Note on "io"

It takes a single optional argument and determines what type of IO::All subclass object to return. With no arguments it returns an "IO::All" object, which has no I/O methods, but has methods to construct subclass objects like "IO::All::File".

In other words, these 2 statements are usually the same:

    $content = io('file.txt')->all;
    $content = io->file('file.txt')->all;

Use the first form when you are demonstrating your Perl virtues of laziness and impatience, and use the second form when your job is on the line.

File Locking

    use IO::All;
    my $io1 = io('myfile')->lock;
    $io1->println('line 1');

    fork or do {
      my $io2 = io('myfile')->lock;
      print $io2->slurp;
      exit;
    };

    sleep 1;
    $io1->println('line 2');
    $io1->println('line 3');
    $io1->unlock;

There are a lot of subtle things going on here. An exclusive lock is issued for $io1 on the first "println". That's because the file isn't actually opened until the first IO operation.

When the child process tries to read the file using $io2, there is a shared lock put on it. Since $io1 has the exclusive lock, the slurp blocks.

The parent process sleeps just to make sure the child process gets a chance. The parent needs to call "unlock" or "close" to release the lock. If all goes well the child will print 3 lines.

In-place Editing

    > cat > x.txt
    The sexy saxophone,

    got the axe.
    ^d

    > perl -MIO::All -e 'map { s/x/X/g; $_ } @{ io(shift) }' x.txt
    > cat x.txt
    The seXy saXophone,

    got the aXe.

 This one liner uses shift() to grab the file from STDIN and create an io
 object that is dereferenced using @{ } and fed to map() like any perl array
 reference.

Round Robin

    my $io = io 'file1.txt';
    $io->autoclose(1);
    while (my $line = $io->getline || $io->getline) {
      print $line;
    }

Reading Backwards

    my @reversed;
    my $io = io('file1.txt');
    $io->backwards;
    while (my $line = $io->getline) {
      push @reversed, $line;
    }

or more simply:

    my @reversed = io('file1.txt')->backwards->getlines;

The "backwards" method returns the IO::All object so that you can chain the calls.

NOTE: This operation requires that you have the File::ReadBackwards module installed.

Client/Server Sockets

In this example, a server will return 3 lines of text, to every client that calls it. Here is the server code:

    use IO::All;

    my $socket = io(':12345')->fork->accept;
    $socket->print($_) while <DATA>;
    $socket->close;

    __DATA__
    On your mark,
    Get set,
    Go!

Here is the client code:

    use IO::All;

    my $io = io('localhost:12345');
    print while $_ = $io->getline;

You can run the server once, and then run the client repeatedly (in another terminal window). It should print the 3 data lines each time.

Note that it is important to close the socket if the server is forking, or else the socket won't go out of scope and close.

A Tiny Web Server

    perl -MIO::All -e 'io(":8080")->fork->accept->(sub { $_[0] < io(-x $1 ? "./$1 |" : $1) if /^GET \/(.*) / })'

There is are a lot of subtle things going on here. First we accept a socket and fork the server. Then we overload the new socket as a code ref. This code ref takes one argument, another code ref, which is used as a callback.

The callback is called once for every line read on the socket. The line is put into $_ and the socket itself is passed in to the callback.

Our callback is scanning the line in $_ for an HTTP GET request. If one is found it parses the file name into $1. Then we use $1 to create an new IO::All file object... with a twist. If the file is executable ("-x"), then we create a piped command as our IO::All object. This somewhat approximates CGI support.

Whatever the resulting object is, we direct the contents back at our socket which is in $_[0]. Pretty simple, eh?

DBM Files

    io("names.db")->{ingy} > io('=');

Since their are several DBM formats available in Perl, IO::All picks the first one of these that is installed on your system:

    DB_File GDBM_File NDBM_File ODBM_File SDBM_File

You can override which DBM you want for each IO::All object:

    my @keys = keys %{io('mydbm')->dbm('SDBM_File')};

File Subclassing

    package NewModule;
    use IO::All -base;

You need to do it this way so that IO::All will export the "io" function. Here is a simple recipe for subclassing:

IO::Dumper inherits everything from IO::All and adds an extra method called "dump", which will dump a data structure to the file we specify in the "io" function. Since it needs Data::Dumper to do the dumping, we override the "open" method to "require Data::Dumper" and then pass control to the real "open".

First the code using the module:

    use IO::Dumper;

    io('./mydump')->dump($hash);

And next the IO::Dumper module itself:

    package IO::Dumper;
    use IO::All -base;
    use Data::Dumper;

    sub dump {
      my $self = shift;
      Dumper(@_) > $self;
    }

    1;

Inline Subclassing

    IO::Dumper->import;
    io('./mydump')->dump($hash);

    package IO::Dumper;
    use IO::All -base;
    use Data::Dumper;

    sub dump {
      my $self = shift;
      Dumper(@_) > $self;
    }

Object Construction and Initialization Methods

new

There are three ways to create a new IO::All object. The first is with the special function "io" which really just calls "IO::All->new". The second is by calling "new" as a class method. The third is calling "new" as an object instance method. In this final case, the new objects attributes are copied from the instance object.

    io(file-descriptor);
    IO::All->new(file-descriptor);
    $io->new(file-descriptor);
    

All three forms take a single argument, a file descriptor. A file descriptor can be any of the following:

    - A file name
    - A file handle
    - A directory name
    - A directory handle
    - A typeglob reference
    - A piped shell command. eg '| ls -al'
    - A socket domain/port.  eg 'perl.com:5678'
    - '-' means STDIN or STDOUT (depending on usage)
    - '=' means STDERR
    - '$' means an in memory filehandle object
    - '?' means a temporary file
    - A URI including: http, https, ftp and mailto
    - An IO::All object
    

If you provide an IO::All object, you will simply get that same object returned from the constructor.

If no file descriptor is provided, an object will still be created, but it must be defined by one of the following methods before it can be used for I/O:

file

    io->file("path/to/my/file.txt");
    

Using the "file" method sets the type of the object to file and sets the pathname of the file if provided.

It might be important to use this method if you had a file whose name was '- ', or if the name might otherwise be confused with a directory or a socket. In this case, either of these statements would work the same:

    my $file = io('-')->file;
    my $file = io->file('-');
    

dir

    io->dir($dir_name);
    

Make the object be of type directory.

socket

    io->socket("${domain}:${port}");
    

Make the object be of type socket.

link

    io->link($link_name);
    

Make the object be of type link.

pipe

    io->pipe($pipe_command);
    

Make the object be of type pipe. The following three statements are equivalent:

    my $io = io('ls -l |');
    my $io = io('ls -l')->pipe;
    my $io = io->pipe('ls -l');
    

dbm

This method takes the names of zero or more DBM modules. The first one that is available is used to process the dbm file.

    io('mydbm')->dbm('NDBM_File', 'SDBM_File')->{author} = 'ingy';
    

If no module names are provided, the first available of the following is used:

    DB_File GDBM_File NDBM_File ODBM_File SDBM_File
    

mldbm

Similar to the "dbm" method, except create a Multi Level DBM object using the MLDBM module.

This method takes the names of zero or more DBM modules and an optional serialization module. The first DBM module that is available is used to process the MLDBM file. The serialization module can be Data::Dumper, Storable or FreezeThaw.

    io('mymldbm')->mldbm('GDBM_File', 'Storable')->{author} =
      {nickname => 'ingy'};
    

string

Make the object be an in memory filehandle. These are equivalent:

    my $io = io('$');
    my $io = io->string;
    

temp

Make the object represent a temporary file. It will automatically be open for both read and write.

stdio

Make the object represent either STDIN or STDOUT depending on how it is used subsequently. These are equivalent:

    my $io = io('-');
    my $io = io->stdin;
    

stdin

Make the object represent STDIN.

stdout

Make the object represent STDOUT.

stderr

Make the object represent STDERR.

handle

    io->handle($io_handle);
    

Forces the object to be created from an pre-existing IO handle. You can chain calls together to indicate the type of handle:

    my $file_object = io->file->handle($file_handle);
    my $dir_object = io->dir->handle($dir_handle);
    

http

Make the object represent an HTTP URI. Requires IO-All-LWP.

https

Make the object represent an HTTPS URI. Requires IO-All-LWP.

ftp

Make the object represent an FTP URI. Requires IO-All-LWP.

mailto

Make the object represent a "mailto:" URI. Requires IO-All-Mailto.

If you need to use the same options to create a lot of objects, and don't want to duplicate the code, just create a dummy object with the options you want, and use that object to spawn other objects.

    my $lt = io->lock->tie;
    ...
    my $io1 = $lt->new('file1');
    my $io2 = $lt->new('file2');

Since the new method copies attributes from the calling object, both $io1 and $io2 will be locked and tied.

Option Setting Methods

Each option can take a single argument of 0 or 1. If no argument is given, the value 1 is assumed. Passing 0 turns the option off.

All of these options return the object reference that was used to invoke them. This is so that the option methods can be chained together. For example:

    my $io = io('path/file')->tie->assert->chomp->lock;

absolute

Indicates that the "pathname" for the object should be made absolute.

    # Print the full path of the current working directory
    # (like pwd).

    use IO::All;

    print io->curdir->absolute;
    

assert

This method ensures that the path for a file or directory actually exists before the file is open. If the path does not exist, it is created.

For example, here is a program called "create-cat-to" that outputs to a file that it creates.

    #!/usr/bin/perl

    # create-cat-to.pl
    # cat to a file that can be created.

    use strict;
    use warnings;

    use IO::All;

    my $filename = shift(@ARGV);

    # Create a file called $filename, including all leading components.
    io('-') > io->file($filename)->assert;
    

Here's an example use of it:

    $ ls -l
    total 0
    $ echo "Hello World" | create-cat-to one/two/three/four.txt
    $ ls -l
    total 4
    drwxr-xr-x 3 shlomif shlomif 4096 2010-10-14 18:03 one/
    $ cat one/two/three/four.txt
    Hello World
    $
    

autoclose

By default, IO::All will close an object opened for input when EOF is reached. By closing the handle early, one can immediately do other operations on the object without first having to close it.

This option is on by default, so if you don't want this behaviour, say so like this:

    $io->autoclose(0);
    

The object will then be closed when $io goes out of scope, or you manually call "$io->close".

autoflush

Proxy for IO::Handle::autoflush

backwards

Sets the object to 'backwards' mode. All subsequent "getline" operations will read backwards from the end of the file.

Requires the File::ReadBackwards CPAN module.

binary

Adds ":raw" to the list of PerlIO layers applied after "open", and applies it immediately on an open handle.

chdir

chdir() to the pathname of a directory object. When object goes out of scope, chdir back to starting directory.

chomp

Indicates that all operations that read lines should chomp the lines. If the "separator" method has been called, chomp will remove that value from the end of each record.

Note that "chomp" may cause the following idiom to halt prematurely (e.g., if "separator" is "\n" (the default) and "chomp" is in effect, then this command will stop reading at the first blank line):

    while ( my $line = $io->getline ) {...}
    

Try the following instead:

    while ( defined(my $line = $io->getline) ) {...}
    

confess

Errors should be reported with the very detailed Carp::confess function.

deep

Indicates that calls to the "all" family of methods should search directories as deep as possible.

fork

Indicates that the process should automatically be forked inside the "accept" socket method.

lock

Indicate that operations on an object should be locked using flock.

rdonly

This option indicates that certain operations like DBM and Tie::File access should be done in read-only mode.

rdwr

This option indicates that DBM and MLDBM files should be opened in read/write mode.

relative

Indicates that the "pathname" for the object should be made relative. If passed an argument, path will be made relative to passed argument.

sort

Indicates whether objects returned from one of the "all" methods will be in sorted order by name. True by default.

tie

Indicate that the object should be tied to itself, thus allowing it to be used as a filehandle in any of Perl's builtin IO operations.

    my $io = io('foo')->tie;
    @lines = <$io>;
    

utf8

Adds ":encoding(UTF-8)" to the list of PerlIO layers applied after "open", and applies it immediately on an open handle.

Configuration Methods

If these methods are passed no argument, they will return their current value. If arguments are passed they will be used to set the current value, and the object reference will be returned for potential method chaining.

bcc

Set the Bcc field for a mailto object.

binmode

Adds the specified layer to the list of PerlIO layers applied after "open", and applies it immediately on an open handle. Does a bare "binmode" when called without argument.

block_size

The default length to be used for "read" and "sysread" calls. Defaults to 1024.

buffer

Returns a reference to the internal buffer, which is a scalar. You can use this method to set the buffer to a scalar of your choice. (You can just pass in the scalar, rather than a reference to it.)

This is the buffer that "read" and "write" will use by default.

You can easily have IO::All objects use the same buffer:

    my $input = io('abc');
    my $output = io('xyz');
    my $buffer;
    $output->buffer($input->buffer($buffer));
    $output->write while $input->read;
    

cc

Set the Cc field for a mailto object.

content

Get or set the content for an LWP operation manually.

domain

Set the domain name or ip address that a socket should use.

encoding

Adds the specified encoding to the list of PerlIO layers applied after "open", and applies it immediately on an open handle. Requires an argument.

errors

Use this to set a subroutine reference that gets called when an internal error is thrown.

filter

Use this to set a subroutine reference that will be used to grep which objects get returned on a call to one of the "all" methods. For example:

    my @odd = io->curdir->filter(sub {$_->size % 2})->All_Files;
    

@odd will contain all the files under the current directory whose size is an odd number of bytes.

from

Indicate the sender for a mailto object.

mailer

Set the mailer program for a mailto transaction. Defaults to 'sendmail'.

mode

Set the mode for which the file should be opened. Examples:

    $io->mode('>>')->open;
    $io->mode(O_RDONLY);

    my $log_appender = io->file('/var/log/my-application.log')
                         ->mode('>>')->open();

    $log_appender->print("Stardate 5987.6: Mission accomplished.");
    

name

Set or get the name of the file or directory represented by the IO::All object.

password

Set the password for an LWP transaction.

perms

Sets the permissions to be used if the file/directory needs to be created.

port

Set the port number that a socket should use.

request

Manually specify the request object for an LWP transaction.

response

Returns the resulting response object from an LWP transaction.

separator

Sets the record (line) separator to whatever value you pass it. Default is "\n". Affects the chomp setting too.

string_ref

Returns a reference to the internal string that is acting like a file.

subject

Set the subject for a mailto transaction.

to

Set the recipient address for a mailto request.

uri

Direct access to the URI used in LWP transactions.

user

Set the user name for an LWP transaction.

IO Action Methods

accept

For sockets. Opens a server socket (LISTEN => 1, REUSE => 1). Returns an IO::All socket object that you are listening on.

If the "fork" method was called on the object, the process will automatically be forked for every connection.

all

Read all contents into a single string.

    compare(io('file1')->all, io('file2')->all);
    

all (For directories)

Returns a list of IO::All objects for all files and subdirectories in a directory.

'.' and '..' are excluded.

Takes an optional argument telling how many directories deep to search. The default is 1. Zero (0) means search as deep as possible.

The filter method can be used to limit the results.

The items returned are sorted by name unless "->sort(0)" is used.

All

Same as all(0).

all_dirs

Same as "all", but only return directories.

All_Dirs

Same as all_dirs(0).

all_files

Same as "all", but only return files.

All_Files

Same as all_files(0).

all_links

Same as "all", but only return links.

All_Links

Same as all_links(0).

append

Same as print, but sets the file mode to '>>'.

appendf

Same as printf, but sets the file mode to '>>'.

appendln

Same as println, but sets the file mode to '>>'.

clear

Clear the internal buffer. This method is called by "write" after it writes the buffer. Returns the object reference for chaining.

close

Close will basically unopen the object, which has different meanings for different objects. For files and directories it will close and release the handle. For sockets it calls shutdown. For tied things it unties them, and it unlocks locked things.

copy

Copies the object to the path passed. Works on both files and directories, but directories require "File::Copy::Recursive" to be installed.

empty

Returns true if a file exists but has no size, or if a directory exists but has no contents.

eof

Proxy for IO::Handle::eof

ext

Returns the extension of the file. Can also be spelled as "extension"

exists

Returns whether or not the file or directory exists.

filename

Return the name portion of the file path in the object. For example:

    io('my/path/file.txt')->filename;
    

would return "file.txt".

fileno

Proxy for IO::Handle::fileno

filepath

Return the path portion of the file path in the object. For example:

    io('my/path/file.txt')->filepath;
    

would return "my/path".

get

Perform an LWP GET request manually.

getc

Proxy for IO::Handle::getc

getline

Calls IO::File::getline. You can pass in an optional record separator.

getlines

Calls IO::File::getlines. You can pass in an optional record separator.

glob

Creates IO::All objects for the files matching the glob in the IO::All::Dir. For example:

    io->dir($ENV{HOME})->glob('*.txt')
    

head

Return the first 10 lines of a file. Takes an optional argument which is the number of lines to return. Works as expected in list and scalar context. Is subject to the current line separator.

io_handle

Direct access to the actual IO::Handle object being used on an opened IO::All object.

is_dir

Returns boolean telling whether or not the IO::All object represents a directory.

is_executable

Returns true if file or directory is executable.

is_dbm

Returns boolean telling whether or not the IO::All object represents a dbm file.

is_file

Returns boolean telling whether or not the IO::All object represents a file.

is_link

Returns boolean telling whether or not the IO::All object represents a symlink.

is_mldbm

Returns boolean telling whether or not the IO::All object represents a mldbm file.

is_open

Indicates whether the IO::All is currently open for input/output.

is_pipe

Returns boolean telling whether or not the IO::All object represents a pipe operation.

is_readable

Returns true if file or directory is readable.

is_socket

Returns boolean telling whether or not the IO::All object represents a socket.

is_stdio

Returns boolean telling whether or not the IO::All object represents a STDIO file handle.

is_string

Returns boolean telling whether or not the IO::All object represents an in memory filehandle.

is_temp

Returns boolean telling whether or not the IO::All object represents a temporary file.

is_writable

Returns true if file or directory is writable. Can also be spelled as "is_writeable".

length

Return the length of the internal buffer.

mimetype

Return the mimetype of the file.

Requires a working installation of the File::MimeInfo CPAN module.

mkdir

Create the directory represented by the object.

mkpath

Create the directory represented by the object, when the path contains more than one directory that doesn't exist. Proxy for File::Path::mkpath.

next

For a directory, this will return a new IO::All object for each file or subdirectory in the directory. Return undef on EOD.

open

Open the IO::All object. Takes two optional arguments "mode" and "perms", which can also be set ahead of time using the "mode" and "perms" methods.

NOTE: Normally you won't need to call open (or mode/perms), since this happens automatically for most operations.

os

Change the object's os representation. Valid options are: "win32", "unix", "vms", "mac", "os2".

pathname

Return the absolute or relative pathname for a file or directory, depending on whether object is in "absolute" or "relative" mode.

print

Proxy for IO::Handle::print

printf

Proxy for IO::Handle::printf

println

Same as print, but adds newline to each argument unless it already ends with one.

put

Perform an LWP PUT request manually.

read

This method varies depending on its context. Read carefully (no pun intended).

For a file, this will proxy IO::File::read. This means you must pass it a buffer, a length to read, and optionally a buffer offset for where to put the data that is read. The function returns the length actually read (which is zero at EOF).

If you don't pass any arguments for a file, IO::All will use its own internal buffer, a default length, and the offset will always point at the end of the buffer. The buffer can be accessed with the "buffer" method. The length can be set with the "block_size" method. The default length is 1024 bytes. The "clear" method can be called to clear the buffer.

For a directory, this will proxy IO::Dir::read.

readdir

Similar to the Perl "readdir" builtin. In scalar context, return the next directory entry (ie file or directory name), or undef on end of directory. In list context, return all directory entries.

Note that "readdir" does not return the special "." and ".." entries.

readline

Same as "getline".

readlink

Calls Perl's readlink function on the link represented by the object. Instead of returning the file path, it returns a new IO::All object using the file path.

recv

Proxy for IO::Socket::recv

rename

    my $new = $io->rename('new-name');
    

Calls Perl's rename function and returns an IO::All object for the renamed file. Returns false if the rename failed.

rewind

Proxy for IO::Dir::rewind

rmdir

Delete the directory represented by the IO::All object.

rmtree

Delete the directory represented by the IO::All object and all the files and directories beneath it. Proxy for File::Path::rmtree.

scalar

Deprecated. Same as "all()".

seek

Proxy for IO::Handle::seek. If you use seek on an unopened file, it will be opened for both read and write.

send

Proxy for IO::Socket::send

shutdown

Proxy for IO::Socket::shutdown

slurp

Read all file content in one operation. Returns the file content as a string. In list context returns every line in the file.

stat

Proxy for IO::Handle::stat

sysread

Proxy for IO::Handle::sysread

syswrite

Proxy for IO::Handle::syswrite

tail

Return the last 10 lines of a file. Takes an optional argument which is the number of lines to return. Works as expected in list and scalar context. Is subject to the current line separator.

tell

Proxy for IO::Handle::tell

throw

This is an internal method that gets called whenever there is an error. It could be useful to override it in a subclass, to provide more control in error handling.

touch

Update the atime and mtime values for a file or directory. Creates an empty file if the file does not exist.

truncate

Proxy for IO::Handle::truncate

type

Returns a string indicated the type of io object. Possible values are:

    file
    dir
    link
    socket
    string
    pipe
    

Returns undef if type is not determinable.

unlink

Unlink (delete) the file represented by the IO::All object.

NOTE: You can unlink a file after it is open, and continue using it until it is closed.

unlock

Release a lock from an object that used the "lock" method.

utime

Proxy for the utime Perl function.

write

Opposite of "read" for file operations only.

NOTE: When used with the automatic internal buffer, "write" will clear the buffer after writing it.

Stat Methods

atime

Last access time in seconds since the epoch

blksize

Preferred block size for file system I/O

blocks

Actual number of blocks allocated

ctime

Inode change time in seconds since the epoch

device

Device number of filesystem

device_id

Device identifier for special files only

gid

Numeric group id of file's owner

inode

Inode number

modes

File mode - type and permissions

mtime

Last modify time in seconds since the epoch

nlink

Number of hard links to the file

size

Total size of file in bytes

uid

Numeric user id of file's owner

File::Spec Methods

abs2rel

Returns the relative path for the absolute path in the IO::All object. Can take an optional argument indicating the base path.

canonpath

Returns the canonical path for the IO::All object. The canonical path is the fully resolved path if the file exists, so any symlinks will be resolved.

case_tolerant

Returns 0 or 1 indicating whether the file system is case tolerant. Since an active IO::All object is not needed for this function, you can code it like:

    IO::All->case_tolerant;
    

or more simply:

    io->case_tolerant;
    

catdir

Concatenate the directory components together, and return a new IO::All object representing the resulting directory.

catfile

Concatenate the directory and file components together, and return a new IO::All object representing the resulting file.

    my $contents = io->catfile(qw(dir subdir file))->slurp;
    

This is a very portable way to read "dir/subdir/file".

catpath

Concatenate the volume, directory and file components together, and return a new IO::All object representing the resulting file.

curdir

Returns an IO::All object representing the current directory.

devnull

Returns an IO::All object representing the "/dev/null" file.

is_absolute

Returns 0 or 1 indicating whether the "name" field of the IO::All object is an absolute path.

join

Same as "catfile".

path

Returns a list of IO::All directory objects for each directory in your path.

rel2abs

Returns the absolute path for the relative path in the IO::All object. Can take an optional argument indicating the base path.

rootdir

Returns an IO::All object representing the root directory on your file system.

splitdir

Returns a list of the directory components of a path in an IO::All object.

splitpath

Returns a volume directory and file component of a path in an IO::All object.

tmpdir

Returns an IO::All object representing a temporary directory on your file system.

updir

Returns an IO::All object representing the current parent directory.

Known Bugs and Deficiencies

If you call a method that does not make sense for a particular object, the result probably won't make sense. Little attempt is made to check for improper usage.