GSP
Quick Navigator

Search Site

Unix VPS
A - Starter
B - Basic
C - Preferred
D - Commercial
MPS - Dedicated
Previous VPSs
* Sign Up! *

Support
Contact Us
Online Help
Handbooks
Domain Status
Man Pages

FAQ
Virtual Servers
Pricing
Billing
Technical

Network
Facilities
Connectivity
Topology Map

Miscellaneous
Server Agreement
Year 2038
Credits
 

USA Flag

 

 

Man Pages
Multibyte(3) User Contributed Perl Documentation Multibyte(3)

String::Multibyte - manipulation of multibyte character strings

    use String::Multibyte;

    $utf8 = String::Multibyte->new('UTF8');
    $utf8_len = $utf8->length($utf8_str);

This module provides some functions which emulate the corresponding "CORE" functions for locale-independent manipulation of multiple-byte character strings.

Why this module is locale-independent? Well, because this module only consider the byte sequence structure of charsets and is not aware of any Locale stuff! Locale-dependent methods like "uc()", "lc()", etc., will not be supported at all.

The definition files are sited under the directory where String::Multibyte is sited. E.g. if String::Multibyte is "perl/site/lib/String/Multibyte.pm", copy String::Multibyte::Foo as "perl/site/lib/String/Multibyte/Foo.pm".

The definition file must return a hashref, having key(s) named as following.

"charset"
The value for the key 'charset' stands for a string of the charset name. In almost case, omission of the 'charset' matters very little, but keep them not conflict among another charset.
"regexp"
The value for the key 'regexp', REQUIRED, is a regular expression that matchs a single character of charset in question. (You may use "qr//" if available.)

If the 'regexp' is omitted, calling any method is croaked.

"nextchar"
The value for the key 'nextchar' must be a coderef that returns the next character to the specified character. If the 'nextchar' coderef is omitted, "mkrange()" and "strtr()" methods don't understand hyphen metacharacter for character ranges.
"cmpchar"
The value for the key 'cmpchar' must be a coderef that compares the specified two characters. If the 'cmpchar' coderef is omitted, "mkrange" and "strtr" functions don't understand reverse character ranges.
"hyphen"
The value for the key 'hyphen' is a character to stand for a character range. The default is '-'.
"escape"
The value for the key 'escape' is an escape character for a "hyphen" character. The default is '\\'. The 'escape' character is valid only before a "hyphen" or another 'escape' (e.g. '\\\\-]' means '\\' to ']'; '\\\\\-]' means '\\', '-', and ']'). If an 'escape' character is followed by any character other than 'escape' or 'hyphen', it is parsed literally.

"$mbcs = String::Multibyte->new(CHARSET)"
"$mbcs = String::Multibyte->new(CHARSET, VERBOSE)"
"CHARSET" is the charset name; exactly speaking, the file name of the definition file (without the suffix .pm). It returns the instance to tell methods in which charset the specified strings should be handled.

"CHARSET" may be a hashref; this is how to define a charset without any .pm file.

    # see perlfaq6  :-)
    my $martian  = String::Multibyte->new({
        charset => "martian",
        regexp => '[A-Z][A-Z]|[^A-Z]',
    });
    

If true value is specified as "VERBOSE", the called method (excepting "islegal") will check its arguments and carps if any of them is not legally encoded.

Otherwise such a check won't be carried out (saves a bit of time, but unsafe, though you can use the "islegal" method if necessary).

"$mbcs->islegal(LIST)"
Returns a boolean indicating whether all the strings in arguments are legally encoded in the concerned charset. Returns false even if one element is illegal in "LIST".

"$mbcs->length(STRING)"
Returns the length in characters of the specified string.

"$mbcs->strrev(STRING)"
Returns a reversed string in characters.
"$mbcs->index(STRING, SUBSTR)"
"$mbcs->index(STRING, SUBSTR, POSITION)"
Returns the position of the first occurrence of "SUBSTR" in "STRING" at or after "POSITION". If "POSITION" is omitted, starts searching from the beginning of the string.

If the substring is not found, returns "-1".

"$mbcs->rindex(STRING, SUBSTR)"
"$mbcs->rindex(STRING, SUBSTR, POSITION)"
Returns the position of the last occurrence of "SUBSTR" in "STRING" at or after "POSITION". If "POSITION" is specified, returns the last occurrence at or before that position.

If the substring is not found, returns "-1".

"$mbcs->strspn(STRING, SEARCHLIST)"
Returns returns the position of the first occurrence of any character not contained in the search list.

  $mbcs->strspn("+0.12345*12", "+-.0123456789");
  # returns 8.
    

If the specified string does not contain any character in the search list, returns 0.

The string consists of characters in the search list, the returned value equals the length of the string.

"SEARCHLIST" can be an "ARRAYREF". e.g. if a charset treats "CRLF" as a single character, "\r\n" is a one-element list of only "\r\n". A two-element list of "\r" and "\n" can be given as "["\r", "\n"]" (of course "\n\r" is also ok since the character order of "SEARCHLIST" doesn't matter in "strspn").

"$mbcs->strcspn(STRING, SEARCHLIST)"
Returns returns the position of the first occurrence of any character contained in the search list.

If the specified string does not contain any character in the search list, the returned value equals the length of the string.

"SEARCHLIST" can be an "ARRAYREF". e.g. if a charset treats "CRLF" as a single character, "\r\n" is a one-element list of only "\r\n". A two-element list of "\r" and "\n" can be given as "["\r", "\n"]" (of course "\n\r" is also ok since the character order of "SEARCHLIST" doesn't matter in "strcspn").

"$mbcs->substr(STRING or SCALAR REF, OFFSET)"
"$mbcs->substr(STRING or SCALAR REF, OFFSET, LENGTH)"
"$mbcs->substr(SCALAR, OFFSET, LENGTH, REPLACEMENT)"
It works like "CORE::substr", but using character semantics of multibyte charset encoding.

If the "REPLACEMENT" as the fourth argument is specified, replaces parts of the "SCALAR" and returns what was there before.

You can utilize the lvalue reference, returned if a reference of scalar variable is used as the first argument.

    ${ $mbcs->substr(\$str,$off,$len) } = $replace;

        works like

    CORE::substr($str,$off,$len) = $replace;
    

The returned lvalue is not multibyte-aware, then successive assignment may lead to odd results.

"$mbcs->strsplit(SEPARATOR, STRING)"
"$mbcs->strsplit(SEPARATOR, STRING, LIMIT)"
This function emulates "CORE::split", but splits on the "SEPARATOR" string, not by a pattern.

If not in list context, only return the number of fields found, but does not split into the @_ array.

If empty string is specified as "SEPARATOR", splits the specified string into characters.

  $bytes->strsplit('', 'This is perl.', 7);
  # ('T', 'h', 'i', 's', ' ', 'i',  's perl.')
    

"$mbcs->mkrange(CHARLIST, ALLOW_REVERSE)"
Returns the character list (not in list context, as a concatenated string) gained by parsing the specified character range.

The result depends on the the character order for the concerned charset. About the character order for each charset, see its definition file.

If the character order is undefined in the definition file, returns an identical string with the specified string.

A character range is specified with a hyphen ('-', but exactly speaking, "$obj->{hyphen}").

The backslashed combinations '\-' and '\\' (exactly speaking, "$obj->{escape}$obj->{hyphen}" and "$obj->{escape}$obj->{escape}") are used instead of the characters '-' and '\', respectively. The hyphen at the beginning or the end of the range is also evaluated as the hyphen itself.

For example, "$mbcs->mkrange('+\-0-9A-F')" returns "('+', '-', '0', '1', '2', '3', '4', '5', '6', '7', '8', '9', 'A', 'B', 'C', 'D', 'E', 'F')" and "scalar $mbcs->mkrange('A-P')" returns 'ABCDEFGHIJKLMNOP'.

If true value is specified as the second argument, reverse character ranges such as '9-0', 'Z-A' are allowed.

  $bytes = String::Multibyte->new('Bytes');
  $bytes->mkrange('p-e-r-l', 1); # ponmlkjihgfefghijklmnopqrqponml
    

"$mbcs->strtr(STRING or SCALAR REF, SEARCHLIST, REPLACEMENTLIST)"
"$mbcs->strtr(STRING or SCALAR REF, SEARCHLIST, REPLACEMENTLIST, MODIFIER)"
Transliterates all occurrences of the characters found in the search list with the corresponding character in the replacement list.

If a reference of scalar variable is specified as the first argument, returns the number of characters replaced or deleted; otherwise, returns the transliterated string and the specified string is unaffected.

If 'h' modifier is specified, returns a hash of histogram in list context; a reference to hash of histogram in scalar context;

SEARCHLIST and REPLACEMENTLIST

Character ranges (internally utilizing "mkrange()") are supported.

If the "REPLACEMENTLIST" is empty (specified as '', not "undef", because the use of uninitialized value causes warning under -w option), the "SEARCHLIST" is replicated.

If the replacement list is shorter than the search list, the final character in the replacement list is replicated till it is long enough (but differently works when the 'd' modifier is used).

"SEARCHLIST" and "REPLACEMENTLIST" can be an "ARRAYREF". e.g. if a charset treats "\r\n" ("CRLF") as a single character, "\r\n" is a one-element list of only "\r\n". A two-element list of "\r" and "\n" should be given as "["\r", "\n"]". Of course "\n\r" is also ok but the character order is different; cf. "strtr($str, ["\r", "\n"], ["\n", "\r"])" that swaps "\n" and "\r".

Each elements of "ARRAYREF" can include character ranges (the modifiers "R" and "r" affect their evaluation as usual).

"["A-C", "h-z"]" is evaluated like "A-Ch-z" if "charset" does not include grapheme "Ch". The former prevents "C" and "h" from evaluation as "Ch" even if the "charset" included grapheme "Ch".

MODIFIER

    c   Complement the SEARCHLIST.
    d   Delete found but unreplaced characters.
    s   Squash duplicate replaced characters.
    h   Return a hash (or a hashref) of histogram.
    R   No use of character ranges.
    r   Allows to use reverse character ranges.
    o   Caches the conversion table internally.
    

If 'R' modifier is specified, '-' is not evaluated as a meta character but hyphen itself like in "tr'''". Compare:

  $mbcs->strtr("90 - 32 = 58", "0-9", "A-J");
    # output: "JA - DC = FI"

  $mbcs->strtr("90 - 32 = 58", "0-9", "A-J", "R");
    # output: "JA - 32 = 58"
    # cf. ($str = "90 - 32 = 58") =~ tr'0-9'A-J';
    # '0' to 'A', '-' to '-', and '9' to 'J'.
    

If 'r' modifier is specified, reverse character ranges are allowed. e.g.

   $mbcs->strtr($str, "0-9", "9-0", "r")

     is equivalent to

   $mbcs->strtr($str, "0123456789", "9876543210")
    

Caching the conversion table

If 'o' modifier is specified, the conversion table is cached internally. e.g.

  foreach (@source_strings) {
    print $mbcs->strtr($_, $from_list, $to_list, 'o');
  }
    

will be almost as efficient as this:

  $trans = $mbcs->trclosure($from_list, $to_list);

  foreach (@source_strings) {
    print &$trans($_);
  }
    

You can use whichever you like.

Without 'o',

  foreach (@source_strings) {
    print $mbcs->strtr($_, $from_list, $to_list);
  }
    

will be very slow since the conversion table is made whenever the function is called.

"$mbcs->trclosure(SEARCHLIST, REPLACEMENTLIST)"
"$mbcs->trclosure(SEARCHLIST, REPLACEMENTLIST, MODIFIER)"
Returns a closure to transliterate the specified string. The return value is an only code reference, not blessed object. By use of this code ref, you can save yourself time as you need not specify arguments every time.

  my $trans = $mbcs->trclosure($from_list, $to_list);
  print &$trans ($string); # ok to perl 5.003
  print $trans->($string); # perl 5.004 or better
    

The functionality of the closure made by "trclosure()" is equivalent to that of "strtr()". Frankly speaking, the "strtr()" calls "trclosure()" internally and uses the returned closure.

"SEARCHLIST" and "REPLACEMENTLIST" can be an "ARRAYREF" same as "strtr()".

$[
This modules supposes $[ is always equal to 0, never 1.
Grapheme manipulation
Since v. 1.01, manipulation of sequence of graphemes is to be supported.

In a grapheme-aware manipulation, notice that the beginning and the end of a string always lie on a grapheme boundary.

E.g. imagine a grapheme set where a grapheme comprises either a leading latin capital letter followed by one or more latin small letters, or a single byte. Such a set can be define as below.

   $gra = String::Multibyte->new({
         regexp => '[A-Z][a-z]*|[\x00-\xFF]',
      });
    

Think about "$gra->index("Perl", "Pe")". As both "Perl" and "Pe" are a single grapheme, they are not equal to each other. So the result of this must be "-1" (meaning no match).

SADAHIRO Tomoyuki <SADAHIRO@cpan.org>

Copyright(C) 2001-2015, SADAHIRO Tomoyuki. Japan. All rights reserved.

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

perl(1).
2015-12-06 perl v5.32.1

Search for    or go to Top of page |  Section 3 |  Main Index

Powered by GSP Visit the GSP FreeBSD Man Page Interface.
Output converted with ManDoc.