Question

On bash, I have the following (typical) scenario:

~/some/dir$ find | grep java

which outputs, for example

./subdir1/subdir2/file1.java
./subdir1/subdir3/file2.java
./subdir1/subdir4/file3.java
./subdir1/subdir2/file4.java
./subdir1/subdir6/file5.java

now I know that file5.java, which is the file I was looking for (for example), is on that subdir, so I execute:

vi subdir1/subdir6/file5.java

but in order to do this I have to either type the path (I know that using the TAB key speeds it up) or copy and paste the path to the file

the question is: is there a shortcut/variable on bash (or any other shell) that references the lines outputted by the previous command, so that I can say

vi [whatever the previous command returned on the 5th line]

, and that would be a nice time saver

thanks!

Answer 1

With BASH history you can come close to [whatever the previous command returned on line 5]

Here is an example

 $ find -name "*.java"
 src/file1.java
 ...
 src/file5.java
 ...
 $ vi $( !! | head -5 | tail -1 )

In this example !! represents [whatever the previous command returned] in the sence that the command is executed again.

The following command
| head -5 | tail -1 is a rather awkward way to say [take line 5] from the input

If you are after a particular filename you could as well specify that in the search expression as

$ vi $( !find | grep file5 )

If find takes a long time to execute you could save the output of find into a file and use that file instead.

$ find -name "*.java" | tee filelist.save
 src/file1.java
 ...
 src/file5.java
 ...
 $ vi $( grep file5 filelist.save )