xref: /openbsd/gnu/usr.bin/perl/lib/utf8.pm (revision 5dea098c)
1package utf8;
2
3use strict;
4use warnings;
5
6our $hint_bits = 0x00800000;
7
8our $VERSION = '1.24';
9our $AUTOLOAD;
10
11sub import {
12    $^H |= $hint_bits;
13}
14
15sub unimport {
16    $^H &= ~$hint_bits;
17}
18
19sub AUTOLOAD {
20    goto &$AUTOLOAD if defined &$AUTOLOAD;
21    require Carp;
22    Carp::croak("Undefined subroutine $AUTOLOAD called");
23}
24
251;
26__END__
27
28=head1 NAME
29
30utf8 - Perl pragma to enable/disable UTF-8 (or UTF-EBCDIC) in source code
31
32=head1 SYNOPSIS
33
34 use utf8;
35 no utf8;
36
37 # Convert the internal representation of a Perl scalar to/from UTF-8.
38
39 $num_octets = utf8::upgrade($string);
40 $success    = utf8::downgrade($string[, $fail_ok]);
41
42 # Change each character of a Perl scalar to/from a series of
43 # characters that represent the UTF-8 bytes of each original character.
44
45 utf8::encode($string);  # "\x{100}"  becomes "\xc4\x80"
46 utf8::decode($string);  # "\xc4\x80" becomes "\x{100}"
47
48 # Convert a code point from the platform native character set to
49 # Unicode, and vice-versa.
50 $unicode = utf8::native_to_unicode(ord('A')); # returns 65 on both
51                                               # ASCII and EBCDIC
52                                               # platforms
53 $native = utf8::unicode_to_native(65);        # returns 65 on ASCII
54                                               # platforms; 193 on
55                                               # EBCDIC
56
57 $flag = utf8::is_utf8($string); # since Perl 5.8.1
58 $flag = utf8::valid($string);
59
60=head1 DESCRIPTION
61
62The C<use utf8> pragma tells the Perl parser to allow UTF-8 in the
63program text in the current lexical scope.  The C<no utf8> pragma tells Perl
64to switch back to treating the source text as literal bytes in the current
65lexical scope.  (On EBCDIC platforms, technically it is allowing UTF-EBCDIC,
66and not UTF-8, but this distinction is academic, so in this document the term
67UTF-8 is used to mean both).
68
69B<Do not use this pragma for anything else than telling Perl that your
70script is written in UTF-8.> The utility functions described below are
71directly usable without C<use utf8;>.
72
73Because it is not possible to reliably tell UTF-8 from native 8 bit
74encodings, you need either a Byte Order Mark at the beginning of your
75source code, or C<use utf8;>, to instruct perl.
76
77When UTF-8 becomes the standard source format, this pragma will
78effectively become a no-op.
79
80See also the effects of the C<-C> switch and its cousin, the
81C<PERL_UNICODE> environment variable, in L<perlrun>.
82
83Enabling the C<utf8> pragma has the following effect:
84
85=over 4
86
87=item *
88
89Bytes in the source text that are not in the ASCII character set will be
90treated as being part of a literal UTF-8 sequence.  This includes most
91literals such as identifier names, string constants, and constant
92regular expression patterns.
93
94=back
95
96Note that if you have non-ASCII, non-UTF-8 bytes in your script (for example
97embedded Latin-1 in your string literals), C<use utf8> will be unhappy.  If
98you want to have such bytes under C<use utf8>, you can disable this pragma
99until the end the block (or file, if at top level) by C<no utf8;>.
100
101=head2 Utility functions
102
103The following functions are defined in the C<utf8::> package by the
104Perl core.  You do not need to say C<use utf8> to use these and in fact
105you should not say that unless you really want to have UTF-8 source code.
106
107=over 4
108
109=item * C<$num_octets = utf8::upgrade($string)>
110
111(Since Perl v5.8.0)
112Converts in-place the internal representation of the string from an octet
113sequence in the native encoding (Latin-1 or EBCDIC) to UTF-8. The
114logical character sequence itself is unchanged.  If I<$string> is already
115upgraded, then this is a no-op. Returns the
116number of octets necessary to represent the string as UTF-8.
117
118If your code needs to be compatible with versions of perl without
119C<use feature 'unicode_strings';>, you can force Unicode semantics on
120a given string:
121
122  # force unicode semantics for $string without the
123  # "unicode_strings" feature
124  utf8::upgrade($string);
125
126For example:
127
128  # without explicit or implicit use feature 'unicode_strings'
129  my $x = "\xDF";    # LATIN SMALL LETTER SHARP S
130  $x =~ /ss/i;       # won't match
131  my $y = uc($x);    # won't convert
132  utf8::upgrade($x);
133  $x =~ /ss/i;       # matches
134  my $z = uc($x);    # converts to "SS"
135
136B<Note that this function does not handle arbitrary encodings>;
137use L<Encode> instead.
138
139=item * C<$success = utf8::downgrade($string[, $fail_ok])>
140
141(Since Perl v5.8.0)
142Converts in-place the internal representation of the string from UTF-8 to the
143equivalent octet sequence in the native encoding (Latin-1 or EBCDIC). The
144logical character sequence itself is unchanged. If I<$string> is already
145stored as native 8 bit, then this is a no-op.  Can be used to make sure that
146the UTF-8 flag is off, e.g. when you want to make sure that the substr() or
147length() function works with the usually faster byte algorithm.
148
149Fails if the original UTF-8 sequence cannot be represented in the
150native 8 bit encoding. On failure dies or, if the value of I<$fail_ok> is
151true, returns false.
152
153Returns true on success.
154
155If your code expects an octet sequence this can be used to validate
156that you've received one:
157
158  # throw an exception if not representable as octets
159  utf8::downgrade($string)
160
161  # or do your own error handling
162  utf8::downgrade($string, 1) or die "string must be octets";
163
164B<Note that this function does not handle arbitrary encodings>;
165use L<Encode> instead.
166
167=item * C<utf8::encode($string)>
168
169(Since Perl v5.8.0)
170Converts in-place the character sequence to the corresponding octet
171sequence in Perl's extended UTF-8. That is, every (possibly wide) character
172gets replaced with a sequence of one or more characters that represent the
173individual UTF-8 bytes of the character.  The UTF8 flag is turned off.
174Returns nothing.
175
176 my $x = "\x{100}"; # $x contains one character, with ord 0x100
177 utf8::encode($x);  # $x contains two characters, with ords (on
178                    # ASCII platforms) 0xc4 and 0x80.  On EBCDIC
179                    # 1047, this would instead be 0x8C and 0x41.
180
181Similar to:
182
183  use Encode;
184  $x = Encode::encode("utf8", $x);
185
186B<Note that this function does not handle arbitrary encodings>;
187use L<Encode> instead.
188
189=item * C<$success = utf8::decode($string)>
190
191(Since Perl v5.8.0)
192Attempts to convert in-place the octet sequence encoded in Perl's extended
193UTF-8 to the corresponding character sequence. That is, it replaces each
194sequence of characters in the string whose ords represent a valid (extended)
195UTF-8 byte sequence, with the corresponding single character.  The UTF-8 flag
196is turned on only if the source string contains multiple-byte UTF-8
197characters.  If I<$string> is invalid as extended UTF-8, returns false;
198otherwise returns true.
199
200 my $x = "\xc4\x80"; # $x contains two characters, with ords
201                     # 0xc4 and 0x80
202 utf8::decode($x);   # On ASCII platforms, $x contains one char,
203                     # with ord 0x100.   Since these bytes aren't
204                     # legal UTF-EBCDIC, on EBCDIC platforms, $x is
205                     # unchanged and the function returns FALSE.
206 my $y = "\xc3\x83\xc2\xab"; This has been encoded twice; this
207                     # example is only for ASCII platforms
208 utf8::decode($y);   # Converts $y to \xc3\xab, returns TRUE;
209 utf8::decode($y);   # Further converts to \xeb, returns TRUE;
210 utf8::decode($y);   # Returns FALSE, leaves $y unchanged
211
212B<Note that this function does not handle arbitrary encodings>;
213use L<Encode> instead.
214
215=item * C<$unicode = utf8::native_to_unicode($code_point)>
216
217(Since Perl v5.8.0)
218This takes an unsigned integer (which represents the ordinal number of a
219character (or a code point) on the platform the program is being run on) and
220returns its Unicode equivalent value.  Since ASCII platforms natively use the
221Unicode code points, this function returns its input on them.  On EBCDIC
222platforms it converts from EBCDIC to Unicode.
223
224A meaningless value will currently be returned if the input is not an unsigned
225integer.
226
227Since Perl v5.22.0, calls to this function are optimized out on ASCII
228platforms, so there is no performance hit in using it there.
229
230=item * C<$native = utf8::unicode_to_native($code_point)>
231
232(Since Perl v5.8.0)
233This is the inverse of C<utf8::native_to_unicode()>, converting the other
234direction.  Again, on ASCII platforms, this returns its input, but on EBCDIC
235platforms it will find the native platform code point, given any Unicode one.
236
237A meaningless value will currently be returned if the input is not an unsigned
238integer.
239
240Since Perl v5.22.0, calls to this function are optimized out on ASCII
241platforms, so there is no performance hit in using it there.
242
243=item * C<$flag = utf8::is_utf8($string)>
244
245(Since Perl 5.8.1)  Test whether I<$string> is marked internally as encoded in
246UTF-8.  Functionally the same as C<Encode::is_utf8($string)>.
247
248Typically only necessary for debugging and testing, if you need to
249dump the internals of an SV, L<Devel::Peek's|Devel::Peek> Dump()
250provides more detail in a compact form.
251
252If you still think you need this outside of debugging, testing or
253dealing with filenames, you should probably read L<perlunitut> and
254L<perlunifaq/What is "the UTF8 flag"?>.
255
256Don't use this flag as a marker to distinguish character and binary
257data: that should be decided for each variable when you write your
258code.
259
260To force unicode semantics in code portable to perl 5.8 and 5.10, call
261C<utf8::upgrade($string)> unconditionally.
262
263=item * C<$flag = utf8::valid($string)>
264
265[INTERNAL] Test whether I<$string> is in a consistent state regarding
266UTF-8.  Will return true if it is well-formed Perl extended UTF-8 and has the
267UTF-8 flag
268on B<or> if I<$string> is held as bytes (both these states are 'consistent').
269The main reason for this routine is to allow Perl's test suite to check
270that operations have left strings in a consistent state.
271
272=back
273
274C<utf8::encode> is like C<utf8::upgrade>, but the UTF8 flag is
275cleared.  See L<perlunicode>, and the C API
276functions C<L<sv_utf8_upgrade|perlapi/sv_utf8_upgrade>>,
277C<L<perlapi/sv_utf8_downgrade>>, C<L<perlapi/sv_utf8_encode>>,
278and C<L<perlapi/sv_utf8_decode>>, which are wrapped by the Perl functions
279C<utf8::upgrade>, C<utf8::downgrade>, C<utf8::encode> and
280C<utf8::decode>.  Also, the functions C<utf8::is_utf8>, C<utf8::valid>,
281C<utf8::encode>, C<utf8::decode>, C<utf8::upgrade>, and C<utf8::downgrade> are
282actually internal, and thus always available, without a C<require utf8>
283statement.
284
285=head1 BUGS
286
287Some filesystems may not support UTF-8 file names, or they may be supported
288incompatibly with Perl.  Therefore UTF-8 names that are visible to the
289filesystem, such as module names may not work.
290
291=head1 SEE ALSO
292
293L<perlunitut>, L<perluniintro>, L<perlrun>, L<bytes>, L<perlunicode>
294
295=cut
296