Linux text comparison-diff&awk

Recently, in order to complete the work, two files A.old and a were compared, then the new additions in a were saved to a, and the other parts were deleted. After finding the relevant data, it is found that there are two better methods.

1. Use the diff command

 diff  old. file  new. file  | grep   " ^>   | sed    s/>//g    > Temp. file   MV  temp. file  new. file  # overwrites the contents of the Temp.file with new. file  

sed ' s/^>//g ' # s means to replace, meaning  to replace all symbols beginning with "" sed's/ The string to be replaced/the new string/g'

2. Using the awk command

awk ' nr==fnr{a[$0]}nr>fnr{if (! ( $ in a)) Print +}' old. file new. file > Temp. file MV Temp. file new. file

awk built-in variables

Variable name	Representative meaning
$	Represents a whole row of data
Nf	Total number of fields per row ($)
Nr	What awk is currently dealing with is the "number of rows" of data
Fs	The current split character, by default, is the space key

The difference between NR and FNR

NR: Current Record Count

FNR: The number of records for the current file, its scope within a file, reopening the file, FNR starting from 1.

The difference between the two is mainly manifested in the number of awk operation files greater than 1 o'clock.

Reference: http://bbs.chinaunix.net/thread-1155924-1-1.html

Http://bbs.chinaunix.net/thread-2309494-1-1.html

Linux text comparison-diff&awk