vim: Notes about using the vim editor, especially useful regular expressions (regex, regexes).

Vim Notebook.md

Vim Notebook

Notes about using the vim editor, especially useful regular expressions (regex, regexes).

• Run an external command on all matching lines

  :g/pattern/.!command

  It's important that the dot command (.) appear between the search pattern and the bang (!) beginning the external command.

  For example, to run the rev shell command on every line that begins with the string "reverse":

  :g/^reverse/.!rev
  

• Delete contiguous duplicate lines without sorting

  :g/^\(.*\)$\n\1$/d
  

• Delete contiguous lines beginning with a hash (pound) sign

  :g/^\#.*$\n#/d
  

  This is similar to removing contiguous duplicate lines, except that the whole contents of the lines don't need to match, they only need to begin with #. This was helpful when editing a .bash_history file. Removing some commands from history left a lot of contiguous history timestamp comment lines, like #1499897624.

• Delete non-contiguous duplicate lines without sorting

  :%!awk '\!a[$0]++'
  

  awk prints the first occurrence of each line by keeping an (associative) array of counters with the text of the line as the index. Whenever the counter for a line is zero (or unset), this expression evaluates to true, which causes awk to print it.

• Collapse single-annotation PHPDoc comments

  :%s;^\(\s*/\*\*\)\n.*\* \(@.*\)\n.*\*/;\1 \2 */;
  

  For example, a PHPDoc like this:

  /**
   * @return bool
   */

  Becomes:

  /** @return bool */

  This helps save screen space, allowing more code to be seen at a time. It's great for declaring the return type of a function (as shown here) or the types of variables (with @var).

  This will probably work with document comments for other languages, too, like JavaDoc, etc.