Diff for "BashFAQ/082"

Differences between revisions 2 and 3

Why is $(...) preferred over `...` (backticks)?

For several reasons:

It makes nesting command substitutions easier. Compare:
- ```
  x=$(grep $(dirname "$path") file)
  x=`grep \`dirname "$path"\` file`
```
It just gets uglier and uglier after two levels.
It's easier to read.
Newbies who see $() don't normally press the wrong keys. On the other hand, newbies who see `cmd` often mangle it into 'cmd' because they don't know what a backtick is.

Backslashes (\) inside backticks are handled in a non-obvious manner:

  echo "`echo \\a`"      # prints a
  echo "`echo \\\a`"     # prints \a
  echo "`echo \\\\a`"    # prints \a

Inside $(), there are no such surprises.

The only time backticks are preferred is when writing code for the oldest Bourne shells, which do not know about $().

-  ⇤ ← Revision 2 as of 2007-05-09 18:20:43 → 
  Size: 792
  Editor: GreyCat
  Comment:
+   ← Revision 3 as of 2007-09-13 19:19:33 → ⇥
  Size: 895
  Editor: BetterWorld
  Comment: Add example for non-obvious backslashing behaviour
-Deletions are marked like this.
+Additions are marked like this.
 Line 16:
- * Backslashes (\) inside backticks are handled in a non-obvious manner.  ''(Example desired!)''  Inside {{{$()}}}, there are no such surprises.
+ * Backslashes (\) inside backticks are handled in a non-obvious manner:

  {{{
  echo "`echo \\a`"      # prints a
  echo "`echo \\\a`"     # prints \a
  echo "`echo \\\\a`"    # prints \a}}}

Inside {{{$()}}}, there are no such surprises.