:37 AM,2335235,1323534,12343
123232,11-Aug-2015 08:33:37 PM,4234324,1321432,34364

Output:
123455,10-Aug-2020 07:33:37 AM,2335235,1323534,12343

What i did was but nothing happens.

Code:

awk -F "," '($2=="10-Aug-2020")' file1.txt

Please advise,

Thanks,

Hi

What about (untested):

Code:

RANGE=10-Aug-2020
awk -v RANGE=$RANGE '/RANGE/ {print $0}' file1.txt

hth

That will print each line containing the literal string RANGE anywhere on the line. But, you're on the right path. Try:

Code:

RANGE=10-Aug-2020
awk -v RANGE=$RANGE -F"," '$2 ~ RANGE {print $0}' file1.txt

or, using default actions, just:

Code:

RANGE=10-Aug-2020
awk -v RANGE=$RANGE -F"," '$2 ~ RANGE' file1.txt

In that case i was refering to:

Code:

awk  '/10-Aug-2020/ {print $0}' file1.txt

(i tend to use/(believe that) variables (are) just like regular strings)
BTW Don, i belive you're missing a -F,.

Just why again does a variable not expand (as in: represents the assigned value) when passed with the first method?

Because, just tonight i've had an issue resolving the properly limited output of such an $2 ~ VARNAME condition, as it printed 2 lines, allthough just one was expected, but the other was similar.
Namely, i did VARNAME=bindir and file contained bindir= and sbindir=, and the values of both were output at each of them, which in regard of the initial purpose has to be considered a malfunction (imo).
Me solved it (there) with ${!VARNAME}, but thats not an option here

Thanks Everyone for your usual support.

Hi sea,
Yes, I missed specifying the field separator. I will fix that in my earlier post shortly. Thank you for pointing it out to me.

Matching in awk uses extended regular expressions (EREs). An ERE in awk is specified by a string or by an ERE token when it is used as the 2nd operand of a ~ or !~ operator; when setting FS inside an awk script; when used as the ERE argument in a gsub(), match(), or sub() function call; and when it is used as the field separator argument in a split() function call. An ERE token is an ERE surrounded by / characters. A string can be a literal string surrounded by " characters, the expansion of a variable, or a concatenation of variable expansions and quoted strings. An ERE token (but not a string) can also be used as the pattern in a pattern/action pair in an awk statement (and in that case /ERE/ is identical in meaning to $0~/ERE/). (A string in this context evaluates to true if the string is not an empty string. A matching operation is not performed in this case.)

A variable in awk does not expand when it is inside quotes and does not expand when it is inside an ERE token just like a variable doesn't expand in the shell when it is included in a string surrounded by single quotes. This is true if the variable is set using a -v variable=value option to awk, is set inside the awk script, or is set as a variable=value operand mixed in with the list of filenames at the end of an awk command invocation.

As with any other RE match, if you want a match to occur at the start or end of a field, you need to use an RE anchor operator (such as in $2 ~ ("^" VARNAME) to find bindir= only if it appears at the start of field 2).