View source for How to Delete Files and Folders via SSH

There are times when you'll need to remove multiple instances of a certain file that is strewn about your websites files.  Maybe you, accidentally created the ''same file'' in every directory within your website and they're too numerous to delete manually. Well, theres a command for that. Multiple actually.

=== Using  RM ===

The simplest way to delete a file a file using the following command: <code>rm</code>

<code>RM</code> allows you to specify multiple files to be deleted in a single command. Like so:

 $ rm file.html file1.html file3.html

However, you would soon notice that listing all the files/folders that need to be deleted can become quite cumbersome and time consuming. Fortunately, <code>rm</code>accepts several arguments, which can assist us in our task. For example; if we take the above example, we could type:

 $ rm file*.html

This would match all files starting with '''file''' and ending in '''.html''' and subsequently delete them.

To delete an ''entire'' directory:

 $ rm -rf foldername/

Or
 $ rmdir foldername/

To delete all files/folders in the current directory, without deleting the directory itself, you would need to use:

 $ rm -rf *

'''Note:''' ''Please be very careful when using rm command, as it might have devastating effects on your website/server, should you delete a system file or a folder.''

=== Using  find and -delete ===
 $ find . -name 'duplicate.html' -type f -delete 

The <code>-delete</code> option means that <code>find</code> will directly delete the matching files as it traverses the directories.

If you need to target only files, using <code>-type f</code> flag will mean that <code>find</code> will only process files and ''not'' directories. To ''target'' directories, you would want to use <code>-type d</code>.

=== Using  find and -exec ===

 $ find . -name \*.html -type f -exec rm -f {} \;
 $ find . -name \*.html -type f -exec rm -f {} +

The <code>-exec</code> option allows <code>find</code> to execute an arbitrary command per line. The first variant <code> {} \;</code> will run the command once per file. Whereas the second will group as many of the commands as possible by replacing <code>{}</code> with as many parameters (files) as possible.

'''$ find . -name \*.html -type f -exec rm -f {} \;''':

An example of the above would be translated into the following if targeting a directory with multiple html files that needed to be removed.

 $ find . -name \*.html -type f -exec rm -f file1.html
 $ find . -name \*.html -type f -exec rm -f file2.html
 $ find . -name \*.html -type f -exec rm -f file3.html
 $ find . -name \*.html -type f -exec rm -f file4.html 


'''$ find . -name \*.html -type f -exec rm -f {} +''':

An example of the above would be translated into the following if targeting a directory with multiple html files that needed to be removed.

 $ find . -name \*.html -type f -exec rm -f file1.html file2.html  file3.html  file4.html

=== Using find and xargs ===

$ find . -name \*.html -type f -print0  | xargs -0 rm -f 

Piping the output to xargs is used form more complex per-file commands than is possible with -exec. The option -print0 tells find to separate matches with ''null'' instead of a ''newline'', and -0 tells xargs to expect ''null''-separated input. This option is safer when dealing with filenames that contain whitespace.

[[Category:SSH]]