JIYIK CN >

The join command is also a member of the pipeline command family. Its function is to connect the data of those rows with equal columns (that is, the first column) in each file.

First we prepare two files /tmp/join1 and /tmp/join2.

# cat /tmp/join1
www onmpw
domain jiyi
w3 blog
join command
ls l
# cat /tmp/join2
www com
domain cn
w3 net
Join org
wc l

After preparing the above two files, let's take a look at the usage of join without options.

# join /tmp/join1 /tmp/join2
www onmpw com
domain jiyi cn
w3 blog net

Do you see the result? The data in the first column of the first three rows in the join1 and join2 files are equal, so these three rows will be concatenated. For the other two rows, the strings in the first column are not equal, so they will not be concatenated.

Here we may notice that there is join in the join1 file, and the first column of the line in join2 is Join. How should this be matched? By default, join is case-sensitive, so the two are considered unequal. To make the two equal, we need to use the -i option of join.

-i ignore case

# join /tmp/join1 /tmp/join2
www onmpw com
domain jiyi cn
w3 blog net
join command org

Let’s see if there is one more line.

-a FILENUM Print the unjoined lines of the file FILENUM after the normal results.

# join –a 2 /tmp/join1 /tmp/join2
www onmpw com
domain jiyi cn
w3 blog net
Join org
wc l
# join –a 1 /tmp/join1 /tmp/join2
www onmpw com
domain jiyi cn
w3 blog net
join command
ls l

You can understand the meaning of the -a option by looking at the above results. In /tmp/join1, join command and ls l are not connected, so when -a is specified as 1, they are printed out. Similarly, when -a is specified as 2, Join org and wc l are printed out.

-v FILENUM This option has similar functions to -a, the only difference is that -v will not output connected lines, but only output unconnected lines in the specified file.

# join –v 1 /tmp/join1 /tmp/join2
join command
ls l
# join –v 2 /tmp/join1 /tmp/join2
Join org
wc l

-t specifies the delimiter. By default, the fields we are talking about are separated by spaces. With this option, we can change the fields.

# cat /tmp/join3
domain:cs
www:com
domain:cn
w3:net
Join:org
wc:l
# cat /tmp/join4
www:onmpw
domain:jiyi
w3:blog
join:command
ls:l

Let's first look at the default use of space splitting

# join /tmp/join3 /tmp/join4
//The result is empty because each column is not equal

Next, use -t to specify the delimiter

# join –t : /tmp/join3 /tmp/join4
www:com:onmpw
domain:cn:jiyi
w3:net:blog

Let's see if there is any result. Yes, this is the usage of -t.

-o specifies the column to be joined. What does this mean? By default, the first column in common between the two files is joined, and then the contents of the two files are joined in the order of file1 and file2. Using this option is to specify the column we want to join, rather than joining all the columns.

Its format is as follows

-o MN,MN

Here M indicates the number of files, and N indicates the Nth column of the connection.

-o 1.2,2.2

This example means connecting the second column of file 1 and the second column of file 2, of course, the premise is that the two meet the connection conditions.

# join /tmp/join1 /tmp/join2 //First look at the default situation
www onmpw com
domain jiyi cn
w3 blog net

We can see that the output results are the same columns first, then join1 and then join2.

# join –o 1.1,2.2 /tmp/join1 /tmp/join2
www com
domain cn
w3 net

Look at the results displayed in the connection field we specified.

# join –o 1.2,2.2 /tmp/join1 /tmp/join2
onmpw com
jiyi cn
blog net

Specifies that the second column of the first file is displayed concatenated with the second column of the second file. Now you should understand the usage of -o!

-1 FIELD specifies the first file field to compare

-2 FIELD specifies the field of the second file to be compared

As we know, by default, the first field of the two files is used for comparison. Use the above options to change the field to be compared.

# cat /tmp/join5
onmpw www
jiyi domain
blog w3
command join
l ls

First look at the results of join5 and join2 by default

# join /tmp/join5 /tmp/join2
//The result is empty

The reason for the empty result is that the first field of the join5 file is not equal to the first field of the join2 file, so there are no rows to join.

# join -1 2 /tmp/join5 /tmp/join2
www onmpw com
domain jiyi cn
w3 blog net

We specified to use the second field of the first file, that is, the join5 file, and compared it with the first field of the join2 file, and the result appeared because these two fields have equal strings.

After understanding the above example, it is not difficult to understand the use of -1 FIELD and -2 FIELD together.

# join -1 2 -2 2 /tmp/join5 /tmp/join2
//The result is empty

-j FIELD This option is equivalent to -1 FIELD -2 FIELD. No more examples are given here.

This is the end of the introduction to the specific usage of the join command. Finally, it should be noted that it is best to sort the files to be processed before using join to connect them, otherwise some of the compared data will be skipped.