xgrep 0.0.5

A library/script for locating/grepping things on linux

Xgrep

A library and script for linux users that combines indexing directories, specifying filter sets, and grepping of files. Essentially it helps the user not have to remember the details of and path inputs for find | xargs grep...

Are you a graybeard whose memory is not what it used to be?

Are you a die-hard emacs user getting somewhat envious of some of the nice find capabilities of sublime?

Do you want to reduce the time it takes to do a find in all files you are interested in from a simple command so you can use it from terminal, emacs, or whatever?

Purpose

The following is the intent:

• Allow persistence of index definitions - i.e. named collections of paths you frequently visit/search using find/grep or updatedb/mlocate/grep.

• Allow creation of file filter sets that can be incorporated into a search to weed out unwanted hits.

• Easily update your persisted indices.

• Provide the find|grep equivalent combo in single script with less to remember

• Provide generation of $HOME/.xgrep.el for easy access to your grepping commands from emacs

Prereqs/Notes

This uses MongoDB for the persistence of indices and filters. It uses updatedb/mlocate for the creation/search of directories. The databases are stored in $HOME/xgreps/...

The default MongoDB uri is "mongodb://127.0.0.1/xgreps". You can override this by setting the XGREP_MONGO_URI environment variable.

If the -e option is used the file $HOME/.xgreps.el is written

Examples

• Create an index of several project directories, in this case with some prune names:

    xgrep -i my_dart \\
    -p \$HOME/dev/open_source/id:.pub:.git \\           
    -p \$HOME/dev/open_source/xgrep:.pub:.git \\
    -p \$HOME/dev/open_source/ebisu:.pub:.git \\
    -p \$HOME/dev/open_source/ebisu_cpp:.pub:.git \\       
  

• Use that index to list all files:

    xgrep -i my_dart -l
  

• If you added/removed files, easily update the index

    xgrep -i my_dart -u
  

• Use the index to grep for stuff, like usage of split or join in all files

    xgrep -i my_dart split join
  

• Create many indices then grep over all files in all indices, since the named indices can be queried with pattern

    xgrep -i.* split join
  

  giving:

    .../ebisu.dart:152:    result.add(guts.join(',\n'));
    .../ebisu.dart:163:    result.add(guts.join(',\n'));
    ...
    .../ebisu_cpp.dart:448:template< ${decls.join(',\n          ')} >''';
    .../ebisu_cpp.dart:672:  return result.join('');
    .../ebisu_cpp.dart:676:    original.split('\n').map((l) => '"$l\\n"').join('\n');
  

• Similarly update all indices

    xgrep -i.* -u
  

• Generate emacs file with handy emacs functions for all your indices

    xgrep -e
  

  which will create functions like:

    (defun xgu-* ()
      "Update all xgrep indices"
      (interactive)
      (shell-command "xgrep -i.* -u" "update all xgrep indices"))
  
    (defun xg-my-dart (args)
      "Do an xgrep -i my_dart with additional args. Look for things in the index"
      (interactive "sEnter args:")
      (grep (concat "xgrep -i my_dart " args))
      (set-buffer "*grep*")
      (rename-buffer (concat "*xg-my-dart " args "*") t))
  
    (defun xgl-my-dart ()
      "List all files in the index my_dart"
      (interactive)
      (compile "xgrep -i my_dart -l")
      (set-buffer "*compilation*")
      (rename-buffer "*list of my_dart*" t))
  
    (defun xgu-my-dart ()
      "Update the index my_dart"
      (interactive)
      (grep "xgrep -i my_dart -u"))