package Catmandu::Util;
use Catmandu::Sane;
use Exporter qw(import);
use Sub::Quote ();
use Scalar::Util ();
use Data::Util ();
use List::Util ();
use Data::Compare ();
use IO::File;
use IO::Handle::Util ();
use File::Spec;
use YAML::Any ();
use JSON ();
our %EXPORT_TAGS = (
io => [qw(io read_file write_file read_yaml read_json join_path
normalize_path segmented_path)],
data => [qw(parse_data_path get_data set_data delete_data data_at)],
array => [qw(array_exists array_group_by array_pluck array_to_sentence
array_sum array_includes array_any array_rest array_uniq)],
string => [qw(as_utf8 trim capitalize)],
is => [qw(is_same is_different)],
check => [qw(check_same check_different)],
human => [qw(human_number human_content_type human_byte_size)],
xml => [qw(xml_declaration xml_escape)],
misc => [qw(require_package use_lib)],
);
our @EXPORT_OK = map { @$_ } values %EXPORT_TAGS;
$EXPORT_TAGS{all} = \@EXPORT_OK;
my $HUMAN_CONTENT_TYPES = {
# txt
'text/plain' => 'Text',
'application/txt' => 'Text',
# pdf
'application/pdf' => 'PDF',
'application/x-pdf' => 'PDF',
'application/acrobat' => 'PDF',
'applications/vnd.pdf' => 'PDF',
'text/pdf' => 'PDF',
'text/x-pdf' => 'PDF',
# doc
'application/doc' => 'Word',
'application/vnd.msword' => 'Word',
'application/vnd.ms-word' => 'Word',
'application/winword' => 'Word',
'application/word' => 'Word',
'application/x-msw6' => 'Word',
'application/x-msword' => 'Word',
# docx
'application/vnd.openxmlformats-officedocument.wordprocessingml.document' => 'Word',
# xls
'application/vnd.ms-excel' => 'Excel',
'application/msexcel' => 'Excel',
'application/x-msexcel' => 'Excel',
'application/x-ms-excel' => 'Excel',
'application/vnd.ms-excel' => 'Excel',
'application/x-excel' => 'Excel',
'application/x-dos_ms_excel' => 'Excel',
'application/xls' => 'Excel',
# xlsx
'application/vnd.openxmlformats-officedocument.spreadsheetml.sheet' => 'Excel',
# ppt
'application/vnd.ms-powerpoint' => 'PowerPoint',
'application/mspowerpoint' => 'PowerPoint',
'application/ms-powerpoint' => 'PowerPoint',
'application/mspowerpnt' => 'PowerPoint',
'application/vnd-mspowerpoint' => 'PowerPoint',
'application/powerpoint' => 'PowerPoint',
'application/x-powerpoint' => 'PowerPoint',
# pptx
'application/vnd.openxmlformats-officedocument.presentationml.presentation' => 'PowerPoint',
# csv
'text/comma-separated-values' => 'CSV',
'text/csv' => 'CSV',
'application/csv' => 'CSV',
# zip
'application/zip' => 'ZIP archive',
};
my $XML_DECLARATION = qq(<?xml version="1.0" encoding="UTF-8"?>\n);
sub TIESCALAR { };
sub io {
my ($arg, %opts) = @_;
my $binmode = $opts{binmode} || $opts{encoding} || ':utf8';
my $mode = $opts{mode} || 'r';
my $io;
if (is_scalar_ref($arg)) {
$io = IO::Handle::Util::io_from_scalar_ref($arg);
binmode $io, $binmode;
} elsif (is_glob_ref(\$arg) || is_glob_ref($arg)) {
$io = IO::Handle->new_from_fd($arg, $mode);
binmode $io, $binmode;
} elsif (is_string($arg)) {
$io = IO::File->new($arg, $mode);
binmode $io, $binmode;
} elsif (is_code_ref($arg) && $mode eq 'r') {
$io = IO::Handle::Util::io_from_getline($arg);
} elsif (is_code_ref($arg) && $mode eq 'w') {
$io = IO::Handle::Util::io_from_write_cb($arg);
} elsif (is_instance($arg, 'IO::Handle')) {
$io = $arg;
binmode $io, $binmode;
} else {
confess "can't make io from argument";
}
$io;
}
sub read_file {
my ($path) = @_;
local $/;
open my $fh, "<", $path or confess qq(can't open "$path" for reading);
my $str = <$fh>;
close $fh;
$str;
}
sub write_file {
my ($path, $str) = @_;
open my $fh, ">", $path or confess qq(can't open "$path" for writing);
print $fh $str;
close $fh;
$path;
}
sub read_yaml {
YAML::Any::LoadFile($_[0]);
}
sub read_json {
JSON::decode_json(read_file($_[0]));
}
sub join_path {
my $path = File::Spec->catfile(@_);
normalize_path($path);
}
sub normalize_path { # taken from Dancer::FileUtils
my ($path) = @_;
$path =~ s!/\./!/!g;
while ($path =~ s![^/]*/\.\./!!) {}
$path;
}
sub segmented_path {
my ($id, %opts) = @_;
my $segment_size = $opts{segment_size} || 3;
my $base_path = $opts{base_path};
$id =~ s/[^0-9a-zA-Z]+//g;
my @path = unpack "(A$segment_size)*", $id;
defined $base_path
? File::Spec->catdir($base_path, @path)
: File::Spec->catdir(@path);
}
sub parse_data_path {
my ($path) = @_;
check_string($path);
$path = [ split /[\/\.]/, $path ];
my $key = pop @$path;
return $path, $key;
}
sub get_data {
my ($data, $key) = @_;
if (is_array_ref($data)) {
given ($key) {
when ('$first') { return unless @$data; $key = 0 }
when ('$last') { return unless @$data; $key = @$data - 1 }
when ('*') { return @$data }
}
if (array_exists($data, $key)) {
return $data->[$key];
}
return;
}
if (is_hash_ref($data) && exists $data->{$key}) {
return $data->{$key};
}
return;
}
sub set_data {
my ($data, $key, @vals) = @_;
return unless @vals;
if (is_array_ref($data)) {
given ($key) {
when ('$first') { return unless @$data; $key = 0 }
when ('$last') { return unless @$data; $key = @$data - 1 }
when ('$prepend') { unshift @$data, $vals[0]; return $vals[0] }
when ('$append') { push @$data, $vals[0]; return $vals[0] }
when ('*') { return splice @$data, 0, @$data, @vals }
}
return $data->[$key] = $vals[0] if is_natural($key);
return;
}
if (is_hash_ref($data)) {
return $data->{$key} = $vals[0];
}
return;
}
sub delete_data {
my ($data, $key) = @_;
if (is_array_ref($data)) {
given ($key) {
when ('$first') { return unless @$data; $key = 0 }
when ('$last') { return unless @$data; $key = @$data - 1 }
when ('*') { return splice @$data, 0, @$data }
}
if (array_exists($data, $key)) {
return splice @$data, $key, 1;
}
return;
}
if (is_hash_ref($data) && exists $data->{$key}) {
return delete $data->{$key};
}
return;
}
sub data_at {
my ($path, $data, %opts) = @_;
$path = [@$path];
my $create = $opts{create};
my $_key = $opts{_key} // $opts{key};
if (defined $opts{key} && $create && @$path) {
push @$path, $_key;
}
my $key;
while (defined(my $key = shift @$path)) {
ref $data || return;
if (is_array_ref($data)) {
if ($key eq '*') {
return map { data_at($path, $_, create => $create, _key => $_key) } @$data;
} else {
given ($key) {
when ('$first') { $key = 0 }
when ('$last') { $key = -1 }
when ('$prepend') { unshift @$data, undef; $key = 0 }
when ('$append') { $key = @$data }
}
is_integer($key) || return;
if ($create && @$path) {
$data = $data->[$key] ||= is_integer($path->[0]) || ord($path->[0]) == ord('$') ? [] : {};
} else {
$data = $data->[$key];
}
}
} elsif ($create && @$path) {
$data = $data->{$key} ||= is_integer($path->[0]) || ord($path->[0]) == ord('$') ? [] : {};
} else {
$data = $data->{$key};
}
if ($create && @$path == 1) {
last;
}
}
$data;
}
sub array_exists {
my ($arr, $i) = @_;
is_natural($i) && $i < @$arr;
}
sub array_group_by {
my ($arr, $key) = @_;
List::Util::reduce { my $k = $b->{$key}; push @{$a->{$k} ||= []}, $b if defined $k; $a } {}, @$arr;
}
sub array_pluck {
my ($arr, $key) = @_;
my @vals = map { $_->{$key} } @$arr;
\@vals;
}
sub array_to_sentence {
my ($arr, $join, $join_last) = @_;
$join //= ', ';
$join_last //= ' and ';
my $size = scalar @$arr;
$size > 2
? join($join_last, join($join, @$arr[0..$size-1]), $arr->[-1])
: join($join_last, @$arr);
}
sub array_sum {
List::Util::sum(0, @{$_[0]});
}
sub array_includes {
my ($arr, $val) = @_;
is_same($val, $_) && return 1 for @$arr;
0;
}
sub array_any {
my ($arr, $sub) = @_;
$sub->($_) && return 1 for @$arr;
0;
}
sub array_rest {
my ($arr) = @_;
@$arr < 2 ? [] : [@$arr[1..(@$arr-1)]];
}
sub array_uniq {
my ($arr) = @_;
my %seen = ();
my @vals = grep { not $seen{$_}++ } @$arr;
\@vals;
}
sub as_utf8 {
my $str = $_[0];
utf8::upgrade($str);
$str;
}
sub trim {
my $str = $_[0];
if ($str) {
$str =~ s/^\s+//s;
$str =~ s/\s+$//s;
}
$str;
}
sub capitalize {
my $str = $_[0];
utf8::upgrade($str);
ucfirst lc $str;
}
*is_same = \&Data::Compare::Compare;
sub is_different {
!is_same(@_);
}
sub check_same {
is_same(@_) || confess('error: should be same');
}
sub check_different {
!is_same(@_) || confess('error: should be different');
}
*is_invocant = \&Data::Util::is_invocant;
*is_scalar_ref = \&Data::Util::is_scalar_ref;
*is_array_ref = \&Data::Util::is_array_ref;
*is_hash_ref = \&Data::Util::is_hash_ref;
*is_code_ref = \&Data::Util::is_code_ref;
*is_regex_ref = \&Data::Util::is_rx;
*is_glob_ref = \&Data::Util::is_glob_ref;
*is_value = \&Data::Util::is_value;
*is_string = \&Data::Util::is_string;
*is_number = \&Data::Util::is_number;
*is_integer = \&Data::Util::is_integer;
sub is_natural {
is_integer($_[0]) && $_[0] >= 0;
}
sub is_positive {
is_integer($_[0]) && $_[0] >= 1;
}
sub is_ref {
ref $_[0] ? 1 : 0;
}
sub is_able {
my $obj = shift;
is_invocant($obj) || return 0;
$obj->can($_) || return 0 for @_;
1;
}
sub check_able {
my $obj = shift;
return $obj if is_able($obj, @_);
confess('type error: should be able to '.array_to_sentence(\@_));
}
sub check_maybe_able {
my $obj = shift;
return $obj if is_maybe_able($obj, @_);
confess('type error: should be undef or able to '.array_to_sentence(\@_));
}
sub is_instance {
my $obj = shift;
Scalar::Util::blessed($obj) || return 0;
$obj->can($_) || return 0 for @_;
1;
}
sub check_instance {
my $obj = shift;
return $obj if is_able($obj, @_);
confess('type error: should be instance of '.array_to_sentence(\@_));
}
sub check_maybe_instance {
my $obj = shift;
return $obj if is_maybe_able($obj, @_);
confess('type error: should be undef or instance of '.array_to_sentence(\@_));
}
for my $sym (qw(able instance invocant ref
scalar_ref array_ref hash_ref code_ref regex_ref glob_ref
value string number integer natural positive)) {
my $pkg = __PACKAGE__;
my $err_name = $sym;
$err_name =~ s/_/ /;
push @EXPORT_OK, "is_$sym", "is_maybe_$sym", "check_$sym", "check_maybe_$sym";
push @{$EXPORT_TAGS{is}}, "is_$sym", "is_maybe_$sym";
push @{$EXPORT_TAGS{check}}, "check_$sym", "check_maybe_$sym";
Sub::Quote::quote_sub("${pkg}::is_maybe_$sym",
"!defined(\$_[0]) || ${pkg}::is_$sym(\@_)")
unless Data::Util::get_code_ref($pkg, "is_maybe_$sym");
Sub::Quote::quote_sub("${pkg}::check_$sym",
"${pkg}::is_$sym(\@_) || ${pkg}::confess('type error: should be $err_name'); \$_[0]")
unless Data::Util::get_code_ref($pkg, "check_$sym");
Sub::Quote::quote_sub("${pkg}::check_maybe_$sym",
"${pkg}::is_maybe_$sym(\@_) || ${pkg}::confess('type error: should be undef or $err_name'); \$_[0]")
unless Data::Util::get_code_ref($pkg, "check_maybe_$sym");
}
sub human_number { # taken from Number::Format
my $num = $_[0];
# add leading 0's so length($num) is divisible by 3
$num = '0'x(3 - (length($num) % 3)).$num;
# split $num into groups of 3 characters and insert commas
$num = join ',', grep { $_ ne '' } split /(...)/, $num;
# strip off leading zeroes and/or comma
$num =~ s/^0+,?//;
length $num ? $num : '0';
}
sub human_byte_size {
my ($size) = @_;
if ($size > 1000000000) {
return sprintf("%.2f GB", $size / 1000000000);
} elsif ($size > 1000000) {
return sprintf("%.2f MB", $size / 1000000);
} elsif ($size > 1000) {
return sprintf("%.2f KB", $size / 1000);
}
"$size bytes";
}
sub human_content_type {
my ($content_type, $default) = @_;
my ($key) = $content_type =~ /^([^;]+)/;
$HUMAN_CONTENT_TYPES->{$key} // $default // $content_type;
}
sub xml_declaration {
$XML_DECLARATION;
}
sub xml_escape {
my ($str) = @_;
utf8::upgrade($str);
$str =~ s/&/&/go;
$str =~ s/</</go;
$str =~ s/>/>/go;
$str =~ s/"/"/go;
$str =~ s/'/'/go;
# remove control chars
$str =~ s/[^\x09\x0A\x0D\x20-\x{D7FF}\x{E000}-\x{FFFD}\x{10000}-\x{10FFFF}]//go;
$str;
}
sub use_lib {
my (@dirs) = @_;
use lib;
lib->import(@dirs);
confess $@ if $@;
1;
}
sub require_package {
my ($pkg, $ns) = @_;
if ($ns) {
unless ($pkg =~ s/^\+// || $pkg =~ /^$ns/) {
$pkg = "${ns}::$pkg";
}
}
return $pkg if is_invocant($pkg);
eval "require $pkg;1;" or do {
my $err = $@;
confess $err;
};
$pkg;
}
1;
__END__
=head1 NAME
Catmandu::Util - A collection of utility functions
=head1 SYNOPSIS
use Catmandu::Util qw(:string);
$str = trim($str);
=head1 FUNCTIONS
=head2 IO functions
use Catmandu::Util qw(:io);
=over 4
=item io($io, %opts)
Takes a file path, glob, glob reference, scalar reference or L<IO::Handle>
object and returns an opened L<IO::Handle> object.
my $fh = io '/path/to/file';
my $fh = io *STDIN;
my $fh = io \*STDOUT, mode => 'w', binmode => ':crlf';
my $write_cb = sub { my $str = $_[0]; ... };
my $fh = io $write_cb, mode => 'w';
my $scalar = "";
my $fh = io \$scalar, mode => 'w';
$fh->print("some text");
Options are:
=over 12
=item mode
Default is C<"r">.
=item binmode
Default is C<":utf8">.
=item encoding
Alias for C<binmode>.
=back
=item read_file($path);
Reads the file at C<$path> into a string.
my $str = read_file('/path/to/file.txt');
=item write_file($path, $str);
Writes the string C<$str> to a file at C<$path>.
write_file('/path/to/file.txt', "contents");
=item read_yaml($path);
my $cfg = read_yaml('config.yaml');
=item read_json($path);
my $cfg = read_json('config.json');
=item join_path(@path);
join_path('/path/..', './to', 'file.txt');
# => "/to/file.txt"
=item normalize_path($path);
normalize_path('/path/../to/./file.txt');
# => "/to/file.txt"
=item segmented_path($path);
my $id = "FB41144C-F0ED-11E1-A9DE-61C894A0A6B4";
segmented_path($id, segment_size => 4);
# => "FB41/144C/F0ED/11E1/A9DE/61C8/94A0/A6B4"
segmented_path($id, segment_size => 2, base_path => "/files");
# => "/files/FB/41/14/4C/F0/ED/11/E1/A9/DE/61/C8/94/A0/A6/B4"
=back
=head2 Array functions
use Catmandu::Util qw(:array);
A collection of functions that operate on array references.
=over 4
=item array_exists($array, $index)
Returns C<1> if C<$index> is in the bounds of C<$array>
array_exists(["a", "b"], 2);
# => 0
array_exists(["a", "b"], 1);
# => 1
=item array_group_by($array, $key)
my $list = [{color => 'black', id => 1},
{color => 'white', id => 2},
{id => 3},
{color => 'black', id => 4}];
array_group_by($list, 'color');
# => {black => [{color => 'black', id => 1}, {color => 'black', id => 4}],
# white => [{color => 'white', id => 2}]}
=item array_pluck($array, $key)
my $list = [{id => 1}, {}, {id => 3}];
array_pluck($list, 'id');
# => [1, undef, 3]
=item array_to_sentence($array)
=item array_to_sentence($array, $join)
=item array_to_sentence($array, $join, $join_last)
array_to_sentence([1,2,3]);
# => "1, 2 and 3"
array_to_sentence([1,2,3], ",");
# => "1,2 and 3"
array_to_sentence([1,2,3], ",", " & ");
# => "1,2 & 3"
=item array_sum($array)
array_sum([1,2,3]);
# => 6
=item array_includes($array, $val)
Returns 1 if C<$array> includes a value that is deeply equal to C<$val>, 0
otherwise. Comparison is done with C<is_same()>.
array_includes([{color => 'black'}], {color => 'white'});
# => 0
array_includes([{color => 'black'}], {color => 'black'});
# => 1
=item array_any($array, \&sub)
array_any(["green", "blue"], sub { my $color = $_[0]; $color eq "blue" });
# => 1
=item array_rest($array)
Returns a copy of C<$array> without the head.
array_rest([1,2,3,4]);
# => [2,3,4]
array_rest([1]);
# => []
=item array_uniq($array)
Returns a copy of C<$array> with all duplicates removed. Comparison is done
with C<is_same()>.
=back
=head2 String functions
use Catmandu::Util qw(:string);
=over 4
=item as_utf8($str)
Returns a copy of C<$str> flagged as UTF-8.
=item trim($str)
Returns a copy of C<$str> with leading and trailing whitespace removed.
=item capitalize($str)
Equivalent to C<< ucfirst lc as_utf8 $str >>.
=back
=head2 Is functions
use Catmandu::Util qw(:is);
is_number(42) ? "it's numeric" : "it's not numeric";
is_maybe_hash_ref({});
# => 1
is_maybe_hash_ref(undef);
# => 1
is_maybe_hash_ref([]);
# => 0
A collection of predicate functions that test the type or value of argument
C<$val>. Each function (except C<is_same()> and C<is_different>) also has a
I<maybe> variant that also tests true if C<$val> is undefined.
Returns C<1> or C<0>.
=over 4
=item is_invocant($val)
=item is_maybe_invocant($val)
Tests if C<$val> is callable (is an existing package or blessed object).
=item is_able($val, @method_names)
=item is_maybe_able($val, @method_names)
Tests if C<$val> is callable and has all methods in C<@method_names>.
=item is_ref($val)
=item is_maybe_ref($val)
Tests if C<$val> is a reference. Equivalent to C<< ref $val ? 1 : 0 >>.
=item is_scalar_ref($val)
=item is_maybe_scalar_ref($val)
Tests if C<$val> is a scalar reference.
=item is_array_ref($val)
=item is_maybe_array_ref($val)
Tests if C<$val> is an array reference.
=item is_hash_ref($val)
=item is_maybe_hash_ref($val)
Tests if C<$val> is a hash reference.
=item is_code_ref($val)
=item is_maybe_code_ref($val)
Tests if C<$val> is a subroutine reference.
=item is_regex_ref($val)
=item is_maybe_regex_ref($val)
Tests if C<$val> is a regular expression reference generated by the C<qr//>
operator.
=item is_glob_ref($val)
=item is_maybe_glob_ref($val)
Tests if C<$val> is a glob reference.
=item is_value($val)
=item is_maybe_value($val)
Tests if C<$val> is a real value (defined, not a reference and not a
glob.
=item is_string($val)
=item is_maybe_string($val)
Tests if C<$val> is a non-empty string.
Equivalent to C<< is_value($val) && length($val) > 0 >>.
=item is_number($val)
=item is_maybe_number($val)
Tests if C<$val> is a number.
=item is_integer($val)
=item is_maybe_integer($val)
Tests if C<$val> is an integer.
=item is_natural($val)
=item is_maybe_natural($val)
Tests if C<$val> is a non-negative integer.
Equivalent to C<< is_integer($val) && $val >= 0 >>.
=item is_positive($val)
=item is_maybe_positive($val)
Tests if C<$val> is a positive integer.
Equivalent to C<< is_integer($val) && $val >= 1 >>.
=item is_same($val, $other_val)
Tests if C<$val> is deeply equal to C<$other_val>.
=item is_different($val, $other_val)
The opposite of C<is_same()>.
=back
=head2 Check functions
use Catmandu::Util qw(:check);
check_hash_ref({color => 'red'});
# => {color => 'red'}
check_hash_ref([]);
# dies
A group of assert functions similar to the C<:is> group, but instead of
returning true or false they return their argument or die.
=over 4
=item check_invocant($val)
=item check_maybe_invocant($val)
=item check_able($val, @method_names)
=item check_maybe_able($val, @method_names)
=item check_ref($val)
=item check_maybe_ref($val)
=item check_scalar_ref($val)
=item check_maybe_scalar_ref($val)
=item check_array_ref($val)
=item check_maybe_array_ref($val)
=item check_hash_ref($val)
=item check_maybe_hash_ref($val)
=item check_code_ref($val)
=item check_maybe_code_ref($val)
=item check_regex_ref($val)
=item check_maybe_regex_ref($val)
=item check_glob_ref($val)
=item check_maybe_glob_ref($val)
=item check_value($val)
=item check_maybe_value($val)
=item check_string($val)
=item check_maybe_string($val)
=item check_number($val)
=item check_maybe_number($val)
=item check_integer($val)
=item check_maybe_integer($val)
=item check_natural($val)
=item check_maybe_natural($val)
=item check_positive($val)
=item check_maybe_positive($val)
=item check_same($val, $other_val)
=item check_different($val, $other_val)
=back
=head2 Human output functions
use Catmandu::Util qw(:human);
=over 4
=item human_number($num)
Insert a comma a 3-digit intervals to make C<$num> more readable. Only works
with I<integers> for now.
human_number(64354);
# => "64,354"
=item human_byte_size($size)
human_byte_size(64);
# => "64 bytes"
human_byte_size(10005000);
# => "10.01 MB"
=item human_content_type($content_type)
=item human_content_type($content_type, $default)
human_content_type('application/x-dos_ms_excel');
# => "Excel"
human_content_type('application/zip');
# => "ZIP archive"
human_content_type('foo/x-unknown');
# => "foo/x-unknown"
human_content_type('foo/x-unknown', 'Unknown');
# => "Unknown"
=back
=head2 XML functions
use Catmandu::Util qw(:xml);
=over 4
=item xml_declaration()
Returns C<< qq(<?xml version="1.0" encoding="UTF-8"?>\n) >>.
=item xml_escape($str)
Returns an XML escaped copy of C<$str>.
=back
=head2 Miscellaneous functions
=over 4
=item require_package($pkg)
=item require_package($pkg, $namespace)
Load package C<$pkg> at runtime with C<require> and return it's full name.
my $pkg = require_package('File::Spec');
my $dir = $pkg->tmpdir();
require_package('Util', 'Catmandu');
# => "Catmandu::Util"
require_package('Catmandu::Util', 'Catmandu');
# => "Catmandu::Util"
=item use_lib(@dirs)
Add directories to C<@INC> at runtime.
=back
=head1 SEE ALSO
L<Data::Util>.
=cut