App::Info::Util(3) Utility class for App::Info subclasses

SYNOPSIS


use App::Info::Util;
my $util = App::Info::Util->new;
# Subclasses File::Spec.
my @paths = $util->paths;
# First directory that exists in a list.
my $dir = $util->first_dir(@paths);
# First directory that exists in a path.
$dir = $util->first_path($ENV{PATH});
# First file that exists in a list.
my $file = $util->first_file('this.txt', '/that.txt', 'C:\\foo.txt');
# First file found among file base names and directories.
my $files = ['this.txt', 'that.txt'];
$file = $util->first_cat_file($files, @paths);

DESCRIPTION

This class subclasses File::Spec and adds its own methods in order to offer utility methods to App::Info classes. Although intended to be used by App::Info subclasses, in truth App::Info::Util's utility may be considered more general, so feel free to use it elsewhere.

The methods added in addition to the usual File::Spec suspects are designed to facilitate locating files and directories on the file system, as well as searching those files. The assumption is that, in order to provide useful meta data about a given software package, an App::Info subclass must find relevant files and directories and parse them with regular expressions. This class offers methods that simplify those tasks.

CONSTRUCTOR

new

  my $util = App::Info::Util->new;

This is a very simple constructor that merely returns an App::Info::Util object. Since, like its File::Spec super class, App::Info::Util manages no internal data itself, all methods may be used as class methods, if one prefers to. The constructor here is provided merely as a convenience.

OBJECT METHODS

In addition to all of the methods offered by its super class, File::Spec, App::Info::Util offers the following methods.

first_dir

  my @paths = $util->paths;
  my $dir = $util->first_dir(@dirs);

Returns the first file system directory in @paths that exists on the local file system. Only the first item in @paths that exists as a directory will be returned; any other paths leading to non-directories will be ignored.

first_path

  my $path = $ENV{PATH};
  $dir = $util->first_path($path);

Takes the $path string and splits it into a list of directory paths, based on the path delimiter on the local file system. Then calls "first_dir()" to return the first directory in the path list that exists on the local file system. The path delimiter is specified for the following file systems:

  • MacOS: ``,''
  • MSWin32: ``;''
  • os2: ``;''
  • VMS: undef

    This method always returns undef on VMS. Patches welcome.

  • epoc: undef

    This method always returns undef on epoch. Patches welcome.

  • Unix: ``:''

    All other operating systems are assumed to be Unix-based.

first_file

  my $file = $util->first_file(@filelist);

Examines each of the files in @filelist and returns the first one that exists on the file system. The file must be a regular file --- directories will be ignored.

first_exe

  my $exe = $util->first_exe(@exelist);

Examines each of the files in @exelist and returns the first one that exists on the file system as an executable file. Directories will be ignored.

first_cat_path

  my $file = $util->first_cat_path('ick.txt', @paths);
  $file = $util->first_cat_path(['this.txt', 'that.txt'], @paths);

The first argument to this method may be either a file or directory base name (that is, a file or directory name without a full path specification), or a reference to an array of file or directory base names. The remaining arguments constitute a list of directory paths. "first_cat_path()" processes each of these directory paths, concatenates (by the method native to the local operating system) each of the file or directory base names, and returns the first one that exists on the file system.

For example, let us say that we were looking for a file called either httpd or apache, and it could be in any of the following paths: /usr/local/bin, /usr/bin/, /bin. The method call looks like this:

  my $httpd = $util->first_cat_path(['httpd', 'apache'], '/usr/local/bin',
                                    '/usr/bin/', '/bin');

If the OS is a Unix variant, "first_cat_path()" will then look for the first file that exists in this order:

/usr/local/bin/httpd
/usr/local/bin/apache
/usr/bin/httpd
/usr/bin/apache
/bin/httpd
/bin/apache

The first of these complete paths to be found will be returned. If none are found, then undef will be returned.

first_cat_dir

  my $dir = $util->first_cat_dir('ick.txt', @paths);
  $dir = $util->first_cat_dir(['this.txt', 'that.txt'], @paths);

Functionally identical to "first_cat_path()", except that it returns the directory path in which the first file was found, rather than the full concatenated path. Thus, in the above example, if the file found was /usr/bin/httpd, while "first_cat_path()" would return that value, "first_cat_dir()" would return /usr/bin instead.

first_cat_exe

  my $exe = $util->first_cat_exe('ick.exe', @paths);
  $exe = $util->first_cat_exe(['this.exe', 'that.exe'], @paths);

Functionally identical to "first_cat_path()", except that it returns the full path to the first executable file found, rather than simply the first file found.

search_file

  my $file = 'foo.txt';
  my $regex = qr/(text\s+to\s+find)/;
  my $value = $util->search_file($file, $regex);

Opens $file and executes the $regex regular expression against each line in the file. Once the line matches and one or more values is returned by the match, the file is closed and the value or values returned.

For example, say foo.txt contains the line ``Version 6.5, patch level 8'', and you need to grab each of the three version parts. All three parts can be grabbed like this:

  my $regex = qr/Version\s+(\d+)\.(\d+),[^\d]*(\d+)/;
  my @nums = $util->search_file($file, $regex);

Now @nums will contain the values "(6, 5, 8)". Note that in a scalar context, the above search would yield an array reference:

  my $regex = qr/Version\s+(\d+)\.(\d+),[^\d]*(\d+)/;
  my $nums = $util->search_file($file, $regex);

So now $nums contains "[6, 5, 8]". The same does not hold true if the match returns only one value, however. Say foo.txt contains the line ``king of the who?'', and you wish to know who the king is king of. Either of the following two calls would get you the data you need:

  my $minions = $util->search_file($file, qr/King\s+of\s+(.*)/);
  my @minions = $util->search_file($file, qr/King\s+of\s+(.*)/);

In the first case, because the regular expression contains only one set of parentheses, "search_file()" will simply return that value: $minions contains the string ``the who?''. In the latter case, @minions of course contains a single element: "("the who?")".

Note that a regular expression without parentheses --- that is, one that doesn't grab values and put them into $1, $2, etc., will never successfully match a line in this method. You must include something to parenthetically match. If you just want to know the value of what was matched, parenthesize the whole thing and if the value returns, you have a match. Also, if you need to match patterns across lines, try using multiple regular expressions with "multi_search_file()", instead.

files_in_dir

  my @files = $util->files_in_dir($dir);
     @files = $util->files_in_dir($dir, $filter);
  my $files = $util->files_in_dir($dir);
     $files = $util->files_in_dir($dir, $filter);

Returns an list or array reference of all of the files and directories in the file system directory $dir. An optional second argument is a code reference that filters the files. The code reference should examine the $_ for a file name and return true if it's a file that you're interested and false if it's not.

multi_search_file

  my @regexen = (qr/(one)/, qr/(two)\s+(three)/);
  my @matches = $util->multi_search_file($file, @regexen);

Like "search_file()", this method opens $file and parses it for regular expression matches. This method, however, can take a list of regular expressions to look for, and will return the values found for all of them. Regular expressions that match and return multiple values will be returned as array references, while those that match and return a single value will return just that single value.

For example, say you are parsing a file with lines like the following:

  #define XML_MAJOR_VERSION 1
  #define XML_MINOR_VERSION 95
  #define XML_MICRO_VERSION 2

You need to get each of these numbers, but calling "search_file()" for each of them would be wasteful, as each call to "search_file()" opens the file and parses it. With "multi_search_file()", on the other hand, the file will be opened only once, and, once all of the regular expressions have returned matches, the file will be closed and the matches returned.

Thus the above values can be collected like this:

  my @regexen = ( qr/XML_MAJOR_VERSION\s+(\d+)$/,
                  qr/XML_MINOR_VERSION\s+(\d+)$/,
                  qr/XML_MICRO_VERSION\s+(\d+)$/ );
  my @nums = $file->multi_search_file($file, @regexen);

The result will be that @nums contains "(1, 95, 2)". Note that "multi_file_search()" tries to do the right thing by only parsing the file until all of the regular expressions have been matched. Thus, a large file with the values you need near the top can be parsed very quickly.

As with "search_file()", "multi_search_file()" can take regular expressions that match multiple values. These will be returned as array references. For example, say the file you're parsing has files like this:

  FooApp Version 4
  Subversion 2, Microversion 6

To get all of the version numbers, you can either use three regular expressions, as in the previous example:

  my @regexen = ( qr/FooApp\s+Version\s+(\d+)$/,
                  qr/Subversion\s+(\d+),/,
                  qr/Microversion\s+(\d$)$/ );
  my @nums = $file->multi_search_file($file, @regexen);

In which case @nums will contain "(4, 2, 6)". Or, you can use just two regular expressions:

  my @regexen = ( qr/FooApp\s+Version\s+(\d+)$/,
                  qr/Subversion\s+(\d+),\s+Microversion\s+(\d$)$/ );
  my @nums = $file->multi_search_file($file, @regexen);

In which case @nums will contain "(4, [2, 6])". Note that the two parentheses that return values in the second regular expression cause the matches to be returned as an array reference.

lib_dirs

  my @dirs = $util->lib_dirs;

Returns a list of possible library directories to be searched. These are gathered from the "libsdirs" and "loclibpth" Config settings. These are useful for passing to "first_cat_dir()" to search typical directories for library files.

SUPPORT

This module is stored in an open GitHub repository <http://github.com/theory/app-info/>. Feel free to fork and contribute!

Please file bug reports via GitHub Issues <http://github.com/theory/app-info/issues/> or by sending mail to [email protected] <mailto:[email protected]>.

AUTHOR

David E. Wheeler <[email protected]>

COPYRIGHT AND LICENSE

Copyright (c) 2002-2011, David E. Wheeler. Some Rights Reserved.

This module is free software; you can redistribute it and/or modify it under the same terms as Perl itself.