NAME

Module::Generic::Finfo - File Info Object Class

SYNOPSIS

use Module::Generic::Finfo qw( :all );
my $finfo = Module::Generic::Finfo->new( '/some/file/path.html' );
# Get access time as a DateTime object
$finfo->atime;
# Block site
$finfo->blksize;
# Number of blocks
$finfo->blocks;
if( $finfo->can_read )
{
    # Do something
}
# Can also use
$finfo->can_write;
$finfo->can_exec;
$finfo->csize;
# Inode change time as a DateTime object
$finfo->ctime;
$finfo->dev;
if( $finfo->exists )
{
    # Do something
}
print "File path is: ", $finfo->filepath;
if( $finfo->filetype == FILETYPE_NOFILE )
{
    # File does not exist
}
# Same as $finfo->filepath
print "File path is: ", $finfo->fname;
print "File group id is: ", $finfo->gid;
# Can also use $finfo->group which will yield the same result
$finfo->ino;
# or $finfo->inode;
if( $finfo->is_block )
{
    # Do something
}
elsif( $finfo->is_char )
{
    # Do something else
}
elsif( $finfo->is_dir )
{
    # It's a directory
}
elsif( $finfo->is_file )
{
    # It's a regular file
}
elsif( $finfo->is_link )
{
    # A file alias
}
elsif( $info->is_pipe )
{
    # A Unix pipe !
}
elsif( $finfo->is_socket )
{
    # It's a socket
}
elsif( ( $info->mode & 0100 ) )
{
    # Can execute
}
$finfo->mtime->strftime( '%A %d %B %Y %H:%m:%S' );
print "File base name is: ", $finfo->name;
printf "File has %d links\n", $finfo->nlink;
print "File permission in hexadecimal: ", $finfo->protection;
$finfo->rdev;
$finfo->size;
my $new_object = $finfo->stat( '/some/other/file.txt' );
# Get the user id
$finfo->uid;
# Or
$finfo->user;

VERSION

v0.5.3

DESCRIPTION

This class provides a file info object oriented api.

The other advantage is that even if a non-existing file is provided, an object is returned. Obviously many of this module's methods will return an empty value since the file does not actually exist.

METHODS

new

This instantiate an object that is used to access other key methods. It takes a file path.:

atime

Returns the file last access time as a Module::Generic::DateTime object, which stringifies to its value in second since epoch. Module::Generic::DateTime is just a thin wrapper around DateTime to allow a DateTime to be used in comparison with another non DateTime value.

For example:

if( $finfo->atime > time() + 86400 )
{
    print( "You are traveling in the future\n" );
}

blksize

Returns the preferred I/O size in bytes for interacting with the file. You can also use block_size.

block_size

Alias for "blksize"

blocks

Returns the actual number of system-specific blocks allocated on disk (often, but not always, 512 bytes each).

can_read

Returns true if the the effective user can read the file.

can_write

Returns true if the the effective user can write to the file.

can_exec

Returns true if the the effective user can execute the file. Same as "execute"

can_execute

Returns true if the the effective user can execute the file. Same as "exec"

csize

Returns the total size of file, in bytes. Same as "size"

ctime

Returns the file inode change time as a Module::Generic::DateTime object, which stringifies to its value in second since epoch. Module::Generic::DateTime is just a thin wrapper around DateTime to allow a DateTime to be used in comparison with another non DateTime value.

dev

Returns the device number of filesystem. Same as "dev"

device

Returns the device number of filesystem. Same as "device"

exists

Returns true if the filetype is not "FILETYPE_NOFILE"

filepath

Returns the file path as a string. Same as "fname"

filetype

Returns the file type which is one of the "CONSTANTS" below.

fname

Returns the file path as a string. Same as "filepath"

gid

Returns the numeric group ID of file's owner. Same as "group"

group

Returns the numeric group ID of file's owner. Same as "gid"

ino

Alias for "ino"

inode

Returns the inode number.

is_block

Returns true if this is a block file, false otherwise.

is_char

Returns true if this is a character file, false otherwise.

is_dir

Returns true if this is a directory, false otherwise.

is_file

Returns true if this is a regular file, false otherwise.

Returns true if this is a symbolic link, false otherwise.

is_pipe

Returns true if this is a pipe, false otherwise.

is_socket

Returns true if this is a socket, false otherwise.

mime_type

my $mime = $finfo->mime_type;
print "MIME type: $mime\n"; # e.g., "text/plain"

This guesses the file mime type and returns it as a scalar object

If File::MMagic::XS is installed, it will use it, otherwise, it will use File::MMagic

If an error occurs, it will set an exception object, and return an empty list in list context, or undef in scalar context.

mode

Returns the file mode. This is equivalent to the mode & 07777, ie without the file type bit.

So you could do something like:

if( $finfo->mode & 0100 )
{
    print( "Owner can execute\n" );
}
if( $finfo->mode & 0001 )
{
    print( "Everyone can execute too!\n" );
}

mode_n2s

Takes a numeric mode (e.g., 0644) of a file or directory, and returns a human-readable string (e.g., rw-r--r--). Returns undef if the mode is invalid:

my $mode_str = $finfo->mode_n2s( 0644 );
print "Mode: $mode_str\n"; # "rw-r--r--"

mode_s2n

Takes a mode string (e.g., rw-r--r--) of a file or directory, and converts it to octal mode (e.g., 0644). Returns undef if the mode string is invalid:

my $mode_num = $finfo->mode_s2n( "rw-r--r--" );
print "Numeric mode: $mode_num\n"; # 0644

mtime

Returns the file last modify time as a Module::Generic::DateTime object, which stringifies to its value in second since epoch. Module::Generic::DateTime is just a wrapper around DateTime to allow a DateTime to be used in comparison with another non DateTime value.

name

Returns the file base name. So if the file is /home/john/www/some/file.html this would return file.html

Interesting to note that "name" in APR::Finfo which is advertised as returning the file base name, actually returns just an empty string. With this module, this uses a workaround to provide the proper value. It use "basename" in File::Basename on the value returned by "fname"

Returns the number of (hard) links to the file.

permission

Returns the file permission as a 4 digits octal value, such as 0755 or 0644

protection

rdev

Returns the device identifier (special files only).

reset

Force Module::Generic::Finfo to reload the filesystem information of the underlying file or directory

size

Returns the total size of file, in bytes. Same as "csize"

stat

Provided with a file path and this returns a new Module::Generic::Finfo object.

user

Returns the numeric user ID of file's owner. Same as "uid"

uid

Returns the numeric user ID of file's owner. Same as "user"

CONSTANTS

FILETYPE_NOFILE

File type constant to indicate the file does not exist.

FILETYPE_REG

Regular file

FILETYPE_DIR

The element is a directory

FILETYPE_CHR

The element is a character block

FILETYPE_BLK

A block device

FILETYPE_PIPE

The file is a FIFO or a pipe

FILETYPE_LNK

The file is a symbolic link

FILETYPE_SOCK

The file is a (unix domain) socket

FILETYPE_UNKFILE

The file is of some other unknown type or the type cannot be determined

SERIALISATION

Serialisation by CBOR, Sereal and Storable::Improved (or the legacy Storable) is supported by this package. To that effect, the following subroutines are implemented: FREEZE, THAW, STORABLE_freeze and STORABLE_thaw

THREAD & PROCESS SAFETY

Module::Generic::Finfo is designed to be fully thread-safe and process-safe, ensuring data integrity across Perl ithreads and mod_perl’s threaded Multi-Processing Modules (MPMs) such as Worker or Event. It leverages Module::Generic::Global for thread-safe management of system-wide state and maintains per-object state for file metadata.

Key considerations for thread and process safety:

  • Shared Variables

    The has_local_tz value in "_datetime", indicating system timezone support, is stored in Module::Generic::Global’s local_tz namespace as a system-wide global (accessed via local_tz = 'system'>). This is shared across all Module::Generic::* modules (e.g., Module::Generic::Finfo, Module::Generic::DateTime) to avoid redundant timezone checks. Access is protected by thread locks, ensuring thread-safe initialization and retrieval:

    use threads;
    my $finfo = Module::Generic::Finfo->new( "example.txt" );
    my $atime = $finfo->atime; # Sets shared has_local_tz
    my $dt = Module::Generic::DateTime->new( DateTime->now );
    $dt->epoch; # Reuses has_local_tz

    The local_tz namespace is highly specific, ensuring negligible risk of collisions with unrelated modules.

  • File Metadata

    File metadata (e.g., "filepath", internal _data array) is stored per-object, ensuring each thread or process operates on its own instance without shared state conflicts:

    use threads;
    my @threads = map
    {
        threads->create(sub
        {
            my $finfo = Module::Generic::Finfo->new( "example.txt" );
            $finfo->size; # Thread-safe, per-object state
            return(1);
        });
    } 1..5;
    $_->join for @threads;
  • External Libraries

    Methods like "mime_type" use File::MMagic::XS or File::MMagic, which are thread-safe as they operate on per-object state. "_datetime" uses DateTime and DateTime::Format::Strptime, both thread-safe when combined with Module::Generic::Global’s locking for has_local_tz.

  • Serialisation

    Serialisation methods ("FREEZE", "THAW") operate on per-object state, making them thread-safe for use with CBOR::XS, Sereal, or Storable::Improved:

    use threads;
    use Sereal;
    my $finfo = Module::Generic::Finfo->new( "example.txt" );
    my $encoded = Sereal::encode_sereal( $finfo ); # Thread-safe
  • mod_perl Considerations

    - Prefork MPM

    File metadata and has_local_tz are isolated per-process, requiring no additional synchronisation. System-level operations (e.g., stat in "reset") are process-safe.

    - Threaded MPMs (Worker/Event)

    The local_tz namespace is protected by Module::Generic::Global’s synchronisation ("lock" in perlfunc for ithreads, APR::ThreadRWLock in rare non-threaded Perl cases). Users should call Module::Generic::Global->cleanup_register in mod_perl handlers to clear shared state after requests, preventing memory leaks.

    - Thread-Unsafe Functions

    Avoid Perl functions like "localtime" in perlfunc or "getgrgid" in perlfunc (used in "group") in threaded MPMs, as they may affect all threads. Module::Generic::Finfo mitigates this by caching results in per-object state. Consult perlthrtut and mod_perl documentation for details.

  • Process Safety

    File operations (e.g., "stat" in perlfunc in "init", "reset") are process-safe, relying on system-level calls. The shared has_local_tz is managed by Module::Generic::Global, ensuring consistent access across processes:

    use POSIX qw( fork );
    my $pid = fork();
    if( $pid == 0 )
    {
        my $finfo = Module::Generic::Finfo->new( "example.txt" );
        $finfo->atime; # Safe, uses shared has_local_tz
        exit;
    }
    waitpid( $pid, 0 );

For debugging in threaded or multi-process environments, use platform-specific commands (e.g., on Linux):

ls -l /proc/$$/fd  # List open file descriptors

See your operating system’s documentation for equivalent commands.

AUTHOR

Jacques Deguest <jack@deguest.jp>

CPAN ID: jdeguest

https://gitlab.com/jackdeguest/Module-Generic

SEE ALSO

Module::Generic::File, Module::Generic

COPYRIGHT & LICENSE

Copyright (c) 2021-2024 DEGUEST Pte. Ltd.

You can use, copy, modify and redistribute this package and associated files under the same terms as Perl itself.