{ "content": [ { "type": "text", "text": "# perluniintro (man)\n\n## NAME\n\nperluniintro - Perl Unicode introduction\n\n## DESCRIPTION\n\nThis document gives a general idea of Unicode and how to use Unicode in Perl. See \"Further\nResources\" for references to more in-depth treatments of Unicode.\n\n## Sections\n\n- **NAME**\n- **DESCRIPTION** (15 subsections)\n- **UNICODE IN OLDER PERLS**\n- **SEE ALSO**\n- **ACKNOWLEDGMENTS** (1 subsections)\n\nUse structuredContent.sections for detailed options, examples, and full documentation.\n" } ], "structuredContent": { "command": "perluniintro", "section": "", "mode": "man", "summary": "perluniintro - Perl Unicode introduction", "synopsis": null, "tldr_summary": null, "tldr_examples": [], "tldr_source": null, "flags": [], "examples": [], "see_also": [], "section_outline": [ { "name": "NAME", "lines": 2, "subsections": [] }, { "name": "DESCRIPTION", "lines": 3, "subsections": [ { "name": "Unicode", "lines": 106 }, { "name": "Perl's Unicode Support", "lines": 21 }, { "name": "Perl's Unicode Model", "lines": 47 }, { "name": "Unicode and EBCDIC", "lines": 11 }, { "name": "Creating Unicode", "lines": 74 }, { "name": "Handling Unicode", "lines": 16 }, { "name": "Legacy Encodings", "lines": 9 }, { "name": "Unicode I/O", "lines": 123 }, { "name": "Displaying Unicode As Text", "lines": 28 }, { "name": "Special Cases", "lines": 31 }, { "name": "Advanced Topics", "lines": 37 }, { "name": "Miscellaneous", "lines": 29 }, { "name": "Questions With Answers", "lines": 162 }, { "name": "Hexadecimal Notation", "lines": 25 }, { "name": "Further Resources", "lines": 36 } ] }, { "name": "UNICODE IN OLDER PERLS", "lines": 14, "subsections": [] }, { "name": "SEE ALSO", "lines": 3, "subsections": [] }, { "name": "ACKNOWLEDGMENTS", "lines": 3, "subsections": [ { "name": "AUTHOR, COPYRIGHT, AND LICENSE", "lines": 7 } ] } ], "sections": { "NAME": { "content": "perluniintro - Perl Unicode introduction\n", "subsections": [] }, "DESCRIPTION": { "content": "This document gives a general idea of Unicode and how to use Unicode in Perl. See \"Further\nResources\" for references to more in-depth treatments of Unicode.\n", "subsections": [ { "name": "Unicode", "content": "Unicode is a character set standard which plans to codify all of the writing systems of the\nworld, plus many other symbols.\n\nUnicode and ISO/IEC 10646 are coordinated standards that unify almost all other modern\ncharacter set standards, covering more than 80 writing systems and hundreds of languages,\nincluding all commercially-important modern languages. All characters in the largest\nChinese, Japanese, and Korean dictionaries are also encoded. The standards will eventually\ncover almost all characters in more than 250 writing systems and thousands of languages.\nUnicode 1.0 was released in October 1991, and 6.0 in October 2010.\n\nA Unicode character is an abstract entity. It is not bound to any particular integer width,\nespecially not to the C language \"char\". Unicode is language-neutral and display-neutral: it\ndoes not encode the language of the text, and it does not generally define fonts or other\ngraphical layout details. Unicode operates on characters and on text built from those\ncharacters.\n\nUnicode defines characters like \"LATIN CAPITAL LETTER A\" or \"GREEK SMALL LETTER ALPHA\" and\nunique numbers for the characters, in this case 0x0041 and 0x03B1, respectively. These\nunique numbers are called code points. A code point is essentially the position of the\ncharacter within the set of all possible Unicode characters, and thus in Perl, the term\nordinal is often used interchangeably with it.\n\nThe Unicode standard prefers using hexadecimal notation for the code points. If numbers like\n0x0041 are unfamiliar to you, take a peek at a later section, \"Hexadecimal Notation\". The\nUnicode standard uses the notation \"U+0041 LATIN CAPITAL LETTER A\", to give the hexadecimal\ncode point and the normative name of the character.\n\nUnicode also defines various properties for the characters, like \"uppercase\" or \"lowercase\",\n\"decimal digit\", or \"punctuation\"; these properties are independent of the names of the\ncharacters. Furthermore, various operations on the characters like uppercasing, lowercasing,\nand collating (sorting) are defined.\n\nA Unicode logical \"character\" can actually consist of more than one internal actual\n\"character\" or code point. For Western languages, this is adequately modelled by a base\ncharacter (like \"LATIN CAPITAL LETTER A\") followed by one or more modifiers (like \"COMBINING\nACUTE ACCENT\"). This sequence of base character and modifiers is called a combining\ncharacter sequence. Some non-western languages require more complicated models, so Unicode\ncreated the grapheme cluster concept, which was later further refined into the extended\ngrapheme cluster. For example, a Korean Hangul syllable is considered a single logical\ncharacter, but most often consists of three actual Unicode characters: a leading consonant\nfollowed by an interior vowel followed by a trailing consonant.\n\nWhether to call these extended grapheme clusters \"characters\" depends on your point of view.\nIf you are a programmer, you probably would tend towards seeing each element in the sequences\nas one unit, or \"character\". However from the user's point of view, the whole sequence could\nbe seen as one \"character\" since that's probably what it looks like in the context of the\nuser's language. In this document, we take the programmer's point of view: one \"character\"\nis one Unicode code point.\n\nFor some combinations of base character and modifiers, there are precomposed characters.\nThere is a single character equivalent, for example, for the sequence \"LATIN CAPITAL LETTER\nA\" followed by \"COMBINING ACUTE ACCENT\". It is called \"LATIN CAPITAL LETTER A WITH ACUTE\".\nThese precomposed characters are, however, only available for some combinations, and are\nmainly meant to support round-trip conversions between Unicode and legacy standards (like ISO\n8859). Using sequences, as Unicode does, allows for needing fewer basic building blocks\n(code points) to express many more potential grapheme clusters. To support conversion\nbetween equivalent forms, various normalization forms are also defined. Thus, \"LATIN CAPITAL\nLETTER A WITH ACUTE\" is in Normalization Form Composed, (abbreviated NFC), and the sequence\n\"LATIN CAPITAL LETTER A\" followed by \"COMBINING ACUTE ACCENT\" represents the same character\nin Normalization Form Decomposed (NFD).\n\nBecause of backward compatibility with legacy encodings, the \"a unique number for every\ncharacter\" idea breaks down a bit: instead, there is \"at least one number for every\ncharacter\". The same character could be represented differently in several legacy encodings.\nThe converse is not true: some code points do not have an assigned character. Firstly, there\nare unallocated code points within otherwise used blocks. Secondly, there are special\nUnicode control characters that do not represent true characters.\n\nWhen Unicode was first conceived, it was thought that all the world's characters could be\nrepresented using a 16-bit word; that is a maximum of 0x10000 (or 65,536) characters would be\nneeded, from 0x0000 to 0xFFFF. This soon proved to be wrong, and since Unicode 2.0 (July\n1996), Unicode has been defined all the way up to 21 bits (0x10FFFF), and Unicode 3.1 (March\n2001) defined the first characters above 0xFFFF. The first 0x10000 characters are called the\nPlane 0, or the Basic Multilingual Plane (BMP). With Unicode 3.1, 17 (yes, seventeen) planes\nin all were defined--but they are nowhere near full of defined characters, yet.\n\nWhen a new language is being encoded, Unicode generally will choose a \"block\" of consecutive\nunallocated code points for its characters. So far, the number of code points in these\nblocks has always been evenly divisible by 16. Extras in a block, not currently needed, are\nleft unallocated, for future growth. But there have been occasions when a later release\nneeded more code points than the available extras, and a new block had to allocated somewhere\nelse, not contiguous to the initial one, to handle the overflow. Thus, it became apparent\nearly on that \"block\" wasn't an adequate organizing principle, and so the \"Script\" property\nwas created. (Later an improved script property was added as well, the \"ScriptExtensions\"\nproperty.) Those code points that are in overflow blocks can still have the same script as\nthe original ones. The script concept fits more closely with natural language: there is\n\"Latin\" script, \"Greek\" script, and so on; and there are several artificial scripts, like\n\"Common\" for characters that are used in multiple scripts, such as mathematical symbols.\nScripts usually span varied parts of several blocks. For more information about scripts, see\n\"Scripts\" in perlunicode. The division into blocks exists, but it is almost completely\naccidental--an artifact of how the characters have been and still are allocated. (Note that\nthis paragraph has oversimplified things for the sake of this being an introduction. Unicode\ndoesn't really encode languages, but the writing systems for them--their scripts; and one\nscript can be used by many languages. Unicode also encodes things that aren't really about\nlanguages, such as symbols like \"BAGGAGE CLAIM\".)\n\nThe Unicode code points are just abstract numbers. To input and output these abstract\nnumbers, the numbers must be encoded or serialised somehow. Unicode defines several\ncharacter encoding forms, of which UTF-8 is the most popular. UTF-8 is a variable length\nencoding that encodes Unicode characters as 1 to 4 bytes. Other encodings include UTF-16 and\nUTF-32 and their big- and little-endian variants (UTF-8 is byte-order independent). The\nISO/IEC 10646 defines the UCS-2 and UCS-4 encoding forms.\n\nFor more information about encodings--for instance, to learn what surrogates and byte order\nmarks (BOMs) are--see perlunicode.\n" }, { "name": "Perl's Unicode Support", "content": "Starting from Perl v5.6.0, Perl has had the capacity to handle Unicode natively. Perl\nv5.8.0, however, is the first recommended release for serious Unicode work. The maintenance\nrelease 5.6.1 fixed many of the problems of the initial Unicode implementation, but for\nexample regular expressions still do not work with Unicode in 5.6.1. Perl v5.14.0 is the\nfirst release where Unicode support is (almost) seamlessly integratable without some gotchas.\n(There are a few exceptions. Firstly, some differences in quotemeta were fixed starting in\nPerl 5.16.0. Secondly, some differences in the range operator were fixed starting in Perl\n5.26.0. Thirdly, some differences in split were fixed started in Perl 5.28.0.)\n\nTo enable this seamless support, you should \"use feature 'unicodestrings'\" (which is\nautomatically selected if you \"use 5.012\" or higher). See feature. (5.14 also fixes a\nnumber of bugs and departures from the Unicode standard.)\n\nBefore Perl v5.8.0, the use of \"use utf8\" was used to declare that operations in the current\nblock or file would be Unicode-aware. This model was found to be wrong, or at least clumsy:\nthe \"Unicodeness\" is now carried with the data, instead of being attached to the operations.\nStarting with Perl v5.8.0, only one case remains where an explicit \"use utf8\" is needed: if\nyour Perl script itself is encoded in UTF-8, you can use UTF-8 in your identifier names, and\nin string and regular expression literals, by saying \"use utf8\". This is not the default\nbecause scripts with legacy 8-bit data in them would break. See utf8.\n" }, { "name": "Perl's Unicode Model", "content": "Perl supports both pre-5.6 strings of eight-bit native bytes, and strings of Unicode\ncharacters. The general principle is that Perl tries to keep its data as eight-bit bytes for\nas long as possible, but as soon as Unicodeness cannot be avoided, the data is transparently\nupgraded to Unicode. Prior to Perl v5.14.0, the upgrade was not completely transparent (see\n\"The \"Unicode Bug\"\" in perlunicode), and for backwards compatibility, full transparency is\nnot gained unless \"use feature 'unicodestrings'\" (see feature) or \"use 5.012\" (or higher) is\nselected.\n\nInternally, Perl currently uses either whatever the native eight-bit character set of the\nplatform (for example Latin-1) is, defaulting to UTF-8, to encode Unicode strings.\nSpecifically, if all code points in the string are 0xFF or less, Perl uses the native eight-\nbit character set. Otherwise, it uses UTF-8.\n\nA user of Perl does not normally need to know nor care how Perl happens to encode its\ninternal strings, but it becomes relevant when outputting Unicode strings to a stream without\na PerlIO layer (one with the \"default\" encoding). In such a case, the raw bytes used\ninternally (the native character set or UTF-8, as appropriate for each string) will be used,\nand a \"Wide character\" warning will be issued if those strings contain a character beyond\n0x00FF.\n\nFor example,\n\nperl -e 'print \"\\x{DF}\\n\", \"\\x{0100}\\x{DF}\\n\"'\n\nproduces a fairly useless mixture of native bytes and UTF-8, as well as a warning:\n\nWide character in print at ...\n\nTo output UTF-8, use the \":encoding\" or \":utf8\" output layer. Prepending\n\nbinmode(STDOUT, \":utf8\");\n\nto this sample program ensures that the output is completely UTF-8, and removes the program's\nwarning.\n\nYou can enable automatic UTF-8-ification of your standard file handles, default \"open()\"\nlayer, and @ARGV by using either the \"-C\" command line switch or the \"PERLUNICODE\"\nenvironment variable, see perlrun for the documentation of the \"-C\" switch.\n\nNote that this means that Perl expects other software to work the same way: if Perl has been\nled to believe that STDIN should be UTF-8, but then STDIN coming in from another command is\nnot UTF-8, Perl will likely complain about the malformed UTF-8.\n\nAll features that combine Unicode and I/O also require using the new PerlIO feature. Almost\nall Perl 5.8 platforms do use PerlIO, though: you can see whether yours is by running \"perl\n-V\" and looking for \"useperlio=define\".\n" }, { "name": "Unicode and EBCDIC", "content": "Perl 5.8.0 added support for Unicode on EBCDIC platforms. This support was allowed to lapse\nin later releases, but was revived in 5.22. Unicode support is somewhat more complex to\nimplement since additional conversions are needed. See perlebcdic for more information.\n\nOn EBCDIC platforms, the internal Unicode encoding form is UTF-EBCDIC instead of UTF-8. The\ndifference is that as UTF-8 is \"ASCII-safe\" in that ASCII characters encode to UTF-8 as-is,\nwhile UTF-EBCDIC is \"EBCDIC-safe\", in that all the basic characters (which includes all those\nthat have ASCII equivalents (like \"A\", \"0\", \"%\", etc.) are the same in both EBCDIC and UTF-\nEBCDIC. Often, documentation will use the term \"UTF-8\" to mean UTF-EBCDIC as well. This is\nthe case in this document.\n" }, { "name": "Creating Unicode", "content": "This section applies fully to Perls starting with v5.22. Various caveats for earlier\nreleases are in the \"Earlier releases caveats\" subsection below.\n\nTo create Unicode characters in literals, use the \"\\N{...}\" notation in double-quoted\nstrings:\n\nmy $smileyfromname = \"\\N{WHITE SMILING FACE}\";\nmy $smileyfromcodepoint = \"\\N{U+263a}\";\n\nSimilarly, they can be used in regular expression literals\n\n$smiley =~ /\\N{WHITE SMILING FACE}/;\n$smiley =~ /\\N{U+263a}/;\n\nor, starting in v5.32:\n\n$smiley =~ /\\p{Name=WHITE SMILING FACE}/;\n$smiley =~ /\\p{Name=whitesmilingface}/;\n\nAt run-time you can use:\n\nuse charnames ();\nmy $hebrewaleffromname\n= charnames::stringvianame(\"HEBREW LETTER ALEF\");\nmy $hebrewaleffromcodepoint = charnames::stringvianame(\"U+05D0\");\n\nNaturally, \"ord()\" will do the reverse: it turns a character into a code point.\n\nThere are other runtime options as well. You can use \"pack()\":\n\nmy $hebrewaleffromcodepoint = pack(\"U\", 0x05d0);\n\nOr you can use \"chr()\", though it is less convenient in the general case:\n\n$hebrewaleffromcodepoint = chr(utf8::unicodetonative(0x05d0));\nutf8::upgrade($hebrewaleffromcodepoint);\n\nThe \"utf8::unicodetonative()\" and \"utf8::upgrade()\" aren't needed if the argument is above\n0xFF, so the above could have been written as\n\n$hebrewaleffromcodepoint = chr(0x05d0);\n\nsince 0x5d0 is above 255.\n\n\"\\x{}\" and \"\\o{}\" can also be used to specify code points at compile time in double-quotish\nstrings, but, for backward compatibility with older Perls, the same rules apply as with\n\"chr()\" for code points less than 256.\n\n\"utf8::unicodetonative()\" is used so that the Perl code is portable to EBCDIC platforms.\nYou can omit it if you're really sure no one will ever want to use your code on a non-ASCII\nplatform. Starting in Perl v5.22, calls to it on ASCII platforms are optimized out, so\nthere's no performance penalty at all in adding it. Or you can simply use the other\nconstructs that don't require it.\n\nSee \"Further Resources\" for how to find all these names and numeric codes.\n\nEarlier releases caveats\n\nOn EBCDIC platforms, prior to v5.22, using \"\\N{U+...}\" doesn't work properly.\n\nPrior to v5.16, using \"\\N{...}\" with a character name (as opposed to a \"U+...\" code point)\nrequired a \"use charnames :full\".\n\nPrior to v5.14, there were some bugs in \"\\N{...}\" with a character name (as opposed to a\n\"U+...\" code point).\n\n\"charnames::stringvianame()\" was introduced in v5.14. Prior to that, \"charnames::vianame()\"\nshould work, but only if the argument is of the form \"U+...\". Your best bet there for\nruntime Unicode by character name is probably:\n\nuse charnames ();\nmy $hebrewaleffromname\n= pack(\"U\", charnames::vianame(\"HEBREW LETTER ALEF\"));\n" }, { "name": "Handling Unicode", "content": "Handling Unicode is for the most part transparent: just use the strings as usual. Functions\nlike \"index()\", \"length()\", and \"substr()\" will work on the Unicode characters; regular\nexpressions will work on the Unicode characters (see perlunicode and perlretut).\n\nNote that Perl considers grapheme clusters to be separate characters, so for example\n\nprint length(\"\\N{LATIN CAPITAL LETTER A}\\N{COMBINING ACUTE ACCENT}\"),\n\"\\n\";\n\nwill print 2, not 1. The only exception is that regular expressions have \"\\X\" for matching\nan extended grapheme cluster. (Thus \"\\X\" in a regular expression would match the entire\nsequence of both the example characters.)\n\nLife is not quite so transparent, however, when working with legacy encodings, I/O, and\ncertain special cases:\n" }, { "name": "Legacy Encodings", "content": "When you combine legacy data and Unicode, the legacy data needs to be upgraded to Unicode.\nNormally the legacy data is assumed to be ISO 8859-1 (or EBCDIC, if applicable).\n\nThe \"Encode\" module knows about many encodings and has interfaces for doing conversions\nbetween those encodings:\n\nuse Encode 'decode';\n$data = decode(\"iso-8859-3\", $data); # convert from legacy\n" }, { "name": "Unicode I/O", "content": "Normally, writing out Unicode data\n\nprint FH $somestringwithunicode, \"\\n\";\n\nproduces raw bytes that Perl happens to use to internally encode the Unicode string. Perl's\ninternal encoding depends on the system as well as what characters happen to be in the string\nat the time. If any of the characters are at code points 0x100 or above, you will get a\nwarning. To ensure that the output is explicitly rendered in the encoding you desire--and to\navoid the warning--open the stream with the desired encoding. Some examples:\n\nopen FH, \">:utf8\", \"file\";\n\nopen FH, \">:encoding(ucs2)\", \"file\";\nopen FH, \">:encoding(UTF-8)\", \"file\";\nopen FH, \">:encoding(shiftjis)\", \"file\";\n\nand on already open streams, use \"binmode()\":\n\nbinmode(STDOUT, \":utf8\");\n\nbinmode(STDOUT, \":encoding(ucs2)\");\nbinmode(STDOUT, \":encoding(UTF-8)\");\nbinmode(STDOUT, \":encoding(shiftjis)\");\n\nThe matching of encoding names is loose: case does not matter, and many encodings have\nseveral aliases. Note that the \":utf8\" layer must always be specified exactly like that; it\nis not subject to the loose matching of encoding names. Also note that currently \":utf8\" is\nunsafe for input, because it accepts the data without validating that it is indeed valid\nUTF-8; you should instead use \":encoding(UTF-8)\" (with or without a hyphen).\n\nSee PerlIO for the \":utf8\" layer, PerlIO::encoding and Encode::PerlIO for the \":encoding()\"\nlayer, and Encode::Supported for many encodings supported by the \"Encode\" module.\n\nReading in a file that you know happens to be encoded in one of the Unicode or legacy\nencodings does not magically turn the data into Unicode in Perl's eyes. To do that, specify\nthe appropriate layer when opening files\n\nopen(my $fh,'<:encoding(UTF-8)', 'anything');\nmy $lineofunicode = <$fh>;\n\nopen(my $fh,'<:encoding(Big5)', 'anything');\nmy $lineofunicode = <$fh>;\n\nThe I/O layers can also be specified more flexibly with the \"open\" pragma. See open, or look\nat the following example.\n\nuse open ':encoding(UTF-8)'; # input/output default encoding will be\n# UTF-8\nopen X, \">file\";\nprint X chr(0x100), \"\\n\";\nclose X;\nopen Y, \"); # this should print 0x100\nclose Y;\n\nWith the \"open\" pragma you can use the \":locale\" layer\n\nBEGIN { $ENV{LCALL} = $ENV{LANG} = 'ruRU.KOI8-R' }\n# the :locale will probe the locale environment variables like\n# LCALL\nuse open OUT => ':locale'; # russki parusski\nopen(O, \">koi8\");\nprint O chr(0x430); # Unicode CYRILLIC SMALL LETTER A = KOI8-R 0xc1\nclose O;\nopen(I, \"), \"\\n\"; # this should print 0xc1\nclose I;\n\nThese methods install a transparent filter on the I/O stream that converts data from the\nspecified encoding when it is read in from the stream. The result is always Unicode.\n\nThe open pragma affects all the \"open()\" calls after the pragma by setting default layers.\nIf you want to affect only certain streams, use explicit layers directly in the \"open()\"\ncall.\n\nYou can switch encodings on an already opened stream by using \"binmode()\"; see \"binmode\" in\nperlfunc.\n\nThe \":locale\" does not currently work with \"open()\" and \"binmode()\", only with the \"open\"\npragma. The \":utf8\" and \":encoding(...)\" methods do work with all of \"open()\", \"binmode()\",\nand the \"open\" pragma.\n\nSimilarly, you may use these I/O layers on output streams to automatically convert Unicode to\nthe specified encoding when it is written to the stream. For example, the following snippet\ncopies the contents of the file \"text.jis\" (encoded as ISO-2022-JP, aka JIS) to the file\n\"text.utf8\", encoded as UTF-8:\n\nopen(my $nihongo, '<:encoding(iso-2022-jp)', 'text.jis');\nopen(my $unicode, '>:utf8', 'text.utf8');\nwhile (<$nihongo>) { print $unicode $ }\n\nThe naming of encodings, both by the \"open()\" and by the \"open\" pragma allows for flexible\nnames: \"koi8-r\" and \"KOI8R\" will both be understood.\n\nCommon encodings recognized by ISO, MIME, IANA, and various other standardisation\norganisations are recognised; for a more detailed list see Encode::Supported.\n\n\"read()\" reads characters and returns the number of characters. \"seek()\" and \"tell()\"\noperate on byte counts, as does \"sysseek()\".\n\n\"sysread()\" and \"syswrite()\" should not be used on file handles with character encoding\nlayers, they behave badly, and that behaviour has been deprecated since perl 5.24.\n\nNotice that because of the default behaviour of not doing any conversion upon input if there\nis no default layer, it is easy to mistakenly write code that keeps on expanding a file by\nrepeatedly encoding the data:\n\n# BAD CODE WARNING\nopen F, \"file\";\nlocal $/; ## read in the whole file of 8-bit characters\n$t = ;\nclose F;\nopen F, \">:encoding(UTF-8)\", \"file\";\nprint F $t; ## convert to UTF-8 on output\nclose F;\n\nIf you run this code twice, the contents of the file will be twice UTF-8 encoded. A \"use\nopen ':encoding(UTF-8)'\" would have avoided the bug, or explicitly opening also the file for\ninput as UTF-8.\n\nNOTE: the \":utf8\" and \":encoding\" features work only if your Perl has been built with PerlIO,\nwhich is the default on most systems.\n" }, { "name": "Displaying Unicode As Text", "content": "Sometimes you might want to display Perl scalars containing Unicode as simple ASCII (or\nEBCDIC) text. The following subroutine converts its argument so that Unicode characters with\ncode points greater than 255 are displayed as \"\\x{...}\", control characters (like \"\\n\") are\ndisplayed as \"\\x..\", and the rest of the characters as themselves:\n\nsub nicestring {\njoin(\"\",\nmap { $ > 255 # if wide character...\n? sprintf(\"\\\\x{%04X}\", $) # \\x{...}\n: chr($) =~ /[[:cntrl:]]/ # else if control character...\n? sprintf(\"\\\\x%02X\", $) # \\x..\n: quotemeta(chr($)) # else quoted or as themselves\n} unpack(\"W*\", $[0])); # unpack Unicode characters\n}\n\nFor example,\n\nnicestring(\"foo\\x{100}bar\\n\")\n\nreturns the string\n\n'foo\\x{0100}bar\\x0A'\n\nwhich is ready to be printed.\n\n(\"\\\\x{}\" is used here instead of \"\\\\N{}\", since it's most likely that you want to see what\nthe native values are.)\n" }, { "name": "Special Cases", "content": "• Starting in Perl 5.28, it is illegal for bit operators, like \"~\", to operate on strings\ncontaining code points above 255.\n\n• The vec() function may produce surprising results if used on strings containing\ncharacters with ordinal values above 255. In such a case, the results are consistent with\nthe internal encoding of the characters, but not with much else. So don't do that, and\nstarting in Perl 5.28, a deprecation message is issued if you do so, becoming illegal in\nPerl 5.32.\n\n• Peeking At Perl's Internal Encoding\n\nNormal users of Perl should never care how Perl encodes any particular Unicode string\n(because the normal ways to get at the contents of a string with Unicode--via input and\noutput--should always be via explicitly-defined I/O layers). But if you must, there are\ntwo ways of looking behind the scenes.\n\nOne way of peeking inside the internal encoding of Unicode characters is to use\n\"unpack(\"C*\", ...\" to get the bytes of whatever the string encoding happens to be, or\n\"unpack(\"U0..\", ...)\" to get the bytes of the UTF-8 encoding:\n\n# this prints c4 80 for the UTF-8 bytes 0xc4 0x80\nprint join(\" \", unpack(\"U0(H2)*\", pack(\"U\", 0x100))), \"\\n\";\n\nYet another way would be to use the Devel::Peek module:\n\nperl -MDevel::Peek -e 'Dump(chr(0x100))'\n\nThat shows the \"UTF8\" flag in FLAGS and both the UTF-8 bytes and Unicode characters in\n\"PV\". See also later in this document the discussion about the \"utf8::isutf8()\"\nfunction.\n" }, { "name": "Advanced Topics", "content": "• String Equivalence\n\nThe question of string equivalence turns somewhat complicated in Unicode: what do you\nmean by \"equal\"?\n\n(Is \"LATIN CAPITAL LETTER A WITH ACUTE\" equal to \"LATIN CAPITAL LETTER A\"?)\n\nThe short answer is that by default Perl compares equivalence (\"eq\", \"ne\") based only on\ncode points of the characters. In the above case, the answer is no (because 0x00C1 !=\n0x0041). But sometimes, any CAPITAL LETTER A's should be considered equal, or even A's\nof any case.\n\nThe long answer is that you need to consider character normalization and casing issues:\nsee Unicode::Normalize, Unicode Technical Report #15, Unicode Normalization Forms\n and sections on case mapping in the\nUnicode Standard .\n\nAs of Perl 5.8.0, the \"Full\" case-folding of Case Mappings/SpecialCasing is implemented,\nbut bugs remain in \"qr//i\" with them, mostly fixed by 5.14, and essentially entirely by\n5.18.\n\n• String Collation\n\nPeople like to see their strings nicely sorted--or as Unicode parlance goes, collated.\nBut again, what do you mean by collate?\n\n(Does \"LATIN CAPITAL LETTER A WITH ACUTE\" come before or after \"LATIN CAPITAL LETTER A\nWITH GRAVE\"?)\n\nThe short answer is that by default, Perl compares strings (\"lt\", \"le\", \"cmp\", \"ge\",\n\"gt\") based only on the code points of the characters. In the above case, the answer is\n\"after\", since 0x00C1 > 0x00C0.\n\nThe long answer is that \"it depends\", and a good answer cannot be given without knowing\n(at the very least) the language context. See Unicode::Collate, and Unicode Collation\nAlgorithm \n" }, { "name": "Miscellaneous", "content": "• Character Ranges and Classes\n\nCharacter ranges in regular expression bracketed character classes ( e.g., \"/[a-z]/\") and\nin the \"tr///\" (also known as \"y///\") operator are not magically Unicode-aware. What\nthis means is that \"[A-Za-z]\" will not magically start to mean \"all alphabetic letters\"\n(not that it does mean that even for 8-bit characters; for those, if you are using\nlocales (perllocale), use \"/[[:alpha:]]/\"; and if not, use the 8-bit-aware property\n\"\\p{alpha}\").\n\nAll the properties that begin with \"\\p\" (and its inverse \"\\P\") are actually character\nclasses that are Unicode-aware. There are dozens of them, see perluniprops.\n\nStarting in v5.22, you can use Unicode code points as the end points of regular\nexpression pattern character ranges, and the range will include all Unicode code points\nthat lie between those end points, inclusive.\n\nqr/ [ \\N{U+03} - \\N{U+20} ] /xx\n\nincludes the code points \"\\N{U+03}\", \"\\N{U+04}\", ..., \"\\N{U+20}\".\n\nThis also works for ranges in \"tr///\" starting in Perl v5.24.\n\n• String-To-Number Conversions\n\nUnicode does define several other decimal--and numeric--characters besides the familiar 0\nto 9, such as the Arabic and Indic digits. Perl does not support string-to-number\nconversion for digits other than ASCII 0 to 9 (and ASCII \"a\" to \"f\" for hexadecimal). To\nget safe conversions from any Unicode string, use \"num()\" in Unicode::UCD.\n" }, { "name": "Questions With Answers", "content": "• Will My Old Scripts Break?\n\nVery probably not. Unless you are generating Unicode characters somehow, old behaviour\nshould be preserved. About the only behaviour that has changed and which could start\ngenerating Unicode is the old behaviour of \"chr()\" where supplying an argument more than\n255 produced a character modulo 255. \"chr(300)\", for example, was equal to \"chr(45)\" or\n\"-\" (in ASCII), now it is LATIN CAPITAL LETTER I WITH BREVE.\n\n• How Do I Make My Scripts Work With Unicode?\n\nVery little work should be needed since nothing changes until you generate Unicode data.\nThe most important thing is getting input as Unicode; for that, see the earlier I/O\ndiscussion. To get full seamless Unicode support, add \"use feature 'unicodestrings'\"\n(or \"use 5.012\" or higher) to your script.\n\n• How Do I Know Whether My String Is In Unicode?\n\nYou shouldn't have to care. But you may if your Perl is before 5.14.0 or you haven't\nspecified \"use feature 'unicodestrings'\" or \"use 5.012\" (or higher) because otherwise\nthe rules for the code points in the range 128 to 255 are different depending on whether\nthe string they are contained within is in Unicode or not. (See \"When Unicode Does Not\nHappen\" in perlunicode.)\n\nTo determine if a string is in Unicode, use:\n\nprint utf8::isutf8($string) ? 1 : 0, \"\\n\";\n\nBut note that this doesn't mean that any of the characters in the string are necessary\nUTF-8 encoded, or that any of the characters have code points greater than 0xFF (255) or\neven 0x80 (128), or that the string has any characters at all. All the \"isutf8()\" does\nis to return the value of the internal \"utf8ness\" flag attached to the $string. If the\nflag is off, the bytes in the scalar are interpreted as a single byte encoding. If the\nflag is on, the bytes in the scalar are interpreted as the (variable-length, potentially\nmulti-byte) UTF-8 encoded code points of the characters. Bytes added to a UTF-8 encoded\nstring are automatically upgraded to UTF-8. If mixed non-UTF-8 and UTF-8 scalars are\nmerged (double-quoted interpolation, explicit concatenation, or printf/sprintf parameter\nsubstitution), the result will be UTF-8 encoded as if copies of the byte strings were\nupgraded to UTF-8: for example,\n\n$a = \"ab\\x80c\";\n$b = \"\\x{100}\";\nprint \"$a = $b\\n\";\n\nthe output string will be UTF-8-encoded \"ab\\x80c = \\x{100}\\n\", but $a will stay byte-\nencoded.\n\nSometimes you might really need to know the byte length of a string instead of the\ncharacter length. For that use the \"bytes\" pragma and the \"length()\" function:\n\nmy $unicode = chr(0x100);\nprint length($unicode), \"\\n\"; # will print 1\nuse bytes;\nprint length($unicode), \"\\n\"; # will print 2\n# (the 0xC4 0x80 of the UTF-8)\nno bytes;\n\n• How Do I Find Out What Encoding a File Has?\n\nYou might try Encode::Guess, but it has a number of limitations.\n\n• How Do I Detect Data That's Not Valid In a Particular Encoding?\n\nUse the \"Encode\" package to try converting it. For example,\n\nuse Encode 'decode';\n\nif (eval { decode('UTF-8', $string, Encode::FBCROAK); 1 }) {\n# $string is valid UTF-8\n} else {\n# $string is not valid UTF-8\n}\n\nOr use \"unpack\" to try decoding it:\n\nuse warnings;\n@chars = unpack(\"C0U*\", $stringofbytesthatIthinkisutf8);\n\nIf invalid, a \"Malformed UTF-8 character\" warning is produced. The \"C0\" means \"process\nthe string character per character\". Without that, the \"unpack(\"U*\", ...)\" would work in\n\"U0\" mode (the default if the format string starts with \"U\") and it would return the\nbytes making up the UTF-8 encoding of the target string, something that will always work.\n\n• How Do I Convert Binary Data Into a Particular Encoding, Or Vice Versa?\n\nThis probably isn't as useful as you might think. Normally, you shouldn't need to.\n\nIn one sense, what you are asking doesn't make much sense: encodings are for characters,\nand binary data are not \"characters\", so converting \"data\" into some encoding isn't\nmeaningful unless you know in what character set and encoding the binary data is in, in\nwhich case it's not just binary data, now is it?\n\nIf you have a raw sequence of bytes that you know should be interpreted via a particular\nencoding, you can use \"Encode\":\n\nuse Encode 'fromto';\nfromto($data, \"iso-8859-1\", \"UTF-8\"); # from latin-1 to UTF-8\n\nThe call to \"fromto()\" changes the bytes in $data, but nothing material about the nature\nof the string has changed as far as Perl is concerned. Both before and after the call,\nthe string $data contains just a bunch of 8-bit bytes. As far as Perl is concerned, the\nencoding of the string remains as \"system-native 8-bit bytes\".\n\nYou might relate this to a fictional 'Translate' module:\n\nuse Translate;\nmy $phrase = \"Yes\";\nTranslate::fromto($phrase, 'english', 'deutsch');\n## phrase now contains \"Ja\"\n\nThe contents of the string changes, but not the nature of the string. Perl doesn't know\nany more after the call than before that the contents of the string indicates the\naffirmative.\n\nBack to converting data. If you have (or want) data in your system's native 8-bit\nencoding (e.g. Latin-1, EBCDIC, etc.), you can use pack/unpack to convert to/from\nUnicode.\n\n$nativestring = pack(\"W*\", unpack(\"U*\", $Unicodestring));\n$Unicodestring = pack(\"U*\", unpack(\"W*\", $nativestring));\n\nIf you have a sequence of bytes you know is valid UTF-8, but Perl doesn't know it yet,\nyou can make Perl a believer, too:\n\n$Unicode = $bytes;\nutf8::decode($Unicode);\n\nor:\n\n$Unicode = pack(\"U0a*\", $bytes);\n\nYou can find the bytes that make up a UTF-8 sequence with\n\n@bytes = unpack(\"C*\", $Unicodestring)\n\nand you can create well-formed Unicode with\n\n$Unicodestring = pack(\"U*\", 0xff, ...)\n\n• How Do I Display Unicode? How Do I Input Unicode?\n\nSee and \n\n• How Does Unicode Work With Traditional Locales?\n\nIf your locale is a UTF-8 locale, starting in Perl v5.26, Perl works well for all\ncategories; before this, starting with Perl v5.20, it works for all categories but\n\"LCCOLLATE\", which deals with sorting and the \"cmp\" operator. But note that the\nstandard \"Unicode::Collate\" and \"Unicode::Collate::Locale\" modules offer much more\npowerful solutions to collation issues, and work on earlier releases.\n\nFor other locales, starting in Perl 5.16, you can specify\n\nuse locale ':notcharacters';\n\nto get Perl to work well with them. The catch is that you have to translate from the\nlocale character set to/from Unicode yourself. See \"Unicode I/O\" above for how to\n\nuse open ':locale';\n\nto accomplish this, but full details are in \"Unicode and UTF-8\" in perllocale, including\ngotchas that happen if you don't specify \":notcharacters\".\n" }, { "name": "Hexadecimal Notation", "content": "The Unicode standard prefers using hexadecimal notation because that more clearly shows the\ndivision of Unicode into blocks of 256 characters. Hexadecimal is also simply shorter than\ndecimal. You can use decimal notation, too, but learning to use hexadecimal just makes life\neasier with the Unicode standard. The \"U+HHHH\" notation uses hexadecimal, for example.\n\nThe \"0x\" prefix means a hexadecimal number, the digits are 0-9 and a-f (or A-F, case doesn't\nmatter). Each hexadecimal digit represents four bits, or half a byte. \"print 0x..., \"\\n\"\"\nwill show a hexadecimal number in decimal, and \"printf \"%x\\n\", $decimal\" will show a decimal\nnumber in hexadecimal. If you have just the \"hex digits\" of a hexadecimal number, you can\nuse the \"hex()\" function.\n\nprint 0x0009, \"\\n\"; # 9\nprint 0x000a, \"\\n\"; # 10\nprint 0x000f, \"\\n\"; # 15\nprint 0x0010, \"\\n\"; # 16\nprint 0x0011, \"\\n\"; # 17\nprint 0x0100, \"\\n\"; # 256\n\nprint 0x0041, \"\\n\"; # 65\n\nprintf \"%x\\n\", 65; # 41\nprintf \"%#x\\n\", 65; # 0x41\n\nprint hex(\"41\"), \"\\n\"; # 65\n" }, { "name": "Further Resources", "content": "• Unicode Consortium\n\n\n\n• Unicode FAQ\n\n\n\n• Unicode Glossary\n\n\n\n• Unicode Recommended Reading List\n\nThe Unicode Consortium has a list of articles and books, some of which give a much more\nin depth treatment of Unicode: \n\n• Unicode Useful Resources\n\n\n\n• Unicode and Multilingual Support in HTML, Fonts, Web Browsers and Other Applications\n\n\n\n• UTF-8 and Unicode FAQ for Unix/Linux\n\n\n\n• Legacy Character Sets\n\n \n\n• You can explore various information from the Unicode data files using the \"Unicode::UCD\"\nmodule.\n" } ] }, "UNICODE IN OLDER PERLS": { "content": "If you cannot upgrade your Perl to 5.8.0 or later, you can still do some Unicode processing\nby using the modules \"Unicode::String\", \"Unicode::Map8\", and \"Unicode::Map\", available from\nCPAN. If you have the GNU recode installed, you can also use the Perl front-end\n\"Convert::Recode\" for character conversions.\n\nThe following are fast conversions from ISO 8859-1 (Latin-1) bytes to UTF-8 bytes and back,\nthe code works even with older Perl 5 versions.\n\n# ISO 8859-1 to UTF-8\ns/([\\x80-\\xFF])/chr(0xC0|ord($1)>>6).chr(0x80|ord($1)&0x3F)/eg;\n\n# UTF-8 to ISO 8859-1\ns/([\\xC2\\xC3])([\\x80-\\xBF])/chr(ord($1)<<6&0xC0|ord($2)&0x3F)/eg;\n", "subsections": [] }, "SEE ALSO": { "content": "perlunitut, perlunicode, Encode, open, utf8, bytes, perlretut, perlrun, Unicode::Collate,\nUnicode::Normalize, Unicode::UCD\n", "subsections": [] }, "ACKNOWLEDGMENTS": { "content": "Thanks to the kind readers of the perl5-porters@perl.org, perl-unicode@perl.org,\nlinux-utf8@nl.linux.org, and unicore@unicode.org mailing lists for their valuable feedback.\n", "subsections": [ { "name": "AUTHOR, COPYRIGHT, AND LICENSE", "content": "Copyright 2001-2011 Jarkko Hietaniemi . Now maintained by Perl 5 Porters.\n\nThis document may be distributed under the same terms as Perl itself.\n\n\n\nperl v5.34.0 2026-06-23 PERLUNIINTRO(1)" } ] } } } }