4 use Encode qw(:fallbacks find_encoding);
5 our $VERSION = do { my @r = (q$Revision: 1.7 $ =~ /\d+/g); sprintf "%d."."%02d" x $#r, @r };
9 our %DEF_SUSPECTS = map { $_ => find_encoding($_) } qw(ascii utf8);
10 $Encode::Encoding{$Canon} =
13 Suspects => { %DEF_SUSPECTS },
16 use base qw(Encode::Encoding);
20 our @EXPORT = qw(guess_encoding);
22 sub import { # Exporter not used so we do it on our own
24 for my $item (@EXPORT){
26 *{"$callpkg\::$item"} = \&{"$item"};
33 my $self = ref($class) ? $class : $Encode::Encoding{$Canon};
34 $self->{Suspects} = { %DEF_SUSPECTS };
35 $self->add_suspects(@_);
40 my $self = ref($class) ? $class : $Encode::Encoding{$Canon};
42 my $e = find_encoding($c) or die "Unknown encoding: $c";
43 $self->{Suspects}{$e->name} = $e;
44 $DEBUG and warn "Added: ", $e->name;
49 my ($obj, $octet, $chk) = @_;
50 my $guessed = guess($obj, $octet);
51 unless (ref($guessed)){
53 Carp::croak($guessed);
55 my $utf8 = $guessed->decode($octet, $chk);
56 $_[1] = $octet if $chk;
61 guess($Encode::Encoding{$Canon}, @_);
66 my $obj = ref($class) ? $class : $Encode::Encoding{$Canon};
70 return unless defined $octet and length $octet;
73 Encode::is_utf8($octet) and return find_encoding('utf8');
76 my $BOM = unpack('n', $octet);
77 return find_encoding('UTF-16')
78 if (defined $BOM and ($BOM == 0xFeFF or $BOM == 0xFFFe));
79 $BOM = unpack('N', $octet);
80 return find_encoding('UTF-32')
81 if (defined $BOM and ($BOM == 0xFeFF or $BOM == 0xFFFe0000));
83 my %try = %{$obj->{Suspects}};
85 my $e = find_encoding($c) or die "Unknown encoding: $c";
87 $DEBUG and warn "Added: ", $e->name;
90 for my $line (split /\r\n?|\n/, $octet){
91 # cheat 2 -- \e in the string
94 delete @try{qw/utf8 ascii/};
96 ref($try{$k}) eq 'Encode::XS' and delete $try{$k};
100 # warn join(",", keys %try);
101 for my $k (keys %try){
103 $try{$k}->decode($scratch, FB_QUIET);
105 $DEBUG and warn sprintf("%4d:%-24s ok\n", $nline, $k);
109 warn sprintf("%4d:%-24s not ok; %d bytes left\n",
110 $nline, $k, bytes::length($scratch));
115 %ok or return "No appropriate encodings found!";
116 if (scalar(keys(%ok)) == 1){
117 my ($retval) = values(%ok);
120 %try = %ok; $nline++;
123 return "Encodings too ambiguous: ", join(" or ", keys %try);
134 Encode::Guess -- Guesses encoding from data
138 # if you are sure $data won't contain anything bogus
141 use Encode::Guess qw/euc-jp shiftjis 7bit-jis/;
142 my $utf8 = decode("Guess", $data);
143 my $data = encode("Guess", $utf8); # this doesn't work!
147 my $enc = guess_encoding($data, qw/euc-jp shiftjis 7bit-jis/);
148 ref($enc) or die "Can't guess: $enc"; # trap error this way
149 $utf8 = $enc->decode($data);
151 $utf8 = decode($enc->name, $data)
155 Encode::Guess enables you to guess in what encoding a given data is
156 encoded, or at least tries to.
160 By default, it checks only ascii, utf8 and UTF-16/32 with BOM.
162 use Encode::Guess; # ascii/utf8/BOMed UTF
164 To use it more practically, you have to give the names of encodings to
165 check (I<suspects> as follows). The name of suspects can either be
166 canonical names or aliases.
168 # tries all major Japanese Encodings as well
169 use Encode::Guess qw/euc-jp shiftjis 7bit-jis/;
173 =item Encode::Guess->set_suspects
175 You can also change the internal suspects list via C<set_suspects>
179 Encode::Guess->set_suspects(qw/euc-jp shiftjis 7bit-jis/);
181 =item Encode::Guess->add_suspects
183 Or you can use C<add_suspects> method. The difference is that
184 C<set_suspects> flushes the current suspects list while
185 C<add_suspects> adds.
188 Encode::Guess->add_suspects(qw/euc-jp shiftjis 7bit-jis/);
189 # now the suspects are euc-jp,shiftjis,7bit-jis, AND
190 # euc-kr,euc-cn, and big5-eten
191 Encode::Guess->add_suspects(qw/euc-kr euc-cn big5-eten/);
193 =item Encode::decode("Guess" ...)
195 When you are content with suspects list, you can now
197 my $utf8 = Encode::decode("Guess", $data);
199 =item Encode::Guess->guess($data)
201 But it will croak if:
207 Two or more suspects remain
215 So you should instead try this;
217 my $decoder = Encode::Guess->guess($data);
219 On success, $decoder is an object that is documented in
220 L<Encode::Encoding>. So you can now do this;
222 my $utf8 = $decoder->decode($data);
224 On failure, $decoder now contains an error message so the whole thing
227 my $decoder = Encode::Guess->guess($data);
228 die $decoder unless ref($decoder);
229 my $utf8 = $decoder->decode($data);
231 =item guess_encoding($data, [, I<list of suspects>])
233 You can also try C<guess_encoding> function which is exported by
234 default. It takes $data to check and it also takes the list of
235 suspects by option. The optional suspect list is I<not reflected> to
236 the internal suspects list.
238 my $decoder = guess_encoding($data, qw/euc-jp euc-kr euc-cn/);
239 die $decoder unless ref($decoder);
240 my $utf8 = $decoder->decode($data);
241 # check only ascii and utf8
242 my $decoder = guess_encoding($data);
252 Because of the algorithm used, ISO-8859 series and other single-byte
253 encodings do not work well unless either one of ISO-8859 is the only
254 one suspect (besides ascii and utf8).
258 my $decoder = guess_encoding($data, 'latin1');
260 my $decoder = guess_encoding($data, qw/latin1 greek/);
262 The reason is that Encode::Guess guesses encoding by trial and error.
263 It first splits $data into lines and tries to decode the line for each
264 suspect. It keeps it going until all but one encoding is eliminated
265 out of suspects list. ISO-8859 series is just too successful for most
266 cases (because it fills almost all code points in \x00-\xff).
270 Do not mix national standard encodings and the corresponding vendor
275 = guess_encoding($data, qw/shiftjis MacJapanese cp932/);
277 The reason is that vendor encoding is usually a superset of national
278 standard so it becomes too ambiguous for most cases.
282 On the other hand, mixing various national standard encodings
283 automagically works unless $data is too short to allow for guessing.
285 # This is ok if $data is long enough
287 guess_encoding($data, qw/euc-cn
288 euc-jp shiftjis 7bit-jis
294 DO NOT PUT TOO MANY SUSPECTS! Don't you try something like this!
296 my $decoder = guess_encoding($data,
297 Encode->encodings(":all"));
301 It is, after all, just a guess. You should alway be explicit when it
302 comes to encodings. But there are some, especially Japanese,
303 environment that guess-coding is a must. Use this module with care.
307 Encode::Guess does not work on EBCDIC platforms.
311 L<Encode>, L<Encode::Encoding>