sh AWK - 标准用途

Posted 2021-05-07

# Simple awk usage - Print all lines
$ awk '{ print }' /etc/passwd
$ awk '{ print $0 }' /etc/passwd

# Print blank lines = document lines
$ awk '{ print "" }' /etc/passwd

# Print hiya = document lines
$ awk '{ print "hiya" }' /etc/passwd

# Print first field - Use declared delimiter
$ awk -F":" '{ print $1 }' /etc/passwd

# Print first and third field - Use declared delimiter
$ awk -F":" '{ print $1 $3 }' /etc/passwd

#######################################################
# Way to call an awk script from the command line
$ awk -f myscript.awk myfile.in


########################################################
# Way to call an awk script and pass parameter to it
# Inline script
awk -v var1=$column '{ x+=$var1 } END { print x }'



########################################################
# Way to call an awk script and pass parameters to it
# Normal script
awk -v var1=$column -v var2=$column2 -f myscript.awk myfilein > myfile.ot


########################################################
# Way to call an awk script that will deal with greek letters
gawk -b -v var1=$aa -f myscript.awk filein > fileot



########################################################
# BASIC STRUCTURE OF AWK SCRIPT

# BEGIN block is executed before file line iteration
# Used traditionally to initialize variables, set record separators (IFS/OFS)
BEGIN { 
        FS=":" 
} 

# MAIN BLOCK 1
# If the block has no declared pattern then it applies to every line
{ 
  print $1 
}

# MAIN BLOCK 2
# If the block has specified conditions then it applies
# only to lines that meet those conditions.
NR>2 {
  
  print $0

}

# END BLOCK
# Executed after reading the last line of input file

# BASIC TYPES
# If block with pattern check in condition
# equivalent to :
# awk '$5 ~ /root/ { print $3 }'
{ 
  if ( $5 ~ /root/ ) { 
          print $3 
  } 
}

# If block with string eq check in condition
# equivalent to :
# awk '$5 == "string" { print $3 }'
{ 
  if ( $5 == "string" ) { 
          print $3 
  } 
}

# If block with numerical eq check in condition
# equivalent to :
# awk '$5 == 7 { print $3 }'
{ 
  if ( $5==7 ) { 
          print $3 
  } 
}

# If block with pattern check and regex in condition
# equivalent to :
# awk '$5 ~ /[A|B][AB][0-9]*/ { print $3 }'
{ 
  if ( $5 ~ /[A|B][AB][0-9]*/ ) { 
          print $3 
  } 
}

############################################
# NESTED CONDITIONALS
{ 
  if ( $1 == "foo" ) { 
           if ( $2 == "foo" ) { 
                    print "uno" 
           } else { 
                    print "one" 
           } 
  } else if ($1 == "bar" ) { 
           print "two" 
  } else { 
           print "three" 
  } 
}

##########################################
# NEGATION
# USING THE !
# Equivalent to:
# awk ' $0 !~ /matchme/ { print $1 $3 $4 }
# 
! /matchme/ { print $1 $3 $4 }
to this:
#
# also equivalent to:
{ 
  if ( $0 !~ /matchme/ ) { 
          print $1 $3 $4 
  } 
}

############################################
# LOGICAL OPERATORS
# Using &&(AND) - ||(OR)

awk '( $1 == "foo" ) && ( $2 == "bar" ) { print }'

# eqyivalent to:
BEGIN {
  
  
}

($1 == "foo" ) && ( $2 == "bar" ) {
  
      print $0
      
}

awk '/foo/ { print }'

# Basic structure :
# One liner : awk ' <condition> { <actions on each line that meets condition}'
# Script:
# 
# BEGIN {
# }
#
# <condition> { <actions on each line that meets it
# }
#
# END {
# }
#

# Regular expression as a condition
/[0-9]+\.[0-9]*/ { print }

# String comparison as condition
$1 == "fred" { print $3 }

# Pattern matching as condition
$5 ~ /root/ { print $3 }

# Numerical comparison as condition
$5 == 4 { print $3 }

# Using internal variables as condition
NR>2 { print $0 }
NF<4 { print $1 $2 }

# Negation of conditions (NOT)

# Multiple conditions on tha same block will need to be bound
# by AND/OR/NOT.

# Search the 6th field for given pattern - print entire line
		awk ' $6 ~ /[OoNn][CcOo][TtvV]/ { print $0 }' sedtest  

# Search the 5th field for values greater > 100 and print entire line
		awk '{ if ($5>100) } { print $0 }' sedtest
		
# Same as above, but with condition set as pattern
		awk ' $5>100  { print $0 }' sedtest
		awk '{if($5>100)   print $0 }' sedtest

# Search in 6th field for the string "Nov" and print entire line
awk ' $6="Nov" { print } ' sedtest
awk ' $6="Nov" { print $0 } ' sedtest
awk '{ if($6="Nov")  print $0 } ' sedtest

# Search in 6th field for the string "Nov" and print substring
awk '{ if($6="Oct") print substr($4,0,2) }' sedtest
awk '$6="Oct" { print substr($4,0,2) }' sedtest
awk '$6="Oct" { print substr($4,0,4) }' sedtest

# Search in 9th field for pattern and print entire line
awk ' $9 ~ /arraytest/ { print $0 }' sedtest

# Search in all lines with even line number and print them
awk ' NR%2 { print $0 } ' sedtest

# Search for lines that begin with -rwx and print them
awk ' /^ -rwx/ { print }' sedtest
awk ' /^ "-rwx"/ { print }' sedtest
awk ' /^ \-rwx/ { print }' sedtest

# Will print the lines that match continuously
awk '{ while($6="Nov")  print } ' sedtest

# Print specified fields and substrings 
awk '  {print $2, $3, $5, substr($8,0,2)}' sedtest

# Print all lines after the fourth - NR in condition
awk ' NR>4 {print $0}' sedtest

# Print all lines that do not contain pattern in the first field
awk ' $1 !~ /total/ { print }' sedtest

# Print 8th field lines that fifth field is greater than 2000
awk '$5>2000 { print $8 }' sedtest 
awk '{ if ($5>2000)  print $8 }' sedtest

# Various ways of pattern matching and negative pattern matching using !
awk ' $9 ~ /columncut[1-3]/ { print }' sedtest
awk ' $9 !~ /columncut[1-3]/ { print }' sedtest
awk '{ if($9 ~ /columncut*/) print $0}' sedtest
awk '{ if($9 ~ /columncut./) print $0}' sedtest
awk '{ if($9 ~ /columncut/) print $0}' sedtest
awk '{ if($9 ~ /columncut..bsh/) print $0}' sedtest

# Various ways of using NR to specify condition
awk '{ if(NR%4!=0) print }' sedtest
awk '{ if(NR%4>1) print }' sedtest
awk '{ if(NR>1) print }' sedtest
awk '{ if(NR>5) print }' sedtest

# Look for match in the beginning of line
awk '/^-/ { print } ' sedtest
awk '/^-rwx/ { print } ' sedtest

# Look for substring of entire line - offset 1 and print entire line
awk '{ if ( substr($0, 1,  1) == "1" ) print $0  }' all.srtmrg.pan

# This snippet details various ways the substring command can be used 
# substring can be placed in BEGIN, END and main code block.
# substring can be used as a "line pattern" or into the code block.



 # Using awk substring in pattern with equality condition 
 awk 'substr($0, 1, 5) == 11131    { print substr($0, 1, 5) }' TEST.OUT
 
 # Using awk substring in pattern with regex condition ( We need to enclose it in / /) 
 awk 'substr($0, 1, 5) ~ /11131/    { print substr($0, 1, 10) }' TEST.OUT
 
 # Using awk substring in pattern with pattern containing escape characters.
 awk 'substr($0, 1, 5) ~ /16\/12\/2018/ { print $0 }' BUT000.TXT > TEST2.OUT
 
 # Using awk substring with delimiter, substring in pattern and pattern containing escape characters
 $ awk -F"," '$11 == 16\/12\/2018 { print $0 }' BUT000.TXT > TEST2.OUT
 
 # Using awk substring in code block with eq. condition 
  awk 'substr($0, 1, 5) == 11131    { if(substr($0, 1, 5)==11131){ a = "panagos"; print a} else { print substr($0, 1, 10)} }' TEST.OUT

 # Using awk substring in code block with regex condition 
 awk 'substr($0, 1, 5) == 11131    { if(substr($0, 1, 5)~/11131/){ a = "panagos"; print a} else { print substr($0, 1, 10)} }' TEST.OUT

 # Using awk substring in print 
awk 'substr($0, 1, 5) ~ /11\131/    { print substr($0, 1, 10) }' TEST.OUT
 
 # Multiple uses on all levels.
  awk 'substr($0, 1, 5) ~ /16\/12\/2018/ { if(substr($0, 1, 2)==3){ varmode = "new"; print $0, substr($0, 1, 20); }}

# gawk -F "," 'substr($9, 0, 1)==3 && length($9)==12 { print $9 }' MISINDYPODIEYREYNISI.TXT > out1.txt^C

gawk -F "," 'substr($10, 0, 3)=3 && length($10)==12 { gsub(/,/,".", $10)l print}'