| 1 | =head1 NAME |
| 2 | |
| 3 | perlunicode - Unicode support in Perl |
| 4 | |
| 5 | =head1 DESCRIPTION |
| 6 | |
| 7 | =head2 Important Caveats |
| 8 | |
| 9 | Unicode support is an extensive requirement. While Perl does not |
| 10 | implement the Unicode standard or the accompanying technical reports |
| 11 | from cover to cover, Perl does support many Unicode features. |
| 12 | |
| 13 | People who want to learn to use Unicode in Perl, should probably read |
| 14 | the L<Perl Unicode tutorial, perlunitut|perlunitut>, before reading |
| 15 | this reference document. |
| 16 | |
| 17 | Also, the use of Unicode may present security issues that aren't obvious. |
| 18 | Read L<Unicode Security Considerations|http://www.unicode.org/reports/tr36>. |
| 19 | |
| 20 | =over 4 |
| 21 | |
| 22 | =item Input and Output Layers |
| 23 | |
| 24 | Perl knows when a filehandle uses Perl's internal Unicode encodings |
| 25 | (UTF-8, or UTF-EBCDIC if in EBCDIC) if the filehandle is opened with |
| 26 | the ":utf8" layer. Other encodings can be converted to Perl's |
| 27 | encoding on input or from Perl's encoding on output by use of the |
| 28 | ":encoding(...)" layer. See L<open>. |
| 29 | |
| 30 | To indicate that Perl source itself is in UTF-8, use C<use utf8;>. |
| 31 | |
| 32 | =item Regular Expressions |
| 33 | |
| 34 | The regular expression compiler produces polymorphic opcodes. That is, |
| 35 | the pattern adapts to the data and automatically switches to the Unicode |
| 36 | character scheme when presented with data that is internally encoded in |
| 37 | UTF-8, or instead uses a traditional byte scheme when presented with |
| 38 | byte data. |
| 39 | |
| 40 | =item C<use utf8> still needed to enable UTF-8/UTF-EBCDIC in scripts |
| 41 | |
| 42 | As a compatibility measure, the C<use utf8> pragma must be explicitly |
| 43 | included to enable recognition of UTF-8 in the Perl scripts themselves |
| 44 | (in string or regular expression literals, or in identifier names) on |
| 45 | ASCII-based machines or to recognize UTF-EBCDIC on EBCDIC-based |
| 46 | machines. B<These are the only times when an explicit C<use utf8> |
| 47 | is needed.> See L<utf8>. |
| 48 | |
| 49 | =item BOM-marked scripts and UTF-16 scripts autodetected |
| 50 | |
| 51 | If a Perl script begins marked with the Unicode BOM (UTF-16LE, UTF16-BE, |
| 52 | or UTF-8), or if the script looks like non-BOM-marked UTF-16 of either |
| 53 | endianness, Perl will correctly read in the script as Unicode. |
| 54 | (BOMless UTF-8 cannot be effectively recognized or differentiated from |
| 55 | ISO 8859-1 or other eight-bit encodings.) |
| 56 | |
| 57 | =item C<use encoding> needed to upgrade non-Latin-1 byte strings |
| 58 | |
| 59 | By default, there is a fundamental asymmetry in Perl's Unicode model: |
| 60 | implicit upgrading from byte strings to Unicode strings assumes that |
| 61 | they were encoded in I<ISO 8859-1 (Latin-1)>, but Unicode strings are |
| 62 | downgraded with UTF-8 encoding. This happens because the first 256 |
| 63 | codepoints in Unicode happens to agree with Latin-1. |
| 64 | |
| 65 | See L</"Byte and Character Semantics"> for more details. |
| 66 | |
| 67 | =back |
| 68 | |
| 69 | =head2 Byte and Character Semantics |
| 70 | |
| 71 | Beginning with version 5.6, Perl uses logically-wide characters to |
| 72 | represent strings internally. |
| 73 | |
| 74 | In future, Perl-level operations will be expected to work with |
| 75 | characters rather than bytes. |
| 76 | |
| 77 | However, as an interim compatibility measure, Perl aims to |
| 78 | provide a safe migration path from byte semantics to character |
| 79 | semantics for programs. For operations where Perl can unambiguously |
| 80 | decide that the input data are characters, Perl switches to |
| 81 | character semantics. For operations where this determination cannot |
| 82 | be made without additional information from the user, Perl decides in |
| 83 | favor of compatibility and chooses to use byte semantics. |
| 84 | |
| 85 | Under byte semantics, when C<use locale> is in effect, Perl uses the |
| 86 | semantics associated with the current locale. Absent a C<use locale>, and |
| 87 | absent a C<use feature 'unicode_strings'> pragma, Perl currently uses US-ASCII |
| 88 | (or Basic Latin in Unicode terminology) byte semantics, meaning that characters |
| 89 | whose ordinal numbers are in the range 128 - 255 are undefined except for their |
| 90 | ordinal numbers. This means that none have case (upper and lower), nor are any |
| 91 | a member of character classes, like C<[:alpha:]> or C<\w>. (But all do belong |
| 92 | to the C<\W> class or the Perl regular expression extension C<[:^alpha:]>.) |
| 93 | |
| 94 | This behavior preserves compatibility with earlier versions of Perl, |
| 95 | which allowed byte semantics in Perl operations only if |
| 96 | none of the program's inputs were marked as being a source of Unicode |
| 97 | character data. Such data may come from filehandles, from calls to |
| 98 | external programs, from information provided by the system (such as %ENV), |
| 99 | or from literals and constants in the source text. |
| 100 | |
| 101 | The C<bytes> pragma will always, regardless of platform, force byte |
| 102 | semantics in a particular lexical scope. See L<bytes>. |
| 103 | |
| 104 | The C<use feature 'unicode_strings'> pragma is intended to always, regardless |
| 105 | of platform, force character (Unicode) semantics in a particular lexical scope. |
| 106 | In release 5.12, it is partially implemented, applying only to case changes. |
| 107 | See L</The "Unicode Bug"> below. |
| 108 | |
| 109 | The C<utf8> pragma is primarily a compatibility device that enables |
| 110 | recognition of UTF-(8|EBCDIC) in literals encountered by the parser. |
| 111 | Note that this pragma is only required while Perl defaults to byte |
| 112 | semantics; when character semantics become the default, this pragma |
| 113 | may become a no-op. See L<utf8>. |
| 114 | |
| 115 | Unless explicitly stated, Perl operators use character semantics |
| 116 | for Unicode data and byte semantics for non-Unicode data. |
| 117 | The decision to use character semantics is made transparently. If |
| 118 | input data comes from a Unicode source--for example, if a character |
| 119 | encoding layer is added to a filehandle or a literal Unicode |
| 120 | string constant appears in a program--character semantics apply. |
| 121 | Otherwise, byte semantics are in effect. The C<bytes> pragma should |
| 122 | be used to force byte semantics on Unicode data, and the C<use feature |
| 123 | 'unicode_strings'> pragma to force Unicode semantics on byte data (though in |
| 124 | 5.12 it isn't fully implemented). |
| 125 | |
| 126 | If strings operating under byte semantics and strings with Unicode |
| 127 | character data are concatenated, the new string will have |
| 128 | character semantics. This can cause surprises: See L</BUGS>, below. |
| 129 | You can choose to be warned when this happens. See L<encoding::warnings>. |
| 130 | |
| 131 | Under character semantics, many operations that formerly operated on |
| 132 | bytes now operate on characters. A character in Perl is |
| 133 | logically just a number ranging from 0 to 2**31 or so. Larger |
| 134 | characters may encode into longer sequences of bytes internally, but |
| 135 | this internal detail is mostly hidden for Perl code. |
| 136 | See L<perluniintro> for more. |
| 137 | |
| 138 | =head2 Effects of Character Semantics |
| 139 | |
| 140 | Character semantics have the following effects: |
| 141 | |
| 142 | =over 4 |
| 143 | |
| 144 | =item * |
| 145 | |
| 146 | Strings--including hash keys--and regular expression patterns may |
| 147 | contain characters that have an ordinal value larger than 255. |
| 148 | |
| 149 | If you use a Unicode editor to edit your program, Unicode characters may |
| 150 | occur directly within the literal strings in UTF-8 encoding, or UTF-16. |
| 151 | (The former requires a BOM or C<use utf8>, the latter requires a BOM.) |
| 152 | |
| 153 | Unicode characters can also be added to a string by using the C<\N{U+...}> |
| 154 | notation. The Unicode code for the desired character, in hexadecimal, |
| 155 | should be placed in the braces, after the C<U>. For instance, a smiley face is |
| 156 | C<\N{U+263A}>. |
| 157 | |
| 158 | Alternatively, you can use the C<\x{...}> notation for characters 0x100 and |
| 159 | above. For characters below 0x100 you may get byte semantics instead of |
| 160 | character semantics; see L</The "Unicode Bug">. On EBCDIC machines there is |
| 161 | the additional problem that the value for such characters gives the EBCDIC |
| 162 | character rather than the Unicode one. |
| 163 | |
| 164 | Additionally, if you |
| 165 | |
| 166 | use charnames ':full'; |
| 167 | |
| 168 | you can use the C<\N{...}> notation and put the official Unicode |
| 169 | character name within the braces, such as C<\N{WHITE SMILING FACE}>. |
| 170 | See L<charnames>. |
| 171 | |
| 172 | =item * |
| 173 | |
| 174 | If an appropriate L<encoding> is specified, identifiers within the |
| 175 | Perl script may contain Unicode alphanumeric characters, including |
| 176 | ideographs. Perl does not currently attempt to canonicalize variable |
| 177 | names. |
| 178 | |
| 179 | =item * |
| 180 | |
| 181 | Regular expressions match characters instead of bytes. "." matches |
| 182 | a character instead of a byte. |
| 183 | |
| 184 | =item * |
| 185 | |
| 186 | Bracketed character classes in regular expressions match characters instead of |
| 187 | bytes and match against the character properties specified in the |
| 188 | Unicode properties database. C<\w> can be used to match a Japanese |
| 189 | ideograph, for instance. |
| 190 | |
| 191 | =item * |
| 192 | |
| 193 | Named Unicode properties, scripts, and block ranges may be used (like bracketed |
| 194 | character classes) by using the C<\p{}> "matches property" construct and |
| 195 | the C<\P{}> negation, "doesn't match property". |
| 196 | See L</"Unicode Character Properties"> for more details. |
| 197 | |
| 198 | You can define your own character properties and use them |
| 199 | in the regular expression with the C<\p{}> or C<\P{}> construct. |
| 200 | See L</"User-Defined Character Properties"> for more details. |
| 201 | |
| 202 | =item * |
| 203 | |
| 204 | The special pattern C<\X> matches a logical character, an "extended grapheme |
| 205 | cluster" in Standardese. In Unicode what appears to the user to be a single |
| 206 | character, for example an accented C<G>, may in fact be composed of a sequence |
| 207 | of characters, in this case a C<G> followed by an accent character. C<\X> |
| 208 | will match the entire sequence. |
| 209 | |
| 210 | =item * |
| 211 | |
| 212 | The C<tr///> operator translates characters instead of bytes. Note |
| 213 | that the C<tr///CU> functionality has been removed. For similar |
| 214 | functionality see pack('U0', ...) and pack('C0', ...). |
| 215 | |
| 216 | =item * |
| 217 | |
| 218 | Case translation operators use the Unicode case translation tables |
| 219 | when character input is provided. Note that C<uc()>, or C<\U> in |
| 220 | interpolated strings, translates to uppercase, while C<ucfirst>, |
| 221 | or C<\u> in interpolated strings, translates to titlecase in languages |
| 222 | that make the distinction (which is equivalent to uppercase in languages |
| 223 | without the distinction). |
| 224 | |
| 225 | =item * |
| 226 | |
| 227 | Most operators that deal with positions or lengths in a string will |
| 228 | automatically switch to using character positions, including |
| 229 | C<chop()>, C<chomp()>, C<substr()>, C<pos()>, C<index()>, C<rindex()>, |
| 230 | C<sprintf()>, C<write()>, and C<length()>. An operator that |
| 231 | specifically does not switch is C<vec()>. Operators that really don't |
| 232 | care include operators that treat strings as a bucket of bits such as |
| 233 | C<sort()>, and operators dealing with filenames. |
| 234 | |
| 235 | =item * |
| 236 | |
| 237 | The C<pack()>/C<unpack()> letter C<C> does I<not> change, since it is often |
| 238 | used for byte-oriented formats. Again, think C<char> in the C language. |
| 239 | |
| 240 | There is a new C<U> specifier that converts between Unicode characters |
| 241 | and code points. There is also a C<W> specifier that is the equivalent of |
| 242 | C<chr>/C<ord> and properly handles character values even if they are above 255. |
| 243 | |
| 244 | =item * |
| 245 | |
| 246 | The C<chr()> and C<ord()> functions work on characters, similar to |
| 247 | C<pack("W")> and C<unpack("W")>, I<not> C<pack("C")> and |
| 248 | C<unpack("C")>. C<pack("C")> and C<unpack("C")> are methods for |
| 249 | emulating byte-oriented C<chr()> and C<ord()> on Unicode strings. |
| 250 | While these methods reveal the internal encoding of Unicode strings, |
| 251 | that is not something one normally needs to care about at all. |
| 252 | |
| 253 | =item * |
| 254 | |
| 255 | The bit string operators, C<& | ^ ~>, can operate on character data. |
| 256 | However, for backward compatibility, such as when using bit string |
| 257 | operations when characters are all less than 256 in ordinal value, one |
| 258 | should not use C<~> (the bit complement) with characters of both |
| 259 | values less than 256 and values greater than 256. Most importantly, |
| 260 | DeMorgan's laws (C<~($x|$y) eq ~$x&~$y> and C<~($x&$y) eq ~$x|~$y>) |
| 261 | will not hold. The reason for this mathematical I<faux pas> is that |
| 262 | the complement cannot return B<both> the 8-bit (byte-wide) bit |
| 263 | complement B<and> the full character-wide bit complement. |
| 264 | |
| 265 | =item * |
| 266 | |
| 267 | You can define your own mappings to be used in C<lc()>, |
| 268 | C<lcfirst()>, C<uc()>, and C<ucfirst()> (or their double-quoted string inlined |
| 269 | versions such as C<\U>). See |
| 270 | L<User-Defined Case-Mappings|/"User-Defined Case Mappings (for serious hackers only)"> |
| 271 | for more details. |
| 272 | |
| 273 | =back |
| 274 | |
| 275 | =over 4 |
| 276 | |
| 277 | =item * |
| 278 | |
| 279 | And finally, C<scalar reverse()> reverses by character rather than by byte. |
| 280 | |
| 281 | =back |
| 282 | |
| 283 | =head2 Unicode Character Properties |
| 284 | |
| 285 | Most Unicode character properties are accessible by using regular expressions. |
| 286 | They are used (like bracketed character classes) by using the C<\p{}> "matches |
| 287 | property" construct and the C<\P{}> negation, "doesn't match property". |
| 288 | |
| 289 | Note that the only time that Perl considers a sequence of individual code |
| 290 | points as a single logical character is in the C<\X> construct, already |
| 291 | mentioned above. Therefore "character" in this discussion means a single |
| 292 | Unicode code point. |
| 293 | |
| 294 | For instance, C<\p{Uppercase}> matches any single character with the Unicode |
| 295 | "Uppercase" property, while C<\p{L}> matches any character with a |
| 296 | General_Category of "L" (letter) property. Brackets are not |
| 297 | required for single letter property names, so C<\p{L}> is equivalent to C<\pL>. |
| 298 | |
| 299 | More formally, C<\p{Uppercase}> matches any single character whose Unicode |
| 300 | Uppercase property value is True, and C<\P{Uppercase}> matches any character |
| 301 | whose Uppercase property value is False, and they could have been written as |
| 302 | C<\p{Uppercase=True}> and C<\p{Uppercase=False}>, respectively. |
| 303 | |
| 304 | This formality is needed when properties are not binary, that is if they can |
| 305 | take on more values than just True and False. For example, the Bidi_Class (see |
| 306 | L</"Bidirectional Character Types"> below), can take on a number of different |
| 307 | values, such as Left, Right, Whitespace, and others. To match these, one needs |
| 308 | to specify the property name (Bidi_Class), and the value being matched against |
| 309 | (Left, Right, etc.). This is done, as in the examples above, by having the |
| 310 | two components separated by an equal sign (or interchangeably, a colon), like |
| 311 | C<\p{Bidi_Class: Left}>. |
| 312 | |
| 313 | All Unicode-defined character properties may be written in these compound forms |
| 314 | of C<\p{property=value}> or C<\p{property:value}>, but Perl provides some |
| 315 | additional properties that are written only in the single form, as well as |
| 316 | single-form short-cuts for all binary properties and certain others described |
| 317 | below, in which you may omit the property name and the equals or colon |
| 318 | separator. |
| 319 | |
| 320 | Most Unicode character properties have at least two synonyms (or aliases if you |
| 321 | prefer), a short one that is easier to type, and a longer one which is more |
| 322 | descriptive and hence it is easier to understand what it means. Thus the "L" |
| 323 | and "Letter" above are equivalent and can be used interchangeably. Likewise, |
| 324 | "Upper" is a synonym for "Uppercase", and we could have written |
| 325 | C<\p{Uppercase}> equivalently as C<\p{Upper}>. Also, there are typically |
| 326 | various synonyms for the values the property can be. For binary properties, |
| 327 | "True" has 3 synonyms: "T", "Yes", and "Y"; and "False has correspondingly "F", |
| 328 | "No", and "N". But be careful. A short form of a value for one property may |
| 329 | not mean the same thing as the same short form for another. Thus, for the |
| 330 | General_Category property, "L" means "Letter", but for the Bidi_Class property, |
| 331 | "L" means "Left". A complete list of properties and synonyms is in |
| 332 | L<perluniprops>. |
| 333 | |
| 334 | Upper/lower case differences in the property names and values are irrelevant, |
| 335 | thus C<\p{Upper}> means the same thing as C<\p{upper}> or even C<\p{UpPeR}>. |
| 336 | Similarly, you can add or subtract underscores anywhere in the middle of a |
| 337 | word, so that these are also equivalent to C<\p{U_p_p_e_r}>. And white space |
| 338 | is irrelevant adjacent to non-word characters, such as the braces and the equals |
| 339 | or colon separators so C<\p{ Upper }> and C<\p{ Upper_case : Y }> are |
| 340 | equivalent to these as well. In fact, in most cases, white space and even |
| 341 | hyphens can be added or deleted anywhere. So even C<\p{ Up-per case = Yes}> is |
| 342 | equivalent. All this is called "loose-matching" by Unicode. The few places |
| 343 | where stricter matching is employed is in the middle of numbers, and the Perl |
| 344 | extension properties that begin or end with an underscore. Stricter matching |
| 345 | cares about white space (except adjacent to the non-word characters) and |
| 346 | hyphens, and non-interior underscores. |
| 347 | |
| 348 | You can also use negation in both C<\p{}> and C<\P{}> by introducing a caret |
| 349 | (^) between the first brace and the property name: C<\p{^Tamil}> is |
| 350 | equal to C<\P{Tamil}>. |
| 351 | |
| 352 | =head3 B<General_Category> |
| 353 | |
| 354 | Every Unicode character is assigned a general category, which is the "most |
| 355 | usual categorization of a character" (from |
| 356 | L<http://www.unicode.org/reports/tr44>). |
| 357 | |
| 358 | The compound way of writing these is like C<\p{General_Category=Number}> |
| 359 | (short, C<\p{gc:n}>). But Perl furnishes shortcuts in which everything up |
| 360 | through the equal or colon separator is omitted. So you can instead just write |
| 361 | C<\pN>. |
| 362 | |
| 363 | Here are the short and long forms of the General Category properties: |
| 364 | |
| 365 | Short Long |
| 366 | |
| 367 | L Letter |
| 368 | LC, L& Cased_Letter (that is: [\p{Ll}\p{Lu}\p{Lt}]) |
| 369 | Lu Uppercase_Letter |
| 370 | Ll Lowercase_Letter |
| 371 | Lt Titlecase_Letter |
| 372 | Lm Modifier_Letter |
| 373 | Lo Other_Letter |
| 374 | |
| 375 | M Mark |
| 376 | Mn Nonspacing_Mark |
| 377 | Mc Spacing_Mark |
| 378 | Me Enclosing_Mark |
| 379 | |
| 380 | N Number |
| 381 | Nd Decimal_Number (also Digit) |
| 382 | Nl Letter_Number |
| 383 | No Other_Number |
| 384 | |
| 385 | P Punctuation (also Punct) |
| 386 | Pc Connector_Punctuation |
| 387 | Pd Dash_Punctuation |
| 388 | Ps Open_Punctuation |
| 389 | Pe Close_Punctuation |
| 390 | Pi Initial_Punctuation |
| 391 | (may behave like Ps or Pe depending on usage) |
| 392 | Pf Final_Punctuation |
| 393 | (may behave like Ps or Pe depending on usage) |
| 394 | Po Other_Punctuation |
| 395 | |
| 396 | S Symbol |
| 397 | Sm Math_Symbol |
| 398 | Sc Currency_Symbol |
| 399 | Sk Modifier_Symbol |
| 400 | So Other_Symbol |
| 401 | |
| 402 | Z Separator |
| 403 | Zs Space_Separator |
| 404 | Zl Line_Separator |
| 405 | Zp Paragraph_Separator |
| 406 | |
| 407 | C Other |
| 408 | Cc Control (also Cntrl) |
| 409 | Cf Format |
| 410 | Cs Surrogate (not usable) |
| 411 | Co Private_Use |
| 412 | Cn Unassigned |
| 413 | |
| 414 | Single-letter properties match all characters in any of the |
| 415 | two-letter sub-properties starting with the same letter. |
| 416 | C<LC> and C<L&> are special cases, which are both aliases for the set consisting of everything matched by C<Ll>, C<Lu>, and C<Lt>. |
| 417 | |
| 418 | Because Perl hides the need for the user to understand the internal |
| 419 | representation of Unicode characters, there is no need to implement |
| 420 | the somewhat messy concept of surrogates. C<Cs> is therefore not |
| 421 | supported. |
| 422 | |
| 423 | =head3 B<Bidirectional Character Types> |
| 424 | |
| 425 | Because scripts differ in their directionality (Hebrew is |
| 426 | written right to left, for example) Unicode supplies these properties in |
| 427 | the Bidi_Class class: |
| 428 | |
| 429 | Property Meaning |
| 430 | |
| 431 | L Left-to-Right |
| 432 | LRE Left-to-Right Embedding |
| 433 | LRO Left-to-Right Override |
| 434 | R Right-to-Left |
| 435 | AL Arabic Letter |
| 436 | RLE Right-to-Left Embedding |
| 437 | RLO Right-to-Left Override |
| 438 | PDF Pop Directional Format |
| 439 | EN European Number |
| 440 | ES European Separator |
| 441 | ET European Terminator |
| 442 | AN Arabic Number |
| 443 | CS Common Separator |
| 444 | NSM Non-Spacing Mark |
| 445 | BN Boundary Neutral |
| 446 | B Paragraph Separator |
| 447 | S Segment Separator |
| 448 | WS Whitespace |
| 449 | ON Other Neutrals |
| 450 | |
| 451 | This property is always written in the compound form. |
| 452 | For example, C<\p{Bidi_Class:R}> matches characters that are normally |
| 453 | written right to left. |
| 454 | |
| 455 | =head3 B<Scripts> |
| 456 | |
| 457 | The world's languages are written in a number of scripts. This sentence |
| 458 | (unless you're reading it in translation) is written in Latin, while Russian is |
| 459 | written in Cyrllic, and Greek is written in, well, Greek; Japanese mainly in |
| 460 | Hiragana or Katakana. There are many more. |
| 461 | |
| 462 | The Unicode Script property gives what script a given character is in, |
| 463 | and the property can be specified with the compound form like |
| 464 | C<\p{Script=Hebrew}> (short: C<\p{sc=hebr}>). Perl furnishes shortcuts for all |
| 465 | script names. You can omit everything up through the equals (or colon), and |
| 466 | simply write C<\p{Latin}> or C<\P{Cyrillic}>. |
| 467 | |
| 468 | A complete list of scripts and their shortcuts is in L<perluniprops>. |
| 469 | |
| 470 | =head3 B<Use of "Is" Prefix> |
| 471 | |
| 472 | For backward compatibility (with Perl 5.6), all properties mentioned |
| 473 | so far may have C<Is> or C<Is_> prepended to their name, so C<\P{Is_Lu}>, for |
| 474 | example, is equal to C<\P{Lu}>, and C<\p{IsScript:Arabic}> is equal to |
| 475 | C<\p{Arabic}>. |
| 476 | |
| 477 | =head3 B<Blocks> |
| 478 | |
| 479 | In addition to B<scripts>, Unicode also defines B<blocks> of |
| 480 | characters. The difference between scripts and blocks is that the |
| 481 | concept of scripts is closer to natural languages, while the concept |
| 482 | of blocks is more of an artificial grouping based on groups of Unicode |
| 483 | characters with consecutive ordinal values. For example, the "Basic Latin" |
| 484 | block is all characters whose ordinals are between 0 and 127, inclusive, in |
| 485 | other words, the ASCII characters. The "Latin" script contains some letters |
| 486 | from this block as well as several more, like "Latin-1 Supplement", |
| 487 | "Latin Extended-A", etc., but it does not contain all the characters from |
| 488 | those blocks. It does not, for example, contain digits, because digits are |
| 489 | shared across many scripts. Digits and similar groups, like punctuation, are in |
| 490 | the script called C<Common>. There is also a script called C<Inherited> for |
| 491 | characters that modify other characters, and inherit the script value of the |
| 492 | controlling character. |
| 493 | |
| 494 | For more about scripts versus blocks, see UAX#24 "Unicode Script Property": |
| 495 | L<http://www.unicode.org/reports/tr24> |
| 496 | |
| 497 | The Script property is likely to be the one you want to use when processing |
| 498 | natural language; the Block property may be useful in working with the nuts and |
| 499 | bolts of Unicode. |
| 500 | |
| 501 | Block names are matched in the compound form, like C<\p{Block: Arrows}> or |
| 502 | C<\p{Blk=Hebrew}>. Unlike most other properties only a few block names have a |
| 503 | Unicode-defined short name. But Perl does provide a (slight) shortcut: You |
| 504 | can say, for example C<\p{In_Arrows}> or C<\p{In_Hebrew}>. For backwards |
| 505 | compatibility, the C<In> prefix may be omitted if there is no naming conflict |
| 506 | with a script or any other property, and you can even use an C<Is> prefix |
| 507 | instead in those cases. But it is not a good idea to do this, for a couple |
| 508 | reasons: |
| 509 | |
| 510 | =over 4 |
| 511 | |
| 512 | =item 1 |
| 513 | |
| 514 | It is confusing. There are many naming conflicts, and you may forget some. |
| 515 | For example, C<\p{Hebrew}> means the I<script> Hebrew, and NOT the I<block> |
| 516 | Hebrew. But would you remember that 6 months from now? |
| 517 | |
| 518 | =item 2 |
| 519 | |
| 520 | It is unstable. A new version of Unicode may pre-empt the current meaning by |
| 521 | creating a property with the same name. There was a time in very early Unicode |
| 522 | releases when C<\p{Hebrew}> would have matched the I<block> Hebrew; now it |
| 523 | doesn't. |
| 524 | |
| 525 | =back |
| 526 | |
| 527 | Some people just prefer to always use C<\p{Block: foo}> and C<\p{Script: bar}> |
| 528 | instead of the shortcuts, for clarity, and because they can't remember the |
| 529 | difference between 'In' and 'Is' anyway (or aren't confident that those who |
| 530 | eventually will read their code will know). |
| 531 | |
| 532 | A complete list of blocks and their shortcuts is in L<perluniprops>. |
| 533 | |
| 534 | =head3 B<Other Properties> |
| 535 | |
| 536 | There are many more properties than the very basic ones described here. |
| 537 | A complete list is in L<perluniprops>. |
| 538 | |
| 539 | Unicode defines all its properties in the compound form, so all single-form |
| 540 | properties are Perl extensions. A number of these are just synonyms for the |
| 541 | Unicode ones, but some are genunine extensions, including a couple that are in |
| 542 | the compound form. And quite a few of these are actually recommended by Unicode |
| 543 | (in L<http://www.unicode.org/reports/tr18>). |
| 544 | |
| 545 | This section gives some details on all the extensions that aren't synonyms for |
| 546 | compound-form Unicode properties (for those, you'll have to refer to the |
| 547 | L<Unicode Standard|http://www.unicode.org/reports/tr44>. |
| 548 | |
| 549 | =over |
| 550 | |
| 551 | =item B<C<\p{All}>> |
| 552 | |
| 553 | This matches any of the 1_114_112 Unicode code points. It is a synonym for |
| 554 | C<\p{Any}>. |
| 555 | |
| 556 | =item B<C<\p{Alnum}>> |
| 557 | |
| 558 | This matches any C<\p{Alphabetic}> or C<\p{Decimal_Number}> character. |
| 559 | |
| 560 | =item B<C<\p{Any}>> |
| 561 | |
| 562 | This matches any of the 1_114_112 Unicode code points. It is a synonym for |
| 563 | C<\p{All}>. |
| 564 | |
| 565 | =item B<C<\p{Assigned}>> |
| 566 | |
| 567 | This matches any assigned code point; that is, any code point whose general |
| 568 | category is not Unassigned (or equivalently, not Cn). |
| 569 | |
| 570 | =item B<C<\p{Blank}>> |
| 571 | |
| 572 | This is the same as C<\h> and C<\p{HorizSpace}>: A character that changes the |
| 573 | spacing horizontally. |
| 574 | |
| 575 | =item B<C<\p{Decomposition_Type: Non_Canonical}>> (Short: C<\p{Dt=NonCanon}>) |
| 576 | |
| 577 | Matches a character that has a non-canonical decomposition. |
| 578 | |
| 579 | To understand the use of this rarely used property=value combination, it is |
| 580 | necessary to know some basics about decomposition. |
| 581 | Consider a character, say H. It could appear with various marks around it, |
| 582 | such as an acute accent, or a circumflex, or various hooks, circles, arrows, |
| 583 | I<etc.>, above, below, to one side and/or the other, etc. There are many |
| 584 | possibilities among the world's languages. The number of combinations is |
| 585 | astronomical, and if there were a character for each combination, it would |
| 586 | soon exhaust Unicode's more than a million possible characters. So Unicode |
| 587 | took a different approach: there is a character for the base H, and a |
| 588 | character for each of the possible marks, and they can be combined variously |
| 589 | to get a final logical character. So a logical character--what appears to be a |
| 590 | single character--can be a sequence of more than one individual characters. |
| 591 | This is called an "extended grapheme cluster". (Perl furnishes the C<\X> |
| 592 | construct to match such sequences.) |
| 593 | |
| 594 | But Unicode's intent is to unify the existing character set standards and |
| 595 | practices, and a number of pre-existing standards have single characters that |
| 596 | mean the same thing as some of these combinations. An example is ISO-8859-1, |
| 597 | which has quite a few of these in the Latin-1 range, an example being "LATIN |
| 598 | CAPITAL LETTER E WITH ACUTE". Because this character was in this pre-existing |
| 599 | standard, Unicode added it to its repertoire. But this character is considered |
| 600 | by Unicode to be equivalent to the sequence consisting of first the character |
| 601 | "LATIN CAPITAL LETTER E", then the character "COMBINING ACUTE ACCENT". |
| 602 | |
| 603 | "LATIN CAPITAL LETTER E WITH ACUTE" is called a "pre-composed" character, and |
| 604 | the equivalence with the sequence is called canonical equivalence. All |
| 605 | pre-composed characters are said to have a decomposition (into the equivalent |
| 606 | sequence) and the decomposition type is also called canonical. |
| 607 | |
| 608 | However, many more characters have a different type of decomposition, a |
| 609 | "compatible" or "non-canonical" decomposition. The sequences that form these |
| 610 | decompositions are not considered canonically equivalent to the pre-composed |
| 611 | character. An example, again in the Latin-1 range, is the "SUPERSCRIPT ONE". |
| 612 | It is kind of like a regular digit 1, but not exactly; its decomposition |
| 613 | into the digit 1 is called a "compatible" decomposition, specifically a |
| 614 | "super" decomposition. There are several such compatibility |
| 615 | decompositions (see L<http://www.unicode.org/reports/tr44>), including one |
| 616 | called "compat" which means some miscellaneous type of decomposition |
| 617 | that doesn't fit into the decomposition categories that Unicode has chosen. |
| 618 | |
| 619 | Note that most Unicode characters don't have a decomposition, so their |
| 620 | decomposition type is "None". |
| 621 | |
| 622 | Perl has added the C<Non_Canonical> type, for your convenience, to mean any of |
| 623 | the compatibility decompositions. |
| 624 | |
| 625 | =item B<C<\p{Graph}>> |
| 626 | |
| 627 | Matches any character that is graphic. Theoretically, this means a character |
| 628 | that on a printer would cause ink to be used. |
| 629 | |
| 630 | =item B<C<\p{HorizSpace}>> |
| 631 | |
| 632 | This is the same as C<\h> and C<\p{Blank}>: A character that changes the |
| 633 | spacing horizontally. |
| 634 | |
| 635 | =item B<C<\p{In=*}>> |
| 636 | |
| 637 | This is a synonym for C<\p{Present_In=*}> |
| 638 | |
| 639 | =item B<C<\p{PerlSpace}>> |
| 640 | |
| 641 | This is the same as C<\s>, restricted to ASCII, namely C<S<[ \f\n\r\t]>>. |
| 642 | |
| 643 | Mnemonic: Perl's (original) space |
| 644 | |
| 645 | =item B<C<\p{PerlWord}>> |
| 646 | |
| 647 | This is the same as C<\w>, restricted to ASCII, namely C<[A-Za-z0-9_]> |
| 648 | |
| 649 | Mnemonic: Perl's (original) word. |
| 650 | |
| 651 | =item B<C<\p{PosixAlnum}>> |
| 652 | |
| 653 | This matches any alphanumeric character in the ASCII range, namely |
| 654 | C<[A-Za-z0-9]>. |
| 655 | |
| 656 | =item B<C<\p{PosixAlpha}>> |
| 657 | |
| 658 | This matches any alphabetic character in the ASCII range, namely C<[A-Za-z]>. |
| 659 | |
| 660 | =item B<C<\p{PosixBlank}>> |
| 661 | |
| 662 | This matches any blank character in the ASCII range, namely C<S<[ \t]>>. |
| 663 | |
| 664 | =item B<C<\p{PosixCntrl}>> |
| 665 | |
| 666 | This matches any control character in the ASCII range, namely C<[\x00-\x1F\x7F]> |
| 667 | |
| 668 | =item B<C<\p{PosixDigit}>> |
| 669 | |
| 670 | This matches any digit character in the ASCII range, namely C<[0-9]>. |
| 671 | |
| 672 | =item B<C<\p{PosixGraph}>> |
| 673 | |
| 674 | This matches any graphical character in the ASCII range, namely C<[\x21-\x7E]>. |
| 675 | |
| 676 | =item B<C<\p{PosixLower}>> |
| 677 | |
| 678 | This matches any lowercase character in the ASCII range, namely C<[a-z]>. |
| 679 | |
| 680 | =item B<C<\p{PosixPrint}>> |
| 681 | |
| 682 | This matches any printable character in the ASCII range, namely C<[\x20-\x7E]>. |
| 683 | These are the graphical characters plus SPACE. |
| 684 | |
| 685 | =item B<C<\p{PosixPunct}>> |
| 686 | |
| 687 | This matches any punctuation character in the ASCII range, namely |
| 688 | C<[\x21-\x2F\x3A-\x40\x5B-\x60\x7B-\x7E]>. These are the |
| 689 | graphical characters that aren't word characters. Note that the Posix standard |
| 690 | includes in its definition of punctuation, those characters that Unicode calls |
| 691 | "symbols." |
| 692 | |
| 693 | =item B<C<\p{PosixSpace}>> |
| 694 | |
| 695 | This matches any space character in the ASCII range, namely |
| 696 | C<S<[ \f\n\r\t\x0B]>> (the last being a vertical tab). |
| 697 | |
| 698 | =item B<C<\p{PosixUpper}>> |
| 699 | |
| 700 | This matches any uppercase character in the ASCII range, namely C<[A-Z]>. |
| 701 | |
| 702 | =item B<C<\p{Present_In: *}>> (Short: C<\p{In=*}>) |
| 703 | |
| 704 | This property is used when you need to know in what Unicode version(s) a |
| 705 | character is. |
| 706 | |
| 707 | The "*" above stands for some two digit Unicode version number, such as |
| 708 | C<1.1> or C<4.0>; or the "*" can also be C<Unassigned>. This property will |
| 709 | match the code points whose final disposition has been settled as of the |
| 710 | Unicode release given by the version number; C<\p{Present_In: Unassigned}> |
| 711 | will match those code points whose meaning has yet to be assigned. |
| 712 | |
| 713 | For example, C<U+0041> "LATIN CAPITAL LETTER A" was present in the very first |
| 714 | Unicode release available, which is C<1.1>, so this property is true for all |
| 715 | valid "*" versions. On the other hand, C<U+1EFF> was not assigned until version |
| 716 | 5.1 when it became "LATIN SMALL LETTER Y WITH LOOP", so the only "*" that |
| 717 | would match it are 5.1, 5.2, and later. |
| 718 | |
| 719 | Unicode furnishes the C<Age> property from which this is derived. The problem |
| 720 | with Age is that a strict interpretation of it (which Perl takes) has it |
| 721 | matching the precise release a code point's meaning is introduced in. Thus |
| 722 | C<U+0041> would match only 1.1; and C<U+1EFF> only 5.1. This is not usually what |
| 723 | you want. |
| 724 | |
| 725 | Some non-Perl implementations of the Age property may change its meaning to be |
| 726 | the same as the Perl Present_In property; just be aware of that. |
| 727 | |
| 728 | Another confusion with both these properties is that the definition is not |
| 729 | that the code point has been assigned, but that the meaning of the code point |
| 730 | has been determined. This is because 66 code points will always be |
| 731 | unassigned, and, so the Age for them is the Unicode version the decision to |
| 732 | make them so was made in. For example, C<U+FDD0> is to be permanently |
| 733 | unassigned to a character, and the decision to do that was made in version 3.1, |
| 734 | so C<\p{Age=3.1}> matches this character and C<\p{Present_In: 3.1}> and up |
| 735 | matches as well. |
| 736 | |
| 737 | =item B<C<\p{Print}>> |
| 738 | |
| 739 | This matches any character that is graphical or blank, except controls. |
| 740 | |
| 741 | =item B<C<\p{SpacePerl}>> |
| 742 | |
| 743 | This is the same as C<\s>, including beyond ASCII. |
| 744 | |
| 745 | Mnemonic: Space, as modified by Perl. (It doesn't include the vertical tab |
| 746 | which both the Posix standard and Unicode consider to be space.) |
| 747 | |
| 748 | =item B<C<\p{VertSpace}>> |
| 749 | |
| 750 | This is the same as C<\v>: A character that changes the spacing vertically. |
| 751 | |
| 752 | =item B<C<\p{Word}>> |
| 753 | |
| 754 | This is the same as C<\w>, including beyond ASCII. |
| 755 | |
| 756 | =back |
| 757 | |
| 758 | =head2 User-Defined Character Properties |
| 759 | |
| 760 | You can define your own binary character properties by defining subroutines |
| 761 | whose names begin with "In" or "Is". The subroutines can be defined in any |
| 762 | package. The user-defined properties can be used in the regular expression |
| 763 | C<\p> and C<\P> constructs; if you are using a user-defined property from a |
| 764 | package other than the one you are in, you must specify its package in the |
| 765 | C<\p> or C<\P> construct. |
| 766 | |
| 767 | # assuming property Is_Foreign defined in Lang:: |
| 768 | package main; # property package name required |
| 769 | if ($txt =~ /\p{Lang::IsForeign}+/) { ... } |
| 770 | |
| 771 | package Lang; # property package name not required |
| 772 | if ($txt =~ /\p{IsForeign}+/) { ... } |
| 773 | |
| 774 | |
| 775 | Note that the effect is compile-time and immutable once defined. |
| 776 | |
| 777 | The subroutines must return a specially-formatted string, with one |
| 778 | or more newline-separated lines. Each line must be one of the following: |
| 779 | |
| 780 | =over 4 |
| 781 | |
| 782 | =item * |
| 783 | |
| 784 | A single hexadecimal number denoting a Unicode code point to include. |
| 785 | |
| 786 | =item * |
| 787 | |
| 788 | Two hexadecimal numbers separated by horizontal whitespace (space or |
| 789 | tabular characters) denoting a range of Unicode code points to include. |
| 790 | |
| 791 | =item * |
| 792 | |
| 793 | Something to include, prefixed by "+": a built-in character |
| 794 | property (prefixed by "utf8::") or a user-defined character property, |
| 795 | to represent all the characters in that property; two hexadecimal code |
| 796 | points for a range; or a single hexadecimal code point. |
| 797 | |
| 798 | =item * |
| 799 | |
| 800 | Something to exclude, prefixed by "-": an existing character |
| 801 | property (prefixed by "utf8::") or a user-defined character property, |
| 802 | to represent all the characters in that property; two hexadecimal code |
| 803 | points for a range; or a single hexadecimal code point. |
| 804 | |
| 805 | =item * |
| 806 | |
| 807 | Something to negate, prefixed "!": an existing character |
| 808 | property (prefixed by "utf8::") or a user-defined character property, |
| 809 | to represent all the characters in that property; two hexadecimal code |
| 810 | points for a range; or a single hexadecimal code point. |
| 811 | |
| 812 | =item * |
| 813 | |
| 814 | Something to intersect with, prefixed by "&": an existing character |
| 815 | property (prefixed by "utf8::") or a user-defined character property, |
| 816 | for all the characters except the characters in the property; two |
| 817 | hexadecimal code points for a range; or a single hexadecimal code point. |
| 818 | |
| 819 | =back |
| 820 | |
| 821 | For example, to define a property that covers both the Japanese |
| 822 | syllabaries (hiragana and katakana), you can define |
| 823 | |
| 824 | sub InKana { |
| 825 | return <<END; |
| 826 | 3040\t309F |
| 827 | 30A0\t30FF |
| 828 | END |
| 829 | } |
| 830 | |
| 831 | Imagine that the here-doc end marker is at the beginning of the line. |
| 832 | Now you can use C<\p{InKana}> and C<\P{InKana}>. |
| 833 | |
| 834 | You could also have used the existing block property names: |
| 835 | |
| 836 | sub InKana { |
| 837 | return <<'END'; |
| 838 | +utf8::InHiragana |
| 839 | +utf8::InKatakana |
| 840 | END |
| 841 | } |
| 842 | |
| 843 | Suppose you wanted to match only the allocated characters, |
| 844 | not the raw block ranges: in other words, you want to remove |
| 845 | the non-characters: |
| 846 | |
| 847 | sub InKana { |
| 848 | return <<'END'; |
| 849 | +utf8::InHiragana |
| 850 | +utf8::InKatakana |
| 851 | -utf8::IsCn |
| 852 | END |
| 853 | } |
| 854 | |
| 855 | The negation is useful for defining (surprise!) negated classes. |
| 856 | |
| 857 | sub InNotKana { |
| 858 | return <<'END'; |
| 859 | !utf8::InHiragana |
| 860 | -utf8::InKatakana |
| 861 | +utf8::IsCn |
| 862 | END |
| 863 | } |
| 864 | |
| 865 | Intersection is useful for getting the common characters matched by |
| 866 | two (or more) classes. |
| 867 | |
| 868 | sub InFooAndBar { |
| 869 | return <<'END'; |
| 870 | +main::Foo |
| 871 | &main::Bar |
| 872 | END |
| 873 | } |
| 874 | |
| 875 | It's important to remember not to use "&" for the first set; that |
| 876 | would be intersecting with nothing (resulting in an empty set). |
| 877 | |
| 878 | =head2 User-Defined Case Mappings (for serious hackers only) |
| 879 | |
| 880 | You can also define your own mappings to be used in C<lc()>, |
| 881 | C<lcfirst()>, C<uc()>, and C<ucfirst()> (or their string-inlined versions, |
| 882 | C<\L>, C<\l>, C<\U>, and C<\u>). The mappings are currently only valid |
| 883 | on strings encoded in UTF-8, but see below for a partial workaround for |
| 884 | this restriction. |
| 885 | |
| 886 | The principle is similar to that of user-defined character |
| 887 | properties: define subroutines that do the mappings. |
| 888 | C<ToLower> is used for C<lc()>, C<\L>, C<lcfirst()>, and C<\l>; C<ToTitle> for |
| 889 | C<ucfirst()> and C<\u>; and C<ToUpper> for C<uc()> and C<\U>. |
| 890 | |
| 891 | C<ToUpper()> should look something like this: |
| 892 | |
| 893 | sub ToUpper { |
| 894 | return <<END; |
| 895 | 0061\t007A\t0041 |
| 896 | 0101\t\t0100 |
| 897 | END |
| 898 | } |
| 899 | |
| 900 | This sample C<ToUpper()> has the effect of mapping "a-z" to "A-Z", 0x101 |
| 901 | to 0x100, and all other characters map to themselves. The first |
| 902 | returned line means to map the code point at 0x61 ("a") to 0x41 ("A"), |
| 903 | the code point at 0x62 ("b") to 0x42 ("B"), ..., 0x7A ("z") to 0x5A |
| 904 | ("Z"). The second line maps just the code point 0x101 to 0x100. Since |
| 905 | there are no other mappings defined, all other code points map to |
| 906 | themselves. |
| 907 | |
| 908 | This mechanism is not well behaved as far as affecting other packages |
| 909 | and scopes. All non-threaded programs have exactly one uppercasing |
| 910 | behavior, one lowercasing behavior, and one titlecasing behavior in |
| 911 | effect for utf8-encoded strings for the duration of the program. Each |
| 912 | of these behaviors is irrevocably determined the first time the |
| 913 | corresponding function is called to change a utf8-encoded string's case. |
| 914 | If a corresponding C<To-> function has been defined in the package that |
| 915 | makes that first call, the mapping defined by that function will be the |
| 916 | mapping used for the duration of the program's execution across all |
| 917 | packages and scopes. If no corresponding C<To-> function has been |
| 918 | defined in that package, the standard official mapping will be used for |
| 919 | all packages and scopes, and any corresponding C<To-> function anywhere |
| 920 | will be ignored. Threaded programs have similar behavior. If the |
| 921 | program's casing behavior has been decided at the time of a thread's |
| 922 | creation, the thread will inherit that behavior. But, if the behavior |
| 923 | hasn't been decided, the thread gets to decide for itself, and its |
| 924 | decision does not affect other threads nor its creator. |
| 925 | |
| 926 | As shown by the example above, you have to furnish a complete mapping; |
| 927 | you can't just override a couple of characters and leave the rest |
| 928 | unchanged. You can find all the official mappings in the directory |
| 929 | C<$Config{privlib}>F</unicore/To/>. The mapping data is returned as the |
| 930 | here-document. The C<utf8::ToSpecI<Foo>> hashes in those files are special |
| 931 | exception mappings derived from |
| 932 | C<$Config{privlib}>F</unicore/SpecialCasing.txt>. (The "Digit" and |
| 933 | "Fold" mappings that one can see in the directory are not directly |
| 934 | user-accessible, one can use either the L<Unicode::UCD> module, or just match |
| 935 | case-insensitively, which is what uses the "Fold" mapping. Neither are user |
| 936 | overridable.) |
| 937 | |
| 938 | If you have many mappings to change, you can take the official mapping data, |
| 939 | change by hand the affected code points, and place the whole thing into your |
| 940 | subroutine. But this will only be valid on Perls that use the same Unicode |
| 941 | version. Another option would be to have your subroutine read the official |
| 942 | mapping file(s) and overwrite the affected code points. |
| 943 | |
| 944 | If you have only a few mappings to change you can use the |
| 945 | following trick (but see below for a big caveat), here illustrated for |
| 946 | Turkish: |
| 947 | |
| 948 | use Config; |
| 949 | use charnames ":full"; |
| 950 | |
| 951 | sub ToUpper { |
| 952 | my $official = do "$Config{privlib}/unicore/To/Upper.pl"; |
| 953 | $utf8::ToSpecUpper{'i'} = |
| 954 | "\N{LATIN CAPITAL LETTER I WITH DOT ABOVE}"; |
| 955 | return $official; |
| 956 | } |
| 957 | |
| 958 | This takes the official mappings and overrides just one, for "LATIN SMALL |
| 959 | LETTER I". Each hash key must be the string of bytes that form the UTF-8 |
| 960 | (on EBCDIC platforms, UTF-EBCDIC) of the character, as illustrated by |
| 961 | the inverse function. |
| 962 | |
| 963 | sub ToLower { |
| 964 | my $official = do $lower; |
| 965 | $utf8::ToSpecLower{"\xc4\xb0"} = "i"; |
| 966 | return $official; |
| 967 | } |
| 968 | |
| 969 | This example is for an ASCII platform, and C<\xc4\xb0> is the string of |
| 970 | bytes that together form the UTF-8 that represents C<\N{LATIN CAPITAL |
| 971 | LETTER I WITH DOT ABOVE}>, C<U+0130>. You can avoid having to figure out |
| 972 | these bytes, and at the same time make it work on all platforms by |
| 973 | instead writing: |
| 974 | |
| 975 | sub ToLower { |
| 976 | my $official = do $lower; |
| 977 | my $sequence = "\N{LATIN CAPITAL LETTER I WITH DOT ABOVE}"; |
| 978 | utf8::encode($sequence); |
| 979 | $utf8::ToSpecLower{$sequence} = "i"; |
| 980 | return $official; |
| 981 | } |
| 982 | |
| 983 | This works because C<utf8::encode()> takes the single character and |
| 984 | converts it to the sequence of bytes that constitute it. Note that we took |
| 985 | advantage of the fact that C<"i"> is the same in UTF-8 or UTF_EBCIDIC as not; |
| 986 | otherwise we would have had to write |
| 987 | |
| 988 | $utf8::ToSpecLower{$sequence} = "\N{LATIN SMALL LETTER I}"; |
| 989 | |
| 990 | in the ToLower example, and in the ToUpper example, use |
| 991 | |
| 992 | my $sequence = "\N{LATIN SMALL LETTER I}"; |
| 993 | utf8::encode($sequence); |
| 994 | |
| 995 | A big caveat to the above trick, and to this whole mechanism in general, |
| 996 | is that they work only on strings encoded in UTF-8. You can partially |
| 997 | get around this by using C<use subs>. For example: |
| 998 | |
| 999 | use subs qw(uc ucfirst lc lcfirst); |
| 1000 | |
| 1001 | sub uc($) { |
| 1002 | my $string = shift; |
| 1003 | utf8::upgrade($string); |
| 1004 | return CORE::uc($string); |
| 1005 | } |
| 1006 | |
| 1007 | sub lc($) { |
| 1008 | my $string = shift; |
| 1009 | utf8::upgrade($string); |
| 1010 | |
| 1011 | # Unless an I is before a dot_above, it turns into a dotless i. |
| 1012 | # (The character class with the combining classes matches non-above |
| 1013 | # marks following the I. Any number of these may be between the 'I' and |
| 1014 | # the dot_above, and the dot_above will still apply to the 'I'. |
| 1015 | use charnames ":full"; |
| 1016 | $string =~ |
| 1017 | s/I |
| 1018 | (?! [^\p{ccc=0}\p{ccc=Above}]* \N{COMBINING DOT ABOVE} ) |
| 1019 | /\N{LATIN SMALL LETTER DOTLESS I}/gx; |
| 1020 | |
| 1021 | # But when the I is followed by a dot_above, remove the |
| 1022 | # dot_above so the end result will be i. |
| 1023 | $string =~ s/I |
| 1024 | ([^\p{ccc=0}\p{ccc=Above}]* ) |
| 1025 | \N{COMBINING DOT ABOVE} |
| 1026 | /i$1/gx; |
| 1027 | return CORE::lc($string); |
| 1028 | } |
| 1029 | |
| 1030 | These examples (also for Turkish) make sure the input is in UTF-8, and then |
| 1031 | call the corresponding official function, which will use the C<ToUpper()> and |
| 1032 | C<ToLower()> functions you have defined. |
| 1033 | (For Turkish, there are other required functions: C<ucfirst>, C<lcfirst>, |
| 1034 | and C<ToTitle>. These are very similar to the ones given above.) |
| 1035 | |
| 1036 | The reason this is a partial work-around is that it doesn't affect the C<\l>, |
| 1037 | C<\L>, C<\u>, and C<\U> case change operations, which still require the source |
| 1038 | to be encoded in utf8 (see L</The "Unicode Bug">). |
| 1039 | |
| 1040 | The C<lc()> example shows how you can add context-dependent casing. Note |
| 1041 | that context-dependent casing suffers from the problem that the string |
| 1042 | passed to the casing function may not have sufficient context to make |
| 1043 | the proper choice. And, it will not be called for C<\l>, C<\L>, C<\u>, |
| 1044 | and C<\U>. |
| 1045 | |
| 1046 | =head2 Character Encodings for Input and Output |
| 1047 | |
| 1048 | See L<Encode>. |
| 1049 | |
| 1050 | =head2 Unicode Regular Expression Support Level |
| 1051 | |
| 1052 | The following list of Unicode support for regular expressions describes |
| 1053 | all the features currently supported. The references to "Level N" |
| 1054 | and the section numbers refer to the Unicode Technical Standard #18, |
| 1055 | "Unicode Regular Expressions", version 11, in May 2005. |
| 1056 | |
| 1057 | =over 4 |
| 1058 | |
| 1059 | =item * |
| 1060 | |
| 1061 | Level 1 - Basic Unicode Support |
| 1062 | |
| 1063 | RL1.1 Hex Notation - done [1] |
| 1064 | RL1.2 Properties - done [2][3] |
| 1065 | RL1.2a Compatibility Properties - done [4] |
| 1066 | RL1.3 Subtraction and Intersection - MISSING [5] |
| 1067 | RL1.4 Simple Word Boundaries - done [6] |
| 1068 | RL1.5 Simple Loose Matches - done [7] |
| 1069 | RL1.6 Line Boundaries - MISSING [8] |
| 1070 | RL1.7 Supplementary Code Points - done [9] |
| 1071 | |
| 1072 | [1] \x{...} |
| 1073 | [2] \p{...} \P{...} |
| 1074 | [3] supports not only minimal list, but all Unicode character |
| 1075 | properties (see L</Unicode Character Properties>) |
| 1076 | [4] \d \D \s \S \w \W \X [:prop:] [:^prop:] |
| 1077 | [5] can use regular expression look-ahead [a] or |
| 1078 | user-defined character properties [b] to emulate set |
| 1079 | operations |
| 1080 | [6] \b \B |
| 1081 | [7] note that Perl does Full case-folding in matching (but with |
| 1082 | bugs), not Simple: for example U+1F88 is equivalent to |
| 1083 | U+1F00 U+03B9, not with 1F80. This difference matters |
| 1084 | mainly for certain Greek capital letters with certain |
| 1085 | modifiers: the Full case-folding decomposes the letter, |
| 1086 | while the Simple case-folding would map it to a single |
| 1087 | character. |
| 1088 | [8] should do ^ and $ also on U+000B (\v in C), FF (\f), CR |
| 1089 | (\r), CRLF (\r\n), NEL (U+0085), LS (U+2028), and PS |
| 1090 | (U+2029); should also affect <>, $., and script line |
| 1091 | numbers; should not split lines within CRLF [c] (i.e. there |
| 1092 | is no empty line between \r and \n) |
| 1093 | [9] UTF-8/UTF-EBDDIC used in perl allows not only U+10000 to |
| 1094 | U+10FFFF but also beyond U+10FFFF [d] |
| 1095 | |
| 1096 | [a] You can mimic class subtraction using lookahead. |
| 1097 | For example, what UTS#18 might write as |
| 1098 | |
| 1099 | [{Greek}-[{UNASSIGNED}]] |
| 1100 | |
| 1101 | in Perl can be written as: |
| 1102 | |
| 1103 | (?!\p{Unassigned})\p{InGreekAndCoptic} |
| 1104 | (?=\p{Assigned})\p{InGreekAndCoptic} |
| 1105 | |
| 1106 | But in this particular example, you probably really want |
| 1107 | |
| 1108 | \p{GreekAndCoptic} |
| 1109 | |
| 1110 | which will match assigned characters known to be part of the Greek script. |
| 1111 | |
| 1112 | Also see the Unicode::Regex::Set module, it does implement the full |
| 1113 | UTS#18 grouping, intersection, union, and removal (subtraction) syntax. |
| 1114 | |
| 1115 | [b] '+' for union, '-' for removal (set-difference), '&' for intersection |
| 1116 | (see L</"User-Defined Character Properties">) |
| 1117 | |
| 1118 | [c] Try the C<:crlf> layer (see L<PerlIO>). |
| 1119 | |
| 1120 | [d] U+FFFF will currently generate a warning message if 'utf8' warnings are |
| 1121 | enabled |
| 1122 | |
| 1123 | =item * |
| 1124 | |
| 1125 | Level 2 - Extended Unicode Support |
| 1126 | |
| 1127 | RL2.1 Canonical Equivalents - MISSING [10][11] |
| 1128 | RL2.2 Default Grapheme Clusters - MISSING [12] |
| 1129 | RL2.3 Default Word Boundaries - MISSING [14] |
| 1130 | RL2.4 Default Loose Matches - MISSING [15] |
| 1131 | RL2.5 Name Properties - MISSING [16] |
| 1132 | RL2.6 Wildcard Properties - MISSING |
| 1133 | |
| 1134 | [10] see UAX#15 "Unicode Normalization Forms" |
| 1135 | [11] have Unicode::Normalize but not integrated to regexes |
| 1136 | [12] have \X but we don't have a "Grapheme Cluster Mode" |
| 1137 | [14] see UAX#29, Word Boundaries |
| 1138 | [15] see UAX#21 "Case Mappings" |
| 1139 | [16] missing loose match [e] |
| 1140 | |
| 1141 | [e] C<\N{...}> allows namespaces (see L<charnames>). |
| 1142 | |
| 1143 | =item * |
| 1144 | |
| 1145 | Level 3 - Tailored Support |
| 1146 | |
| 1147 | RL3.1 Tailored Punctuation - MISSING |
| 1148 | RL3.2 Tailored Grapheme Clusters - MISSING [17][18] |
| 1149 | RL3.3 Tailored Word Boundaries - MISSING |
| 1150 | RL3.4 Tailored Loose Matches - MISSING |
| 1151 | RL3.5 Tailored Ranges - MISSING |
| 1152 | RL3.6 Context Matching - MISSING [19] |
| 1153 | RL3.7 Incremental Matches - MISSING |
| 1154 | ( RL3.8 Unicode Set Sharing ) |
| 1155 | RL3.9 Possible Match Sets - MISSING |
| 1156 | RL3.10 Folded Matching - MISSING [20] |
| 1157 | RL3.11 Submatchers - MISSING |
| 1158 | |
| 1159 | [17] see UAX#10 "Unicode Collation Algorithms" |
| 1160 | [18] have Unicode::Collate but not integrated to regexes |
| 1161 | [19] have (?<=x) and (?=x), but look-aheads or look-behinds |
| 1162 | should see outside of the target substring |
| 1163 | [20] need insensitive matching for linguistic features other |
| 1164 | than case; for example, hiragana to katakana, wide and |
| 1165 | narrow, simplified Han to traditional Han (see UTR#30 |
| 1166 | "Character Foldings") |
| 1167 | |
| 1168 | =back |
| 1169 | |
| 1170 | =head2 Unicode Encodings |
| 1171 | |
| 1172 | Unicode characters are assigned to I<code points>, which are abstract |
| 1173 | numbers. To use these numbers, various encodings are needed. |
| 1174 | |
| 1175 | =over 4 |
| 1176 | |
| 1177 | =item * |
| 1178 | |
| 1179 | UTF-8 |
| 1180 | |
| 1181 | UTF-8 is a variable-length (1 to 6 bytes, current character allocations |
| 1182 | require 4 bytes), byte-order independent encoding. For ASCII (and we |
| 1183 | really do mean 7-bit ASCII, not another 8-bit encoding), UTF-8 is |
| 1184 | transparent. |
| 1185 | |
| 1186 | The following table is from Unicode 3.2. |
| 1187 | |
| 1188 | Code Points 1st Byte 2nd Byte 3rd Byte 4th Byte |
| 1189 | |
| 1190 | U+0000..U+007F 00..7F |
| 1191 | U+0080..U+07FF * C2..DF 80..BF |
| 1192 | U+0800..U+0FFF E0 * A0..BF 80..BF |
| 1193 | U+1000..U+CFFF E1..EC 80..BF 80..BF |
| 1194 | U+D000..U+D7FF ED 80..9F 80..BF |
| 1195 | U+D800..U+DFFF +++++++ utf16 surrogates, not legal utf8 +++++++ |
| 1196 | U+E000..U+FFFF EE..EF 80..BF 80..BF |
| 1197 | U+10000..U+3FFFF F0 * 90..BF 80..BF 80..BF |
| 1198 | U+40000..U+FFFFF F1..F3 80..BF 80..BF 80..BF |
| 1199 | U+100000..U+10FFFF F4 80..8F 80..BF 80..BF |
| 1200 | |
| 1201 | Note the gaps before several of the byte entries above marked by '*'. These are |
| 1202 | caused by legal UTF-8 avoiding non-shortest encodings: it is technically |
| 1203 | possible to UTF-8-encode a single code point in different ways, but that is |
| 1204 | explicitly forbidden, and the shortest possible encoding should always be used |
| 1205 | (and that is what Perl does). |
| 1206 | |
| 1207 | Another way to look at it is via bits: |
| 1208 | |
| 1209 | Code Points 1st Byte 2nd Byte 3rd Byte 4th Byte |
| 1210 | |
| 1211 | 0aaaaaaa 0aaaaaaa |
| 1212 | 00000bbbbbaaaaaa 110bbbbb 10aaaaaa |
| 1213 | ccccbbbbbbaaaaaa 1110cccc 10bbbbbb 10aaaaaa |
| 1214 | 00000dddccccccbbbbbbaaaaaa 11110ddd 10cccccc 10bbbbbb 10aaaaaa |
| 1215 | |
| 1216 | As you can see, the continuation bytes all begin with "10", and the |
| 1217 | leading bits of the start byte tell how many bytes there are in the |
| 1218 | encoded character. |
| 1219 | |
| 1220 | =item * |
| 1221 | |
| 1222 | UTF-EBCDIC |
| 1223 | |
| 1224 | Like UTF-8 but EBCDIC-safe, in the way that UTF-8 is ASCII-safe. |
| 1225 | |
| 1226 | =item * |
| 1227 | |
| 1228 | UTF-16, UTF-16BE, UTF-16LE, Surrogates, and BOMs (Byte Order Marks) |
| 1229 | |
| 1230 | The followings items are mostly for reference and general Unicode |
| 1231 | knowledge, Perl doesn't use these constructs internally. |
| 1232 | |
| 1233 | UTF-16 is a 2 or 4 byte encoding. The Unicode code points |
| 1234 | C<U+0000..U+FFFF> are stored in a single 16-bit unit, and the code |
| 1235 | points C<U+10000..U+10FFFF> in two 16-bit units. The latter case is |
| 1236 | using I<surrogates>, the first 16-bit unit being the I<high |
| 1237 | surrogate>, and the second being the I<low surrogate>. |
| 1238 | |
| 1239 | Surrogates are code points set aside to encode the C<U+10000..U+10FFFF> |
| 1240 | range of Unicode code points in pairs of 16-bit units. The I<high |
| 1241 | surrogates> are the range C<U+D800..U+DBFF> and the I<low surrogates> |
| 1242 | are the range C<U+DC00..U+DFFF>. The surrogate encoding is |
| 1243 | |
| 1244 | $hi = ($uni - 0x10000) / 0x400 + 0xD800; |
| 1245 | $lo = ($uni - 0x10000) % 0x400 + 0xDC00; |
| 1246 | |
| 1247 | and the decoding is |
| 1248 | |
| 1249 | $uni = 0x10000 + ($hi - 0xD800) * 0x400 + ($lo - 0xDC00); |
| 1250 | |
| 1251 | If you try to generate surrogates (for example by using chr()), you |
| 1252 | will get a warning, if warnings are turned on, because those code |
| 1253 | points are not valid for a Unicode character. |
| 1254 | |
| 1255 | Because of the 16-bitness, UTF-16 is byte-order dependent. UTF-16 |
| 1256 | itself can be used for in-memory computations, but if storage or |
| 1257 | transfer is required either UTF-16BE (big-endian) or UTF-16LE |
| 1258 | (little-endian) encodings must be chosen. |
| 1259 | |
| 1260 | This introduces another problem: what if you just know that your data |
| 1261 | is UTF-16, but you don't know which endianness? Byte Order Marks, or |
| 1262 | BOMs, are a solution to this. A special character has been reserved |
| 1263 | in Unicode to function as a byte order marker: the character with the |
| 1264 | code point C<U+FEFF> is the BOM. |
| 1265 | |
| 1266 | The trick is that if you read a BOM, you will know the byte order, |
| 1267 | since if it was written on a big-endian platform, you will read the |
| 1268 | bytes C<0xFE 0xFF>, but if it was written on a little-endian platform, |
| 1269 | you will read the bytes C<0xFF 0xFE>. (And if the originating platform |
| 1270 | was writing in UTF-8, you will read the bytes C<0xEF 0xBB 0xBF>.) |
| 1271 | |
| 1272 | The way this trick works is that the character with the code point |
| 1273 | C<U+FFFE> is guaranteed not to be a valid Unicode character, so the |
| 1274 | sequence of bytes C<0xFF 0xFE> is unambiguously "BOM, represented in |
| 1275 | little-endian format" and cannot be C<U+FFFE>, represented in big-endian |
| 1276 | format". (Actually, C<U+FFFE> is legal for use by your program, even for |
| 1277 | input/output, but better not use it if you need a BOM. But it is "illegal for |
| 1278 | interchange", so that an unsuspecting program won't get confused.) |
| 1279 | |
| 1280 | =item * |
| 1281 | |
| 1282 | UTF-32, UTF-32BE, UTF-32LE |
| 1283 | |
| 1284 | The UTF-32 family is pretty much like the UTF-16 family, expect that |
| 1285 | the units are 32-bit, and therefore the surrogate scheme is not |
| 1286 | needed. The BOM signatures will be C<0x00 0x00 0xFE 0xFF> for BE and |
| 1287 | C<0xFF 0xFE 0x00 0x00> for LE. |
| 1288 | |
| 1289 | =item * |
| 1290 | |
| 1291 | UCS-2, UCS-4 |
| 1292 | |
| 1293 | Encodings defined by the ISO 10646 standard. UCS-2 is a 16-bit |
| 1294 | encoding. Unlike UTF-16, UCS-2 is not extensible beyond C<U+FFFF>, |
| 1295 | because it does not use surrogates. UCS-4 is a 32-bit encoding, |
| 1296 | functionally identical to UTF-32. |
| 1297 | |
| 1298 | =item * |
| 1299 | |
| 1300 | UTF-7 |
| 1301 | |
| 1302 | A seven-bit safe (non-eight-bit) encoding, which is useful if the |
| 1303 | transport or storage is not eight-bit safe. Defined by RFC 2152. |
| 1304 | |
| 1305 | =back |
| 1306 | |
| 1307 | =head2 Security Implications of Unicode |
| 1308 | |
| 1309 | Read L<Unicode Security Considerations|http://www.unicode.org/reports/tr36>. |
| 1310 | Also, note the following: |
| 1311 | |
| 1312 | =over 4 |
| 1313 | |
| 1314 | =item * |
| 1315 | |
| 1316 | Malformed UTF-8 |
| 1317 | |
| 1318 | Unfortunately, the specification of UTF-8 leaves some room for |
| 1319 | interpretation of how many bytes of encoded output one should generate |
| 1320 | from one input Unicode character. Strictly speaking, the shortest |
| 1321 | possible sequence of UTF-8 bytes should be generated, |
| 1322 | because otherwise there is potential for an input buffer overflow at |
| 1323 | the receiving end of a UTF-8 connection. Perl always generates the |
| 1324 | shortest length UTF-8, and with warnings on, Perl will warn about |
| 1325 | non-shortest length UTF-8 along with other malformations, such as the |
| 1326 | surrogates, which are not real Unicode code points. |
| 1327 | |
| 1328 | =item * |
| 1329 | |
| 1330 | Regular expressions behave slightly differently between byte data and |
| 1331 | character (Unicode) data. For example, the "word character" character |
| 1332 | class C<\w> will work differently depending on if data is eight-bit bytes |
| 1333 | or Unicode. |
| 1334 | |
| 1335 | In the first case, the set of C<\w> characters is either small--the |
| 1336 | default set of alphabetic characters, digits, and the "_"--or, if you |
| 1337 | are using a locale (see L<perllocale>), the C<\w> might contain a few |
| 1338 | more letters according to your language and country. |
| 1339 | |
| 1340 | In the second case, the C<\w> set of characters is much, much larger. |
| 1341 | Most importantly, even in the set of the first 256 characters, it will |
| 1342 | probably match different characters: unlike most locales, which are |
| 1343 | specific to a language and country pair, Unicode classifies all the |
| 1344 | characters that are letters I<somewhere> as C<\w>. For example, your |
| 1345 | locale might not think that LATIN SMALL LETTER ETH is a letter (unless |
| 1346 | you happen to speak Icelandic), but Unicode does. |
| 1347 | |
| 1348 | As discussed elsewhere, Perl has one foot (two hooves?) planted in |
| 1349 | each of two worlds: the old world of bytes and the new world of |
| 1350 | characters, upgrading from bytes to characters when necessary. |
| 1351 | If your legacy code does not explicitly use Unicode, no automatic |
| 1352 | switch-over to characters should happen. Characters shouldn't get |
| 1353 | downgraded to bytes, either. It is possible to accidentally mix bytes |
| 1354 | and characters, however (see L<perluniintro>), in which case C<\w> in |
| 1355 | regular expressions might start behaving differently. Review your |
| 1356 | code. Use warnings and the C<strict> pragma. |
| 1357 | |
| 1358 | =back |
| 1359 | |
| 1360 | =head2 Unicode in Perl on EBCDIC |
| 1361 | |
| 1362 | The way Unicode is handled on EBCDIC platforms is still |
| 1363 | experimental. On such platforms, references to UTF-8 encoding in this |
| 1364 | document and elsewhere should be read as meaning the UTF-EBCDIC |
| 1365 | specified in Unicode Technical Report 16, unless ASCII vs. EBCDIC issues |
| 1366 | are specifically discussed. There is no C<utfebcdic> pragma or |
| 1367 | ":utfebcdic" layer; rather, "utf8" and ":utf8" are reused to mean |
| 1368 | the platform's "natural" 8-bit encoding of Unicode. See L<perlebcdic> |
| 1369 | for more discussion of the issues. |
| 1370 | |
| 1371 | =head2 Locales |
| 1372 | |
| 1373 | Usually locale settings and Unicode do not affect each other, but |
| 1374 | there are exceptions: |
| 1375 | |
| 1376 | =over 4 |
| 1377 | |
| 1378 | =item * |
| 1379 | |
| 1380 | You can enable automatic UTF-8-ification of your standard file |
| 1381 | handles, default C<open()> layer, and C<@ARGV> by using either |
| 1382 | the C<-C> command line switch or the C<PERL_UNICODE> environment |
| 1383 | variable, see L<perlrun> for the documentation of the C<-C> switch. |
| 1384 | |
| 1385 | =item * |
| 1386 | |
| 1387 | Perl tries really hard to work both with Unicode and the old |
| 1388 | byte-oriented world. Most often this is nice, but sometimes Perl's |
| 1389 | straddling of the proverbial fence causes problems. Here's an example |
| 1390 | of how things can go wrong. A locale can define a code point to be |
| 1391 | anything it wants. It could make 'A' into a control character, for example. |
| 1392 | But strings encoded in utf8 always have Unicode semantics, so an 'A' in |
| 1393 | such a string is always an uppercase letter, never a control, no matter |
| 1394 | what the locale says it should be. |
| 1395 | |
| 1396 | =back |
| 1397 | |
| 1398 | =head2 When Unicode Does Not Happen |
| 1399 | |
| 1400 | While Perl does have extensive ways to input and output in Unicode, |
| 1401 | and few other 'entry points' like the @ARGV which can be interpreted |
| 1402 | as Unicode (UTF-8), there still are many places where Unicode (in some |
| 1403 | encoding or another) could be given as arguments or received as |
| 1404 | results, or both, but it is not. |
| 1405 | |
| 1406 | The following are such interfaces. Also, see L</The "Unicode Bug">. |
| 1407 | For all of these interfaces Perl |
| 1408 | currently (as of 5.8.3) simply assumes byte strings both as arguments |
| 1409 | and results, or UTF-8 strings if the C<encoding> pragma has been used. |
| 1410 | |
| 1411 | One reason why Perl does not attempt to resolve the role of Unicode in |
| 1412 | these cases is that the answers are highly dependent on the operating |
| 1413 | system and the file system(s). For example, whether filenames can be |
| 1414 | in Unicode, and in exactly what kind of encoding, is not exactly a |
| 1415 | portable concept. Similarly for the qx and system: how well will the |
| 1416 | 'command line interface' (and which of them?) handle Unicode? |
| 1417 | |
| 1418 | =over 4 |
| 1419 | |
| 1420 | =item * |
| 1421 | |
| 1422 | chdir, chmod, chown, chroot, exec, link, lstat, mkdir, |
| 1423 | rename, rmdir, stat, symlink, truncate, unlink, utime, -X |
| 1424 | |
| 1425 | =item * |
| 1426 | |
| 1427 | %ENV |
| 1428 | |
| 1429 | =item * |
| 1430 | |
| 1431 | glob (aka the <*>) |
| 1432 | |
| 1433 | =item * |
| 1434 | |
| 1435 | open, opendir, sysopen |
| 1436 | |
| 1437 | =item * |
| 1438 | |
| 1439 | qx (aka the backtick operator), system |
| 1440 | |
| 1441 | =item * |
| 1442 | |
| 1443 | readdir, readlink |
| 1444 | |
| 1445 | =back |
| 1446 | |
| 1447 | =head2 The "Unicode Bug" |
| 1448 | |
| 1449 | The term, the "Unicode bug" has been applied to an inconsistency with the |
| 1450 | Unicode characters whose ordinals are in the Latin-1 Supplement block, that |
| 1451 | is, between 128 and 255. Without a locale specified, unlike all other |
| 1452 | characters or code points, these characters have very different semantics in |
| 1453 | byte semantics versus character semantics. |
| 1454 | |
| 1455 | In character semantics they are interpreted as Unicode code points, which means |
| 1456 | they have the same semantics as Latin-1 (ISO-8859-1). |
| 1457 | |
| 1458 | In byte semantics, they are considered to be unassigned characters, meaning |
| 1459 | that the only semantics they have is their ordinal numbers, and that they are |
| 1460 | not members of various character classes. None are considered to match C<\w> |
| 1461 | for example, but all match C<\W>. (On EBCDIC platforms, the behavior may |
| 1462 | be different from this, depending on the underlying C language library |
| 1463 | functions.) |
| 1464 | |
| 1465 | The behavior is known to have effects on these areas: |
| 1466 | |
| 1467 | =over 4 |
| 1468 | |
| 1469 | =item * |
| 1470 | |
| 1471 | Changing the case of a scalar, that is, using C<uc()>, C<ucfirst()>, C<lc()>, |
| 1472 | and C<lcfirst()>, or C<\L>, C<\U>, C<\u> and C<\l> in regular expression |
| 1473 | substitutions. |
| 1474 | |
| 1475 | =item * |
| 1476 | |
| 1477 | Using caseless (C</i>) regular expression matching |
| 1478 | |
| 1479 | =item * |
| 1480 | |
| 1481 | Matching a number of properties in regular expressions, namely C<\b>, |
| 1482 | C<\B>, C<\s>, C<\S>, C<\w>, C<\W>, and all the Posix character classes |
| 1483 | I<except> C<[[:ascii:]]>. |
| 1484 | |
| 1485 | =item * |
| 1486 | |
| 1487 | User-defined case change mappings. You can create a C<ToUpper()> function, for |
| 1488 | example, which overrides Perl's built-in case mappings. The scalar must be |
| 1489 | encoded in utf8 for your function to actually be invoked. |
| 1490 | |
| 1491 | =back |
| 1492 | |
| 1493 | This behavior can lead to unexpected results in which a string's semantics |
| 1494 | suddenly change if a code point above 255 is appended to or removed from it, |
| 1495 | which changes the string's semantics from byte to character or vice versa. As |
| 1496 | an example, consider the following program and its output: |
| 1497 | |
| 1498 | $ perl -le' |
| 1499 | $s1 = "\xC2"; |
| 1500 | $s2 = "\x{2660}"; |
| 1501 | for ($s1, $s2, $s1.$s2) { |
| 1502 | print /\w/ || 0; |
| 1503 | } |
| 1504 | ' |
| 1505 | 0 |
| 1506 | 0 |
| 1507 | 1 |
| 1508 | |
| 1509 | If there's no C<\w> in C<s1> or in C<s2>, why does their concatenation have one? |
| 1510 | |
| 1511 | This anomaly stems from Perl's attempt to not disturb older programs that |
| 1512 | didn't use Unicode, and hence had no semantics for characters outside of the |
| 1513 | ASCII range (except in a locale), along with Perl's desire to add Unicode |
| 1514 | support seamlessly. The result wasn't seamless: these characters were |
| 1515 | orphaned. |
| 1516 | |
| 1517 | Work is being done to correct this, but only some of it is complete. |
| 1518 | What has been finished is: |
| 1519 | |
| 1520 | =over |
| 1521 | |
| 1522 | =item * |
| 1523 | |
| 1524 | the matching of C<\b>, C<\s>, C<\w> and the Posix |
| 1525 | character classes and their complements in regular expressions |
| 1526 | |
| 1527 | =item * |
| 1528 | |
| 1529 | case changing (but not user-defined casing) |
| 1530 | |
| 1531 | =item * |
| 1532 | |
| 1533 | case-insensitive (C</i>) regular expression matching for [bracketed |
| 1534 | character classes] only, except for some bugs with C<LATIN SMALL |
| 1535 | LETTER SHARP S> (which is supposed to match the two character sequence |
| 1536 | "ss" (or "Ss" or "sS" or "SS"), but Perl has a number of bugs for all |
| 1537 | such multi-character case insensitive characters, of which this is just |
| 1538 | one example. |
| 1539 | |
| 1540 | =back |
| 1541 | |
| 1542 | Due to concerns, and some evidence, that older code might |
| 1543 | have come to rely on the existing behavior, the new behavior must be explicitly |
| 1544 | enabled by the feature C<unicode_strings> in the L<feature> pragma, even though |
| 1545 | no new syntax is involved. |
| 1546 | |
| 1547 | See L<perlfunc/lc> for details on how this pragma works in combination with |
| 1548 | various others for casing. |
| 1549 | |
| 1550 | Even though the implementation is incomplete, it is planned to have this |
| 1551 | pragma affect all the problematic behaviors in later releases: you can't |
| 1552 | have one without them all. |
| 1553 | |
| 1554 | In the meantime, a workaround is to always call utf8::upgrade($string), or to |
| 1555 | use the standard module L<Encode>. Also, a scalar that has any characters |
| 1556 | whose ordinal is above 0x100, or which were specified using either of the |
| 1557 | C<\N{...}> notations will automatically have character semantics. |
| 1558 | |
| 1559 | =head2 Forcing Unicode in Perl (Or Unforcing Unicode in Perl) |
| 1560 | |
| 1561 | Sometimes (see L</"When Unicode Does Not Happen"> or L</The "Unicode Bug">) |
| 1562 | there are situations where you simply need to force a byte |
| 1563 | string into UTF-8, or vice versa. The low-level calls |
| 1564 | utf8::upgrade($bytestring) and utf8::downgrade($utf8string[, FAIL_OK]) are |
| 1565 | the answers. |
| 1566 | |
| 1567 | Note that utf8::downgrade() can fail if the string contains characters |
| 1568 | that don't fit into a byte. |
| 1569 | |
| 1570 | Calling either function on a string that already is in the desired state is a |
| 1571 | no-op. |
| 1572 | |
| 1573 | =head2 Using Unicode in XS |
| 1574 | |
| 1575 | If you want to handle Perl Unicode in XS extensions, you may find the |
| 1576 | following C APIs useful. See also L<perlguts/"Unicode Support"> for an |
| 1577 | explanation about Unicode at the XS level, and L<perlapi> for the API |
| 1578 | details. |
| 1579 | |
| 1580 | =over 4 |
| 1581 | |
| 1582 | =item * |
| 1583 | |
| 1584 | C<DO_UTF8(sv)> returns true if the C<UTF8> flag is on and the bytes |
| 1585 | pragma is not in effect. C<SvUTF8(sv)> returns true if the C<UTF8> |
| 1586 | flag is on; the bytes pragma is ignored. The C<UTF8> flag being on |
| 1587 | does B<not> mean that there are any characters of code points greater |
| 1588 | than 255 (or 127) in the scalar or that there are even any characters |
| 1589 | in the scalar. What the C<UTF8> flag means is that the sequence of |
| 1590 | octets in the representation of the scalar is the sequence of UTF-8 |
| 1591 | encoded code points of the characters of a string. The C<UTF8> flag |
| 1592 | being off means that each octet in this representation encodes a |
| 1593 | single character with code point 0..255 within the string. Perl's |
| 1594 | Unicode model is not to use UTF-8 until it is absolutely necessary. |
| 1595 | |
| 1596 | =item * |
| 1597 | |
| 1598 | C<uvchr_to_utf8(buf, chr)> writes a Unicode character code point into |
| 1599 | a buffer encoding the code point as UTF-8, and returns a pointer |
| 1600 | pointing after the UTF-8 bytes. It works appropriately on EBCDIC machines. |
| 1601 | |
| 1602 | =item * |
| 1603 | |
| 1604 | C<utf8_to_uvchr(buf, lenp)> reads UTF-8 encoded bytes from a buffer and |
| 1605 | returns the Unicode character code point and, optionally, the length of |
| 1606 | the UTF-8 byte sequence. It works appropriately on EBCDIC machines. |
| 1607 | |
| 1608 | =item * |
| 1609 | |
| 1610 | C<utf8_length(start, end)> returns the length of the UTF-8 encoded buffer |
| 1611 | in characters. C<sv_len_utf8(sv)> returns the length of the UTF-8 encoded |
| 1612 | scalar. |
| 1613 | |
| 1614 | =item * |
| 1615 | |
| 1616 | C<sv_utf8_upgrade(sv)> converts the string of the scalar to its UTF-8 |
| 1617 | encoded form. C<sv_utf8_downgrade(sv)> does the opposite, if |
| 1618 | possible. C<sv_utf8_encode(sv)> is like sv_utf8_upgrade except that |
| 1619 | it does not set the C<UTF8> flag. C<sv_utf8_decode()> does the |
| 1620 | opposite of C<sv_utf8_encode()>. Note that none of these are to be |
| 1621 | used as general-purpose encoding or decoding interfaces: C<use Encode> |
| 1622 | for that. C<sv_utf8_upgrade()> is affected by the encoding pragma |
| 1623 | but C<sv_utf8_downgrade()> is not (since the encoding pragma is |
| 1624 | designed to be a one-way street). |
| 1625 | |
| 1626 | =item * |
| 1627 | |
| 1628 | C<is_utf8_char(s)> returns true if the pointer points to a valid UTF-8 |
| 1629 | character. |
| 1630 | |
| 1631 | =item * |
| 1632 | |
| 1633 | C<is_utf8_string(buf, len)> returns true if C<len> bytes of the buffer |
| 1634 | are valid UTF-8. |
| 1635 | |
| 1636 | =item * |
| 1637 | |
| 1638 | C<UTF8SKIP(buf)> will return the number of bytes in the UTF-8 encoded |
| 1639 | character in the buffer. C<UNISKIP(chr)> will return the number of bytes |
| 1640 | required to UTF-8-encode the Unicode character code point. C<UTF8SKIP()> |
| 1641 | is useful for example for iterating over the characters of a UTF-8 |
| 1642 | encoded buffer; C<UNISKIP()> is useful, for example, in computing |
| 1643 | the size required for a UTF-8 encoded buffer. |
| 1644 | |
| 1645 | =item * |
| 1646 | |
| 1647 | C<utf8_distance(a, b)> will tell the distance in characters between the |
| 1648 | two pointers pointing to the same UTF-8 encoded buffer. |
| 1649 | |
| 1650 | =item * |
| 1651 | |
| 1652 | C<utf8_hop(s, off)> will return a pointer to a UTF-8 encoded buffer |
| 1653 | that is C<off> (positive or negative) Unicode characters displaced |
| 1654 | from the UTF-8 buffer C<s>. Be careful not to overstep the buffer: |
| 1655 | C<utf8_hop()> will merrily run off the end or the beginning of the |
| 1656 | buffer if told to do so. |
| 1657 | |
| 1658 | =item * |
| 1659 | |
| 1660 | C<pv_uni_display(dsv, spv, len, pvlim, flags)> and |
| 1661 | C<sv_uni_display(dsv, ssv, pvlim, flags)> are useful for debugging the |
| 1662 | output of Unicode strings and scalars. By default they are useful |
| 1663 | only for debugging--they display B<all> characters as hexadecimal code |
| 1664 | points--but with the flags C<UNI_DISPLAY_ISPRINT>, |
| 1665 | C<UNI_DISPLAY_BACKSLASH>, and C<UNI_DISPLAY_QQ> you can make the |
| 1666 | output more readable. |
| 1667 | |
| 1668 | =item * |
| 1669 | |
| 1670 | C<foldEQ_utf8(s1, pe1, l1, u1, s2, pe2, l2, u2)> can be used to |
| 1671 | compare two strings case-insensitively in Unicode. For case-sensitive |
| 1672 | comparisons you can just use C<memEQ()> and C<memNE()> as usual, except |
| 1673 | if one string is in utf8 and the other isn't. |
| 1674 | |
| 1675 | =back |
| 1676 | |
| 1677 | For more information, see L<perlapi>, and F<utf8.c> and F<utf8.h> |
| 1678 | in the Perl source code distribution. |
| 1679 | |
| 1680 | =head2 Hacking Perl to work on earlier Unicode versions (for very serious hackers only) |
| 1681 | |
| 1682 | Perl by default comes with the latest supported Unicode version built in, but |
| 1683 | you can change to use any earlier one. |
| 1684 | |
| 1685 | Download the files in the version of Unicode that you want from the Unicode web |
| 1686 | site L<http://www.unicode.org>). These should replace the existing files in |
| 1687 | C<\$Config{privlib}>/F<unicore>. (C<\%Config> is available from the Config |
| 1688 | module.) Follow the instructions in F<README.perl> in that directory to change |
| 1689 | some of their names, and then run F<make>. |
| 1690 | |
| 1691 | It is even possible to download them to a different directory, and then change |
| 1692 | F<utf8_heavy.pl> in the directory C<\$Config{privlib}> to point to the new |
| 1693 | directory, or maybe make a copy of that directory before making the change, and |
| 1694 | using C<@INC> or the C<-I> run-time flag to switch between versions at will |
| 1695 | (but because of caching, not in the middle of a process), but all this is |
| 1696 | beyond the scope of these instructions. |
| 1697 | |
| 1698 | =head1 BUGS |
| 1699 | |
| 1700 | =head2 Interaction with Locales |
| 1701 | |
| 1702 | Use of locales with Unicode data may lead to odd results. Currently, |
| 1703 | Perl attempts to attach 8-bit locale info to characters in the range |
| 1704 | 0..255, but this technique is demonstrably incorrect for locales that |
| 1705 | use characters above that range when mapped into Unicode. Perl's |
| 1706 | Unicode support will also tend to run slower. Use of locales with |
| 1707 | Unicode is discouraged. |
| 1708 | |
| 1709 | =head2 Problems with characters in the Latin-1 Supplement range |
| 1710 | |
| 1711 | See L</The "Unicode Bug"> |
| 1712 | |
| 1713 | =head2 Problems with case-insensitive regular expression matching |
| 1714 | |
| 1715 | There are problems with case-insensitive matches, including those involving |
| 1716 | character classes (enclosed in [square brackets]), characters whose fold |
| 1717 | is to multiple characters (such as the single character LATIN SMALL LIGATURE |
| 1718 | FFL matches case-insensitively with the 3-character string C<ffl>), and |
| 1719 | characters in the Latin-1 Supplement. |
| 1720 | |
| 1721 | =head2 Interaction with Extensions |
| 1722 | |
| 1723 | When Perl exchanges data with an extension, the extension should be |
| 1724 | able to understand the UTF8 flag and act accordingly. If the |
| 1725 | extension doesn't know about the flag, it's likely that the extension |
| 1726 | will return incorrectly-flagged data. |
| 1727 | |
| 1728 | So if you're working with Unicode data, consult the documentation of |
| 1729 | every module you're using if there are any issues with Unicode data |
| 1730 | exchange. If the documentation does not talk about Unicode at all, |
| 1731 | suspect the worst and probably look at the source to learn how the |
| 1732 | module is implemented. Modules written completely in Perl shouldn't |
| 1733 | cause problems. Modules that directly or indirectly access code written |
| 1734 | in other programming languages are at risk. |
| 1735 | |
| 1736 | For affected functions, the simple strategy to avoid data corruption is |
| 1737 | to always make the encoding of the exchanged data explicit. Choose an |
| 1738 | encoding that you know the extension can handle. Convert arguments passed |
| 1739 | to the extensions to that encoding and convert results back from that |
| 1740 | encoding. Write wrapper functions that do the conversions for you, so |
| 1741 | you can later change the functions when the extension catches up. |
| 1742 | |
| 1743 | To provide an example, let's say the popular Foo::Bar::escape_html |
| 1744 | function doesn't deal with Unicode data yet. The wrapper function |
| 1745 | would convert the argument to raw UTF-8 and convert the result back to |
| 1746 | Perl's internal representation like so: |
| 1747 | |
| 1748 | sub my_escape_html ($) { |
| 1749 | my($what) = shift; |
| 1750 | return unless defined $what; |
| 1751 | Encode::decode_utf8(Foo::Bar::escape_html( |
| 1752 | Encode::encode_utf8($what))); |
| 1753 | } |
| 1754 | |
| 1755 | Sometimes, when the extension does not convert data but just stores |
| 1756 | and retrieves them, you will be in a position to use the otherwise |
| 1757 | dangerous Encode::_utf8_on() function. Let's say the popular |
| 1758 | C<Foo::Bar> extension, written in C, provides a C<param> method that |
| 1759 | lets you store and retrieve data according to these prototypes: |
| 1760 | |
| 1761 | $self->param($name, $value); # set a scalar |
| 1762 | $value = $self->param($name); # retrieve a scalar |
| 1763 | |
| 1764 | If it does not yet provide support for any encoding, one could write a |
| 1765 | derived class with such a C<param> method: |
| 1766 | |
| 1767 | sub param { |
| 1768 | my($self,$name,$value) = @_; |
| 1769 | utf8::upgrade($name); # make sure it is UTF-8 encoded |
| 1770 | if (defined $value) { |
| 1771 | utf8::upgrade($value); # make sure it is UTF-8 encoded |
| 1772 | return $self->SUPER::param($name,$value); |
| 1773 | } else { |
| 1774 | my $ret = $self->SUPER::param($name); |
| 1775 | Encode::_utf8_on($ret); # we know, it is UTF-8 encoded |
| 1776 | return $ret; |
| 1777 | } |
| 1778 | } |
| 1779 | |
| 1780 | Some extensions provide filters on data entry/exit points, such as |
| 1781 | DB_File::filter_store_key and family. Look out for such filters in |
| 1782 | the documentation of your extensions, they can make the transition to |
| 1783 | Unicode data much easier. |
| 1784 | |
| 1785 | =head2 Speed |
| 1786 | |
| 1787 | Some functions are slower when working on UTF-8 encoded strings than |
| 1788 | on byte encoded strings. All functions that need to hop over |
| 1789 | characters such as length(), substr() or index(), or matching regular |
| 1790 | expressions can work B<much> faster when the underlying data are |
| 1791 | byte-encoded. |
| 1792 | |
| 1793 | In Perl 5.8.0 the slowness was often quite spectacular; in Perl 5.8.1 |
| 1794 | a caching scheme was introduced which will hopefully make the slowness |
| 1795 | somewhat less spectacular, at least for some operations. In general, |
| 1796 | operations with UTF-8 encoded strings are still slower. As an example, |
| 1797 | the Unicode properties (character classes) like C<\p{Nd}> are known to |
| 1798 | be quite a bit slower (5-20 times) than their simpler counterparts |
| 1799 | like C<\d> (then again, there 268 Unicode characters matching C<Nd> |
| 1800 | compared with the 10 ASCII characters matching C<d>). |
| 1801 | |
| 1802 | =head2 Problems on EBCDIC platforms |
| 1803 | |
| 1804 | There are a number of known problems with Perl on EBCDIC platforms. If you |
| 1805 | want to use Perl there, send email to perlbug@perl.org. |
| 1806 | |
| 1807 | In earlier versions, when byte and character data were concatenated, |
| 1808 | the new string was sometimes created by |
| 1809 | decoding the byte strings as I<ISO 8859-1 (Latin-1)>, even if the |
| 1810 | old Unicode string used EBCDIC. |
| 1811 | |
| 1812 | If you find any of these, please report them as bugs. |
| 1813 | |
| 1814 | =head2 Porting code from perl-5.6.X |
| 1815 | |
| 1816 | Perl 5.8 has a different Unicode model from 5.6. In 5.6 the programmer |
| 1817 | was required to use the C<utf8> pragma to declare that a given scope |
| 1818 | expected to deal with Unicode data and had to make sure that only |
| 1819 | Unicode data were reaching that scope. If you have code that is |
| 1820 | working with 5.6, you will need some of the following adjustments to |
| 1821 | your code. The examples are written such that the code will continue |
| 1822 | to work under 5.6, so you should be safe to try them out. |
| 1823 | |
| 1824 | =over 4 |
| 1825 | |
| 1826 | =item * |
| 1827 | |
| 1828 | A filehandle that should read or write UTF-8 |
| 1829 | |
| 1830 | if ($] > 5.007) { |
| 1831 | binmode $fh, ":encoding(utf8)"; |
| 1832 | } |
| 1833 | |
| 1834 | =item * |
| 1835 | |
| 1836 | A scalar that is going to be passed to some extension |
| 1837 | |
| 1838 | Be it Compress::Zlib, Apache::Request or any extension that has no |
| 1839 | mention of Unicode in the manpage, you need to make sure that the |
| 1840 | UTF8 flag is stripped off. Note that at the time of this writing |
| 1841 | (October 2002) the mentioned modules are not UTF-8-aware. Please |
| 1842 | check the documentation to verify if this is still true. |
| 1843 | |
| 1844 | if ($] > 5.007) { |
| 1845 | require Encode; |
| 1846 | $val = Encode::encode_utf8($val); # make octets |
| 1847 | } |
| 1848 | |
| 1849 | =item * |
| 1850 | |
| 1851 | A scalar we got back from an extension |
| 1852 | |
| 1853 | If you believe the scalar comes back as UTF-8, you will most likely |
| 1854 | want the UTF8 flag restored: |
| 1855 | |
| 1856 | if ($] > 5.007) { |
| 1857 | require Encode; |
| 1858 | $val = Encode::decode_utf8($val); |
| 1859 | } |
| 1860 | |
| 1861 | =item * |
| 1862 | |
| 1863 | Same thing, if you are really sure it is UTF-8 |
| 1864 | |
| 1865 | if ($] > 5.007) { |
| 1866 | require Encode; |
| 1867 | Encode::_utf8_on($val); |
| 1868 | } |
| 1869 | |
| 1870 | =item * |
| 1871 | |
| 1872 | A wrapper for fetchrow_array and fetchrow_hashref |
| 1873 | |
| 1874 | When the database contains only UTF-8, a wrapper function or method is |
| 1875 | a convenient way to replace all your fetchrow_array and |
| 1876 | fetchrow_hashref calls. A wrapper function will also make it easier to |
| 1877 | adapt to future enhancements in your database driver. Note that at the |
| 1878 | time of this writing (October 2002), the DBI has no standardized way |
| 1879 | to deal with UTF-8 data. Please check the documentation to verify if |
| 1880 | that is still true. |
| 1881 | |
| 1882 | sub fetchrow { |
| 1883 | # $what is one of fetchrow_{array,hashref} |
| 1884 | my($self, $sth, $what) = @_; |
| 1885 | if ($] < 5.007) { |
| 1886 | return $sth->$what; |
| 1887 | } else { |
| 1888 | require Encode; |
| 1889 | if (wantarray) { |
| 1890 | my @arr = $sth->$what; |
| 1891 | for (@arr) { |
| 1892 | defined && /[^\000-\177]/ && Encode::_utf8_on($_); |
| 1893 | } |
| 1894 | return @arr; |
| 1895 | } else { |
| 1896 | my $ret = $sth->$what; |
| 1897 | if (ref $ret) { |
| 1898 | for my $k (keys %$ret) { |
| 1899 | defined |
| 1900 | && /[^\000-\177]/ |
| 1901 | && Encode::_utf8_on($_) for $ret->{$k}; |
| 1902 | } |
| 1903 | return $ret; |
| 1904 | } else { |
| 1905 | defined && /[^\000-\177]/ && Encode::_utf8_on($_) for $ret; |
| 1906 | return $ret; |
| 1907 | } |
| 1908 | } |
| 1909 | } |
| 1910 | } |
| 1911 | |
| 1912 | |
| 1913 | =item * |
| 1914 | |
| 1915 | A large scalar that you know can only contain ASCII |
| 1916 | |
| 1917 | Scalars that contain only ASCII and are marked as UTF-8 are sometimes |
| 1918 | a drag to your program. If you recognize such a situation, just remove |
| 1919 | the UTF8 flag: |
| 1920 | |
| 1921 | utf8::downgrade($val) if $] > 5.007; |
| 1922 | |
| 1923 | =back |
| 1924 | |
| 1925 | =head1 SEE ALSO |
| 1926 | |
| 1927 | L<perlunitut>, L<perluniintro>, L<perluniprops>, L<Encode>, L<open>, L<utf8>, L<bytes>, |
| 1928 | L<perlretut>, L<perlvar/"${^UNICODE}"> |
| 1929 | L<http://www.unicode.org/reports/tr44>). |
| 1930 | |
| 1931 | =cut |