Awk replacement matching

Awk replacement matching awk replacement text content learning feels good [root @ localhosttest] # catawk1a9, 100.341b1, 999.001c5, 656.55 [root @ localhosttest] # awk & amp; #39; {sub (/1/, & quot; test & quot;)} {print & quot; \ n & quot; awk replacement matching
Awk replaces text content to learn [root @ localhost test] # cat awk
1a 9,100.34
1b 1,999.00
1c 5,656.55
[Root @ localhost test] # awk '{sub (/1/, "test")} {print "\ n", $1, $2}' awk
Testa 9,100.34
Testb 1,999.00.
Test C 5,656.55
[Root @ localhost test] # awk '{gsub (/1/, "test")} {print "\ n", $1, $2}' awk
Testa 9, test00.34
Testb test, 999.00
Test C 5,656.55
[Root @ localhost test] # awk '{sub (/[0-9] +/, "")} {print "\ n", $1, $2} 'awk
A 9,100.34
B 1,999.00
C 5,656.55 prints awk' $1 ~ with only four characters in $1 ~ /^ .... $/{Print $1} 'file: The learning result is Yes, but the time and system time cannot match [root @ localhost test] # cat awk
1 A 9,100.34 dkjfjkdkjf 45 lopo
1b 1,999.00 dgfg 456 ll
1c 5,656.55 fghgf 465 df [root @ localhost test] # awk '{$2 = strftime ("% F % T", $2); print $1, $2, $3> "bbb.txt"; print $1, $2, $4> "ccc.txt"} 'awk
[Root @ localhost test] # cat bbb.txt
1a 1969-12-31 16:00:09 dkjfjkdkjf
1b 1969-12-31 16:00:01 dgfg
1c 1969-12-31 16:00:05 fghgf
[Root @ localhost test] # date
Wed Dec 14 22:49:28 PST 2011
[Root @ localhost test] # cat ccc.txt
1a 1969-12-31 16:00:09 45
1b 1969-12-31 16:00:01 456
1c 1969-12-31 16:00:05 465 [root @ localhost test] # date
Wed Dec 14 23:07:09 PST 2011
The problem has been solved. remove $2 from {$2 = strftime ("% F % T", $2) to get the correct format. for details, see a file. the number of columns is different, if there are 5 columns, take the first 4 columns. if there are 6 columns, take the first 5 columns and print them when the first column is greater than 2.