=head1 NAME

urpmf - Finds the packages matching some conditions

=head1 SYNOPSIS

    urpmf [options] <pattern-expression>

=head1 DESCRIPTION

By default, urpmf outputs the list of the known packages that contain a
file whose name or path was specified as an argument on the command-line.
<I<pattern-expression>> is the name or part of the name of the file you
want to locate. Perl-style regular expressions are allowed. urpmf will
then print a list of matching files, with their corresponding packages. By
using options, you may format output differently, and search through other
rpm tags. The default behaviour corresponds to the B<--files> option;
using other options makes urpmf search through other tags' contents.

Note that urpmf searches through the urpmi media (B<not> in
the system's rpm database -- for that you should use rpm(8) instead).

=head1 OPTIONS

=over

=item B<--help>

Prints a help message and exits.

=item B<--version>

Prints the version number and exits.

=item B<--env> I<directory>

Use a different environment directly out of a bug report tarball.

=item B<--excludemedia> I<medium1,...,mediumN>

Do not use the given media.

=item B<--literal>

Match literally instead of interpreting the argument as a regular
expression.

=item B<-l>

Equivalent to B<--literal>.

=item B<--media> I<medium1,...,mediumN>

Select specific media to be used, instead of defaulting to all available
media (or all update media if B<--update> is used).

=item B<--sortmedia> I<medium1,...,mediumN>

Sort the given media, substrings may be used to simplify grouping.
This way, C<media1> will be taken into account first, then C<media2>, and
so on. Media which aren't listed are taken into account after the others.

=item B<--synthesis> I<file>

Use the given synthesis file instead of urpmi db for searching packages.

=item B<--uniq>

Do not print identical lines twice.

=item B<--update>

Use only update media. This means urpmf will only search through media
marked as update.

=item B<--verbose>

Verbose mode. urpmf will emit various messages related to the parsing of
hdlist files for your media.

=item B<-i>

Ignore case distinctions in the patterns that follow.

=item B<-I>

Honor case distinctions in the patterns that follow. This is the default
behaviour.

=item B<-F> I<string>

Change field separator to I<string> for the rest of the output line (it
defaults to ':')

=item B<--qf> I<format-string>

Specify a printf-like output format.

=item B<--arch>

Prints the architecture tag.

=item B<--buildhost>

Prints the BuildHost tag.

=item B<--buildtime>

Prints the
BuildTime tag.

=item B<--conffiles>

Prints the list of configuration files contained
in the package. (Might be empty.)

=item B<--conflicts>

Prints the Conflicts tags.

=item B<--description>

Prints the Description tag.

=item B<--distribution>

Prints the Distribution tag.

=item B<--epoch>

Prints the Epoch tag.

=item B<--filename>

Prints package file names.

=item B<--files>

Prints the list of files contained in the rpm (this is the default if you
specify no field).

=item B<--group>

Prints the Group tag.

=item B<--name>

Prints package names.

=item B<--obsoletes>

Prints the Obsoletes tags.

=item B<--packager>

Prints the Packager tag.

=item B<--provides>

Prints the Provides tags.

=item B<--requires>

Prints the Requires tag.

=item B<--size>

Prints the Size tag.

=item B<--sourcerpm>

Prints the names of source rpms.

=item B<--summary>

Prints the Summary tag.

=item B<--url>

Prints the packages' URL.

=item B<--vendor>

Prints the Vendor tag.

=item B<-m>

Print the name of the media in which the
package has been found.

=item B<-f>

Print version, release and arch along with name
of package. This modifies the effect of the I<--name> option.

=item B<-e>

Include code directly
in the perl search expression. Use it with B<--debug> to look at the generated
perl code. In any cases, use it only if you know perl.

=item B<-a>

Binary AND operator, true if both expressions are true.

=item B<-o>

Binary OR operator, true if one expression is true.

=item B<!>

Unary NOT, true if expression is false.

=item B<(>

Left parenthesis to open a group expression.

=item B<)>

Right parenthesis to close a group expression.

=item B<--debug>

Includes debugging output.

=back

=head1 FORMAT

The options to select tags (B<--arch>, B<--buildhost> and so on) control
the output format of urpmf. If not explicitly specified somewhere else,
the name of the rpm is printed first. Fields are separated by C<:> until
specified otherwise.

You can also use B<--qf> for that. It takes a format string, where tags to
be replaced are specified as B<%>I<[-][number]>B<tagname>. The optional
number indicates the length of the space-padding and the B<-> character
its justification (like in printf(3)).

=head1 EXPRESSIONS

The boolean operator switches allow you to construct complex expressions
from regexps. Those expressions are then matched against the whole output
line.

=head1 EXAMPLE

This command will list all C<perl-Apache> modules, and list them along
with the media in which they're been found, nicely indented :

    urpmf --qf '%-30name is found in media %media' perl-Apache

=head1 FILES

See urpmi.files(5).

=head1 SEE ALSO

urpmi.addmedia(8), urpmi.update(8), urpmi.removemedia(8), urpmi(8),
urpmq(8), urpmi.files(5).

=head1 AUTHOR

Pascal Rigaux <pixel@mandriva.com> (original author and current maintainer), 
FranE<ccedil>ois Pons, Rafael Garcia-Suarez <rgarciasuarez@mandriva.com>