UTF-8#fallback and auto-detection

{{See also|Universal Coded Character Set#History}}

The International Organization for Standardization (ISO) set out to compose a universal multi-byte character set in 1989. The draft ISO 10646 standard contained a non-required annex called UTF-1 that provided a byte stream encoding of its 32-bit code points. This encoding was not satisfactory on performance grounds, among other problems, and the biggest problem was probably that it did not have a clear separation between ASCII and non-ASCII: new UTF-1 tools would be backward compatible with ASCII-encoded text, but UTF-1-encoded text could confuse existing code expecting ASCII (or extended ASCII), because it could contain continuation bytes in the range {{mono|0x21}}–{{mono|0x7E}} that meant something else in ASCII, e.g., {{mono|0x2F}} for /, the Unix path directory separator.

In July 1992, the X/Open committee XoJIG was looking for a better encoding. Dave Prosser of Unix System Laboratories submitted a proposal for one that had faster implementation characteristics and introduced the improvement that 7-bit ASCII characters would only represent themselves; multi-byte sequences would only include bytes with the high bit set. The name File System Safe UCS Transformation Format (FSS-UTF){{cite web|url=https://www.unicode.org/L2/Historical/wg20-n193-fss-utf.pdf|title=File System Safe UCS — Transformation Format (FSS-UTF) - X/Open Preliminary Specification|website=unicode.org}} and most of the text of this proposal were later preserved in the final specification.{{cite journal |title=Appendix F. FSS-UTF / File System Safe UCS Transformation format |journal=The Unicode Standard 1.1 |url=https://www.unicode.org/versions/Unicode1.1.0/appF.pdf |access-date=2016-06-07 |url-status=live |archive-url=https://web.archive.org/web/20160607215950/https://www.unicode.org/versions/Unicode1.1.0/appF.pdf |archive-date=2016-06-07}}{{cite web |title=FSS-UTF, UTF-2, UTF-8, and UTF-16 |author-first=Kenneth |author-last=Whistler |date=2001-06-12 |url=https://unicode.org/mail-arch/unicode-ml/y2001-m06/0318.html |access-date=2006-06-07 |url-status=live |archive-url=https://web.archive.org/web/20160607220249/https://unicode.org/mail-arch/unicode-ml/y2001-m06/0318.html |archive-date=2016-06-07 }}{{cite web |url=https://www.cl.cam.ac.uk/~mgk25/ucs/utf-8-history.txt |title=UTF-8 history |author-first=Rob |author-last=Pike |author-link=Rob Pike |date=2003-04-30 |access-date=2012-09-07}} In August 1992, this proposal was circulated by an IBM X/Open representative to interested parties. A modification by Ken Thompson of the Plan 9 operating system group at Bell Labs made it self-synchronizing, letting a reader start anywhere and immediately detect character boundaries, at the cost of being somewhat less bit-efficient than the previous proposal. It also abandoned the use of biases that prevented overlong encodings.At that time subtraction was slower than bit logic on many computers, and speed was considered necessary for acceptance.{{citation needed|date=October 2024}} Thompson's design was outlined on September 2, 1992, on a placemat in a New Jersey diner with Rob Pike. In the following days, Pike and Thompson implemented it and updated Plan 9 to use it throughout,{{cite book |chapter-url=https://www.cl.cam.ac.uk/~mgk25/ucs/UTF-8-Plan9-paper.pdf |chapter=Hello World or Καλημέρα κόσμε or こんにちは 世界 |title=Proceedings of the Winter 1993 USENIX Conference |first1=Rob |last1=Pike |first2=Ken |last2=Thompson |year=1993}} and then communicated their success back to X/Open, which accepted it as the specification for FSS-UTF.

UTF-8 was first officially presented at the USENIX conference in San Diego, from January 25 to 29, 1993.{{cite web|url=https://www.usenix.org/legacy/publications/library/proceedings/sd93/| title=USENIX Winter 1993 Conference Proceedings|website=usenix.org}} The Internet Engineering Task Force adopted UTF-8 in its Policy on Character Sets and Languages in RFC 2277 (BCP 18) for future internet standards work in January 1998, replacing Single Byte Character Sets such as Latin-1 in older RFCs.{{cite IETF |rfc=2277 |bcp=18 |title=IETF Policy on Character Sets and Languages |date=January 1998 |last1=Alvestrand |first1=Harald T. |author-link=Harald Alvestrand |publisher=IETF}}

In November 2003, UTF-8 was restricted by {{IETF RFC|3629}} to match the constraints of the UTF-16 character encoding: explicitly prohibiting code points corresponding to the high and low surrogate characters removed more than 3% of the three-byte sequences, and ending at {{tt|U+10FFFF}} removed more than 48% of the four-byte sequences and all five- and six-byte sequences.{{cite web |author-last=Pike |author-first=Rob |author-link=Rob Pike |date=2012-09-06 |title=UTF-8 turned 20 years old yesterday |url=https://plus.google.com/u/0/101960720994009339267/posts/Rz1udTvtiMg |url-status=dead |archive-url=https://commandcenter.blogspot.com/2020/01/utf-8-turned-20-years-old-in-2012.html |archive-date=2020-01-26 |access-date=2012-09-07}}

UTF-8 encodes code points in one to four bytes, depending on the value of the code point. In the following table, the characters {{mono|u}} to {{mono|z}} are replaced by the bits of the code point, from the positions {{mono|U+uvwxyz}}:

The first 128 code points (ASCII) need 1 byte. The next 1,920 code points need two bytes to encode, which covers the remainder of almost all Latin-script alphabets, and also IPA extensions, Greek, Cyrillic, Coptic, Armenian, Hebrew, Arabic, Syriac, Thaana and N'Ko alphabets, as well as Combining Diacritical Marks. Three bytes are needed for the remaining 61,440 codepoints of the Basic Multilingual Plane (BMP), including most Chinese, Japanese and Korean characters. Four bytes are needed for the 1,048,576 non-BMP code points, which include emoji, less common CJK characters, and other useful characters.{{Cite web |last=Lunde |first=Ken |date=2022-01-09 |title=2022 Top Ten List: Why Support Beyond-BMP Code Points? |url=https://ken-lunde.medium.com/2022-top-ten-list-why-support-beyond-bmp-code-points-6a946d7735f9 |website=Medium |language=en|access-date=2024-01-07}}

UTF-8 is a prefix code and it is unnecessary to read past the last byte of a code point to decode it. Unlike many earlier multi-byte text encodings such as Shift-JIS, it is self-synchronizing so searches for short strings or characters are possible and that the start of a code point can be found from a random position by backing up at most 3 bytes. The values chosen for the lead bytes means sorting a list of UTF-8 strings puts them in the same order as sorting UTF-32 strings.

{{anchor|overlong encodings}}

Using a row in the above table to encode a code point less than "First code point" (thus using more bytes than necessary) is termed an overlong encoding. These are a security problem because they allow character sequences such as malicious JavaScript and ../ to bypass security validations, which has been reported in numerous high-profile products such as Microsoft's IIS web server{{ cite report | first = Marvin |last = Marin | date = 2000-10-17 | title = Windows NT UNICODE vulnerability analysis | department = Web server folder traversal | id = MS00-078 | series = Malware FAQ | website=SANS Institute | url=https://www.sans.org/resources/malwarefaq/wnt-unicode.php | url-status=dead | archive-url=https://web.archive.org/web/20140827001204/http://www.sans.org/security-resources/malwarefaq/wnt-unicode.php | archive-date=Aug 27, 2014 }} and Apache's Tomcat servlet container.{{ cite web | title = CVE-2008-2938 | year = 2008 | website = National Vulnerability Database (nvd.nist.gov) | publisher = U.S. National Institute of Standards and Technology | url = https://web.nvd.nist.gov/view/vuln/detail?vulnId=CVE-2008-2938 }} Overlong encodings should therefore be considered an error and never decoded. Modified UTF-8 allows an overlong encoding of {{tt|U+0000}}.

The chart below gives the detailed meaning of each byte in a stream encoded in UTF-8.

{{UTF-8 byte map}}

Not all sequences of bytes are valid UTF-8. A UTF-8 decoder should be prepared for:

• Bytes that never appear in UTF-8: {{tt|0xC0}}, {{tt|0xC1}}, {{tt|0xF5}}{{ndash}}{{tt|0xFF}}
• A "continuation byte" ({{tt|0x80}}{{ndash}}{{tt|0xBF}}) at the start of a character
• A non-continuation byte (or the string ending) before the end of a character
• An overlong encoding ({{tt|0xE0}} followed by less than {{tt|0xA0}}, or {{tt|0xF0}} followed by less than {{tt|0x90}})
• A 4-byte sequence that decodes to a value greater than {{tt|U+10FFFF}} ({{tt|0xF4}} followed by {{tt|0x90}} or greater)

Many of the first UTF-8 decoders would decode these, ignoring incorrect bits. Carefully crafted invalid UTF-8 could make them either skip or create ASCII characters such as {{mono|NUL}}, slash, or quotes, leading to security vulnerabilities. It is also common to throw an exception or truncate the string at an error{{ cite web | title = DataInput | series = Java Platform SE 8 | website = docs.oracle.com | url = https://docs.oracle.com/javase/8/docs/api/java/io/DataInput.html | access-date = 2021-03-24 }} but this turns what would otherwise be harmless errors (i.e. "file not found") into a denial of service, for instance early versions of Python 3.0 would exit immediately if the command line or environment variables contained invalid UTF-8.{{ cite web | title = Non-decodable bytes in system character interfaces | date = 2009-04-22 | website = python.org | url = https://www.python.org/dev/peps/pep-0383/ | access-date = 2014-08-13 }}

{{nobr|RFC 3629}} states "Implementations of the decoding algorithm MUST protect against decoding invalid sequences."{{cite IETF |title=UTF-8, a transformation format of ISO 10646 |rfc=3629 |std=63 |last1=Yergeau |first1=F. |date=November 2003 |publisher=IETF |access-date=August 20, 2020}} The Unicode Standard requires decoders to: "... treat any ill-formed code unit sequence as an error condition. This guarantees that it will neither interpret nor emit an ill-formed code unit sequence." The standard now recommends replacing each error with the replacement character "�" ({{tt|U+FFFD}}) and continue decoding.

Some decoders consider the sequence {{mono|E1,A0,20}} (a truncated 3-byte code followed by a space) as a single error. This is not a good idea as a search for a space character would find the one hidden in the error. Since Unicode 6 (October 2010){{ cite report | title = Unicode 6.0.0 | date = October 2010 | website = unicode.org | url = https://www.unicode.org/versions/Unicode6.0.0/ }} the standard (chapter 3) has recommended a "best practice" where the error is either one continuation byte, or ends at the first byte that is disallowed, so {{mono|E1,A0,20}} is a two-byte error followed by a space. This means an error is no more than three bytes long and never contains the start of a valid character, and there are {{val|21952|fmt=commas}} different possible errors. Technically this makes UTF-8 no longer a prefix code (the decoder has to read one byte past some errors to figure out they are an error), but searching still works if the searched-for string does not contain any errors.

Making each byte be an error, in which case {{mono|E1,A0,20}} is two errors followed by a space, also still allows searching for a valid string. This means there are only 128 different errors which makes it practical to store the errors in the output string, or replace them with characters from a legacy encoding.

Only a small subset of possible byte strings are error-free UTF-8: several bytes cannot appear; a byte with the high bit set cannot be alone; and in a truly random string a byte with a high bit set has only a {{frac|1|15}} chance of starting a valid UTF-8 character. This has the consequence of making it easy to detect if a legacy text encoding is accidentally used instead of UTF-8, making conversion of a system to UTF-8 easier and avoiding the need to require a Byte Order Mark or any other metadata.

Since RFC 3629 (November 2003), the high and low surrogates used by UTF-16 ({{tt|U+D800}} through {{tt|U+DFFF}}) are not legal Unicode values, and their UTF-8 encodings must be treated as an invalid byte sequence. These encodings all start with {{tt|0xED}} followed by {{tt|0xA0}} or higher. This rule is often ignored as surrogates are allowed in Windows filenames and this means there must be a way to store them in a string.{{ cite web | title = Change Windows filesystem encoding to UTF-8 | id = PEP 529 | website = Python.org |language = en | url = https://www.python.org/dev/peps/pep-0529/ | access-date = 2022-05-10 }} UTF-8 that allows these surrogate halves has been (informally) called {{visible anchor|WTF-8}},{{cite web | title = The WTF-8 encoding | url = https://simonsapin.github.io/wtf-8/}} while another variation that also encodes all non-BMP characters as two surrogates (6 bytes instead of 4) is called CESU-8.

If the Unicode byte-order mark {{tt|U+FEFF}} is at the start of a UTF-8 file, the first three bytes will be {{mono|0xEF}}, {{mono|0xBB}}, {{mono|0xBF}}.

The Unicode Standard neither requires nor recommends the use of the BOM for UTF-8, but warns that it may be encountered at the start of a file trans-coded from another encoding.{{citation | chapter-url = https://www.unicode.org/versions/Unicode15.0.0/ch02.pdf | title = The Unicode Standard — Version 15.0.0 | chapter = Chapter 2 | page = 39 }} While ASCII text encoded using UTF-8 is backward compatible with ASCII, this is not true when Unicode Standard recommendations are ignored and a BOM is added. A BOM can confuse software that isn't prepared for it but can otherwise accept UTF-8, e.g. programming languages that permit non-ASCII bytes in string literals but not at the start of the file. Nevertheless, there was and still is software that always inserts a BOM when writing UTF-8, and refuses to correctly interpret UTF-8 unless the first character is a BOM (or the file only contains ASCII).{{Cite web |title=UTF-8 and Unicode FAQ for Unix/Linux |url=https://www.cl.cam.ac.uk/~mgk25/unicode.html}}

{{See also|Comparison of Unicode encodings}}

{{unsourced section|find=UTF-8|find2=comparison to UTF-16|date=December 2024}}

For a long time there was considerable argument as to whether it was better to process text in UTF-16 or in UTF-8.

The primary advantage of UTF-16 is that the Windows API required it for access to all Unicode characters (UTF-8 was not fully supported in Windows until May 2019). This caused several libraries such as Qt to also use UTF-16 strings which propagates this requirement to non-Windows platforms.

In the early days of Unicode there were no characters greater than {{tt|U+FFFF}} and combining characters were rarely used, so the 16-bit encoding was effectively fixed-size. Some believed fixed-size encoding could make processing more efficient, but any such advantages were lost as soon as UTF-16 became variable width as well.

The code points {{tt|U+0800}}{{ndash}}{{tt|U+FFFF}} take 3 bytes in UTF-8 but only 2 in UTF-16. This led to the idea that text in Chinese and other languages would take more space in UTF-8. However, text is only larger if there are more of these code points than 1-byte ASCII code points, and this rarely happens in the real-world documents due to spaces, newlines, digits, punctuation, English words, and (depending on document format) markup.

UTF-8 has the advantages of being trivial to retrofit to any system that could handle an extended ASCII, not having byte-order problems, and taking about half the space for any language using mostly Latin letters.

File:UTF-8 takes over.png

File:Utf8webgrowth.svg

{{See also|Popularity of text encodings}}

UTF-8 has been the most common encoding for the World Wide Web since 2008.{{cite web | first=Mark |last=Davis |author-link=Mark Davis (Unicode) | date=2008-05-05 | title=Moving to Unicode 5.1 | website=Official Google blog |language=en | url=https://googleblog.blogspot.com/2008/05/moving-to-unicode-51.html | access-date=2023-03-13 }} {{As of|2025|01}}, UTF-8 is used by 98.5% of surveyed web sites.{{Cite web|url=https://w3techs.com/technologies/cross/character_encoding/ranking|title=Usage Survey of Character Encodings broken down by Ranking |website=W3Techs |language = en | date = January 2025 |access-date=2025-01-07}} Although many pages only use ASCII characters to display content, very few websites now declare their encoding to only be ASCII instead of UTF-8.{{cite web |url=https://w3techs.com/technologies/details/en-usascii |title = Usage statistics and market share of ASCII for websites | date = January 2025 | website = W3Techs | access-date = 2025-01-07 }} Virtually all countries and languages have 95% or more use of UTF-8 encodings on the web.

Many standards only support UTF-8, e.g. JSON exchange requires it (without a byte-order mark (BOM)).{{ cite IETF | last = Bray | first = Tim | editor-last = Bray | editor-first = Tim | date = December 2017 | title = The JavaScript Object Notation (JSON) Data Interchange Format | publisher = IETF | doi = 10.17487/RFC8259 | access-date = 16 February 2018 | rfc = 8259 }} UTF-8 is also the recommendation from the WHATWG for HTML and DOM specifications, and stating "UTF-8 encoding is the most appropriate encoding for interchange of Unicode"{{ cite web | title = Encoding Standard | website = encoding.spec.whatwg.org | url = https://encoding.spec.whatwg.org/#preface | access-date = 2020-04-15 }} and the Internet Mail Consortium recommends that all e‑mail programs be able to display and create mail using UTF-8.{{ cite web | url = https://www.imc.org/mail-i18n.html | title = Using International Characters in Internet Mail | publisher = Internet Mail Consortium | date = 1998-08-01 | access-date = 2007-11-08 | url-status = dead | archive-url = https://web.archive.org/web/20071026103104/https://www.imc.org/mail-i18n.html | archive-date = 2007-10-26}}{{ cite web | title = Encoding Standard | website = encoding.spec.whatwg.org |language = en | url = https://encoding.spec.whatwg.org/#security-background | access-date = 2018-11-15 }} The World Wide Web Consortium recommends UTF-8 as the default encoding in XML and HTML (and not just using UTF-8, also declaring it in metadata), "even when all characters are in the ASCII range ... Using non-UTF-8 encodings can have unexpected results".{{ cite report | section = Specifying the document's character encoding | title = HTML 5.2 | date = 14 December 2017 | publisher = World Wide Web Consortium | url = https://www.w3.org/TR/html5/document-metadata.html | section-url = https://www.w3.org/TR/html5/document-metadata.html#charset | access-date = 2018-06-03 | mode = cs1 }}

Many software programs have the ability to read/write UTF-8. It may require the user to change options from the normal settings, or may require a BOM (byte-order mark) as the first character to read the file. Examples of software supporting UTF-8 include Microsoft Word,{{ cite web | title=Choose text encoding when you open and save files | website=Microsoft Support (support.microsoft.com) | url=https://support.microsoft.com/en-us/office/choose-text-encoding-when-you-open-and-save-files-60d59c21-88b5-4006-831c-d536d42fd861 | access-date=2021-11-01 }}{{ cite web | title=UTF-8 - Character encoding of Microsoft Word DOC and DOCX files? | website=Stack Overflow | url=https://stackoverflow.com/questions/28172022/character-encoding-of-microsoft-word-doc-and-docx-files | access-date=2021-11-01 }}{{ cite web | title = Exporting a UTF-8 .txt file from Word | website = support.3playmedia.com | date = 14 March 2023 | url = https://support.3playmedia.com/hc/en-us/articles/227730088-Exporting-a-UTF-8-txt-file-from-Word }} Microsoft Excel (2016 and later),{{ cite web | title = Are XLSX files UTF-8 encoded, by definition? | series = Excel | website = Stack Overflow | url = https://stackoverflow.com/questions/45194771/are-xlsx-files-utf-8-encoded-by-definition | access-date = 2021-11-01 }}{{ cite web | author1 = Abhinav, Ankit | author2 = Xu, Jazlyn | date = April 13, 2020 | title = How to open UTF-8 CSV file in Excel without mis-conversion of characters in Japanese and Chinese language for both Mac and Windows? | website = Microsoft Support Community | language = en-US | url = https://answers.microsoft.com/en-us/msoffice/forum/all/how-to-open-utf-8-csv-file-in-excel-without-mis/1eb15700-d235-441e-8b99-db10fafff3c2 | access-date = 2021-11-01 }} Google Drive, LibreOffice and most databases.

Software that "defaults" to UTF-8 (meaning it writes it without the user changing settings, and it reads it without a BOM) has become more common since 2010.{{ cite web | last=Galloway |first=Matt | date=October 2012 | title=Character encoding for iOS developers; or, UTF-8 what now? | website=www.galloway.me.uk | language=en-UK | url=https://www.galloway.me.uk/2012/10/character-encoding-for-ios-developers-utf8/ | access-date=2021-01-02 | quote = ... in reality, you usually just assume UTF-8 since that is by far the most common encoding. }} Windows Notepad, in all currently supported versions of Windows, defaults to writing UTF-8 without a BOM (a change from {{nobr|Windows 7}} Notepad), bringing it into line with most other text editors.{{ cite web | title=Windows 10 Notepad is getting better UTF-8 encoding support | website=BleepingComputer | url=https://www.bleepingcomputer.com/news/microsoft/windows-10-notepad-is-getting-better-utf-8-encoding-support/ | access-date=2021-03-24 | quote=Microsoft is now defaulting to saving new text files as UTF-8 without BOM, as shown below. | language=en-us }} Some system files on Windows 11 require UTF-8{{ cite web | title = Customize the Windows 11 Start menu | url=https://docs.microsoft.com/en-us/windows-hardware/customize/desktop/customize-the-windows-11-start-menu | access-date=2021-06-29 | website=docs.microsoft.com | language=en-us | quote=Make sure your LayoutModification.json uses UTF-8 encoding. }} with no requirement for a BOM, and almost all files on macOS and Linux are required to be UTF-8 without a BOM.{{citation needed|date=June 2021}} Programming languages that default to UTF-8 for I/O include Ruby 3.0,{{ cite web | title = Set default for Encoding.default_external to UTF-8 on Windows | series = Ruby master | id = Feature #16604 | website = Ruby Issue Tracking System (bugs.ruby-lang.org) | url = https://bugs.ruby-lang.org/issues/16604 | access-date = 2022-08-01 }}{{ cite web | title = Feature #12650: Use UTF-8 encoding for ENV on Windows | series = Ruby master | website = Ruby Issue Tracking System (bugs.ruby-lang.org) | url = https://bugs.ruby-lang.org/issues/12650 | access-date = 2022-08-01 }} R 4.2.2,{{ cite web | title = New features in R 4.2.0 | date = 2022-04-01 | website = R bloggers (r-bloggers.com) | series = The Jumping Rivers Blog | url = https://www.r-bloggers.com/2022/04/new-features-in-r-4-2-0/ | access-date = 2022-08-01 | language = en-US }} Raku and Java 18.{{ cite web | title = UTF-8 by default | id = JEP 400 | website = openjdk.java.net | url = https://openjdk.java.net/jeps/400 | access-date=2022-03-30 }} Although the current version of Python requires an option to open() to read/write UTF-8,{{ cite web | title = add a new UTF-8 mode | website = peps.python.org | id = PEP 540 | url = https://peps.python.org/pep-0540/ | access-date = 2022-09-23 }} plans exist to make UTF-8 I/O the default in Python 3.15.{{ cite web | title = Make UTF-8 mode default | website = peps.python.org | id = PEP 686 | url = https://peps.python.org/pep-0686/ | access-date=2023-07-26 }} C++23 adopts UTF-8 as the only portable source code file format.{{ cite report | title = Support for UTF-8 as a portable source file encoding | year = 2022 | id = p2295r6 | website = open-std.org | url = https://www.open-std.org/jtc1/sc22/wg21/docs/papers/2022/p2295r6.pdf }}

Backwards compatibility is a serious impediment to changing code and APIs using UTF-16 to use UTF-8, but this is happening. {{As of|2019|05}}, Microsoft added the capability for an application to set UTF-8 as the "code page" for the Windows API, removing the need to use UTF-16; and more recently has recommended programmers use UTF-8,{{ cite web | title=Use UTF-8 code pages in Windows apps | website=Microsoft Learn | date=20 August 2024 |language=en-us | url=https://learn.microsoft.com/en-us/windows/apps/design/globalizing/use-utf8-code-page | access-date=2024-09-24}} and even states "UTF-16 [...] is a unique burden that Windows places on code that targets multiple platforms".{{ cite web | title=UTF-8 support in the Microsoft GDK | series = Microsoft Game Development Kit (GDK) | website = Microsoft Learn |language=en-us | url=https://learn.microsoft.com/en-us/gaming/gdk/_content/gc/system/overviews/utf-8 | access-date = 2023-03-05 }}

The default string primitive in Go,{{ cite report | section=Source code representation | title=The Go Programming Language Specification | website=golang.org | section-url=https://golang.org/ref/spec#Source_code_representation | access-date=2021-02-10 }}

Julia, Rust, Swift (since version 5),{{ cite web | last=Tsai |first=Michael J. | date=21 March 2019 | title=UTF-8 string in Swift 5 | type=blog post |language=en | url=https://mjtsai.com/blog/2019/03/21/utf-8-string-in-swift-5/ | access-date=2021-03-15 }} and PyPy{{ cite web | title=PyPy v7.1 released; now uses UTF-8 internally for Unicode strings | department=Mattip | date=2019-03-24 | website=PyPy status blog | url=https://morepypy.blogspot.com/2019/03/pypy-v71-released-now-uses-utf-8.html | access-date=2020-11-21 }} uses UTF-8 internally in all cases. Python (since version 3.3) uses UTF-8 internally for Python C API extensions{{ cite web | title = Flexible String Representation | id = PEP 393 | website = Python.org |language=en | url = https://peps.python.org/pep-0393 | access-date = 2022-05-18 }}{{Cite web |title=Common Object Structures |url=https://docs.python.org/3/c-api/structures.html |access-date=2024-05-29 |website=Python documentation |language=en}} and sometimes for strings{{ cite web | title=Unicode objects and codecs | url=https://docs.python.org/3/c-api/unicode.html | access-date=2023-08-19 |website=Python documentation | quote=UTF-8 representation is created on demand and cached in the Unicode object.}} and a future version of Python is planned to store strings as UTF-8 by default.{{ cite web | title=PEP 623 – remove wstr from Unicode | website=Python.org |language=en | url=https://www.python.org/dev/peps/pep-0623/ | access-date=2020-11-21 }}{{ cite web | last=Wouters |first=Thomas | date=2023-07-11 | title=Python 3.12.0 beta 4 released | website = Python Insider (pythoninsider.blogspot.com) | type = blog post | url=https://pythoninsider.blogspot.com/2023/07/pleased-to-announce-release-of-python-3.html | access-date=2023-07-26 | quote=The deprecated wstr and wstr_length members of the C implementation of unicode objects were removed, per PEP 623. }} Modern versions of Microsoft Visual Studio use UTF-8 internally.{{ cite web | title=validate-charset (validate for compatible characters) | website=docs.microsoft.com |language=en-us | url=https://docs.microsoft.com/en-us/cpp/build/reference/validate-charset-validate-for-compatible-characters | access-date=2021-07-19 | quote=Visual Studio uses UTF-8 as the internal character encoding during conversion between the source character set and the execution character set. }} Microsoft's SQL Server 2019 added support for UTF-8, and using it results in a 35% speed increase, and "nearly 50% reduction in storage requirements."{{ cite web | title = Introducing UTF-8 support for SQL Server | date = 2019-07-02 | website = techcommunity.microsoft.com | url = https://techcommunity.microsoft.com/t5/sql-server/introducing-utf-8-support-for-sql-server/ba-p/734928 | access-date = 2021-08-24 | language = en-US }}

{{anchor|Modified UTF-8}}

Java internally uses UTF-16 for the char data type and, consequentially, the Character, String, and the StringBuffer classes,{{cite web |title=Character (Java SE 24 & JDK 24) |url=https://docs.oracle.com/en/java/javase/24/docs/api/java.base/java/lang/Character.html#unicode |year=2025 |publisher=Oracle Corporation |access-date=2025-04-08}} but for I/O uses Modified UTF-8 (MUTF-8), in which the null character {{tt|U+0000}} uses the two-byte overlong encoding {{tt|0xC0}}, {{tt|0x80}}, instead of just {{tt|0x00}}.{{cite web |title=Java SE documentation for Interface java.io.DataInput, subsection on Modified UTF-8 |url=https://docs.oracle.com/javase/8/docs/api/java/io/DataInput.html#modified-utf-8 |year=2015 |publisher=Oracle Corporation |access-date=2015-10-16}} Modified UTF-8 strings never contain any actual null bytes but can contain all Unicode code points including {{tt|U+0000}},{{cite web |url=https://docs.oracle.com/javase/specs/jvms/se8/html/jvms-4.html#jvms-4.4.7 |title=The Java Virtual Machine Specification, section 4.4.7: "The CONSTANT_Utf8_info Structure" |publisher=Oracle Corporation |year=2015 |access-date=2015-10-16}} which allows such strings (with a null byte appended) to be processed by traditional null-terminated string functions. Java reads and writes normal UTF-8 to files and streams,{{Javadoc:SE|java/io|InputStreamReader}} and {{Javadoc:SE|java/io|OutputStreamWriter}} but it uses Modified UTF-8 for object serialization,{{cite web |title=Java Object Serialization Specification, chapter 6: Object Serialization Stream Protocol, section 2: Stream Elements |url=https://docs.oracle.com/javase/8/docs/platform/serialization/spec/protocol.html#a8299 |year=2010 |publisher=Oracle Corporation |access-date=2015-10-16}}{{Javadoc:SE|java/io|DataInput}} and {{Javadoc:SE|java/io|DataOutput}} for the Java Native Interface,{{cite web |url=https://docs.oracle.com/javase/8/docs/technotes/guides/jni/spec/types.html#modified_utf_8_strings |title=Java Native Interface Specification, chapter 3: JNI Types and Data Structures, section: Modified UTF-8 Strings |publisher=Oracle Corporation |year=2015 |access-date=2015-10-16}} and for embedding constant strings in class files.{{cite web |title=The Java Virtual Machine Specification, section 4.4.7: "The CONSTANT_Utf8_info Structure" |url=https://docs.oracle.com/javase/specs/jvms/se8/html/jvms-4.html#jvms-4.4.7 |publisher=Oracle Corporation |year=2015 |access-date=2015-10-16}} The dex format defined by Dalvik also uses the same modified UTF-8 to represent string values.{{cite web |url=https://source.android.com/tech/dalvik/dex-format.html |title=ART and Dalvik |work=Android Open Source Project |access-date=2013-04-09 |url-status=dead |archive-url=https://web.archive.org/web/20130426010617/https://source.android.com/tech/dalvik/dex-format.html |archive-date=2013-04-26 }} Tcl also uses the same modified UTF-8{{cite web |title=UTF-8 bit by bit |date=2001-02-28 |url=https://wiki.tcl-lang.org/page/UTF-8+bit+by+bit |access-date=2022-09-03 |website=Tcler's Wiki}} as Java for internal representation of Unicode data, but uses strict CESU-8 for external data. All known Modified UTF-8 implementations also treat the surrogate pairs as in CESU-8.

The Raku programming language (formerly Perl 6) uses utf-8 encoding by default for I/O (Perl 5 also supports it); though that choice in Raku also implies "normalization into Unicode NFC (normalization form canonical). In some cases the user will want to ensure no normalization is done; for this utf8-c8" can be used.{{Cite web |title=encoding {{!}} Raku Documentation |url=https://docs.raku.org/routine/encoding |access-date=2024-10-06 |website=docs.raku.org}} That UTF-8 Clean-8 variant, implemented by Raku, is an encoder/decoder that preserves bytes as is (even illegal UTF-8 sequences) and allows for Normal Form Grapheme synthetics.{{Cite web |title=Unicode {{!}} Raku Documentation |url=https://docs.raku.org/language/unicode#UTF8-C8 |access-date=2024-10-06 |website=docs.raku.org}}

Version 3 of the Python programming language treats each byte of an invalid UTF-8 bytestream as an error (see also changes with new UTF-8 mode in Python 3.7{{Cite web|title=PEP 540 -- Add a new UTF-8 Mode|url=https://www.python.org/dev/peps/pep-0540/|access-date=2021-03-24|website=Python.org|language=en}}); this gives 128 different possible errors. Extensions have been created to allow any byte sequence that is assumed to be UTF-8 to be losslessly transformed to UTF-16 or UTF-32, by translating the 128 possible error bytes to 128 reserved code points, and transforming those code points back to error bytes to output UTF-8. The most common approach is to translate the codes to {{tt|U+DC80}}...{{tt|U+DCFF}} which are low (trailing) surrogate values and thus "invalid" UTF-16, as used by Python's PEP 383 (or "surrogateescape") approach.{{cite web |id=PEP 383 |title=Non-decodable Bytes in System Character Interfaces |url=https://www.python.org/dev/peps/pep-0383 |publisher=Python Software Foundation |language=en |first=Martin |last=von Löwis |date=2009-04-22}} Another encoding called MirBSD OPTU-8/16 converts them to {{tt|U+EF80}}...{{tt|U+EFFF}} in a Private Use Area.{{cite web |title=RTFM optu8to16(3), optu8to16vis(3) |url=https://www.mirbsd.org/htman/i386/man3/optu8to16.htm |website=www.mirbsd.org}} In either approach, the byte value is encoded in the low eight bits of the output code point. These encodings are needed if invalid UTF-8 is to survive translation to and then back from the UTF-16 used internally by Python, and as Unix filenames can contain invalid UTF-8 it is necessary for this to work.{{cite web |url=https://www.unicode.org/reports/tr36/#EnablingLosslessConversion |last1=Davis |first1=Mark |author-link1=Mark Davis (Unicode) |first2=Michel |last2=Suignard |title=3.7 Enabling Lossless Conversion |work=Unicode Security Considerations |id=Unicode Technical Report #36 |year=2014}}

The official name for the encoding is {{code|UTF-8}}, the spelling used in all Unicode Consortium documents. The hyphen-minus is required and no spaces are allowed. Some other names used are:

• Most standards are also case-insensitive and utf-8 is often used.{{citation needed|date=March 2023}}
• Web standards (which include CSS, HTML, XML, and HTTP headers) also allow {{code|utf8}} and many other aliases.{{cite web|url=https://encoding.spec.whatwg.org/#names-and-labels|title=Encoding Standard § 4.2. Names and labels|publisher=WHATWG|access-date=2018-04-29}}
• The official Internet Assigned Numbers Authority lists {{code|csUTF8}} as the only alias,{{cite web |publisher=Internet Assigned Numbers Authority |url=https://www.iana.org/assignments/character-sets |title=Character Sets |date=2013-01-23 |access-date=2013-02-08}} which is rarely used.
• In some locales {{code|UTF-8N}} means UTF-8 without a byte-order mark (BOM), and in this case {{code|UTF-8}} may imply there is a BOM.{{cite web |url=https://suika.fam.cx/~wakaba/wiki/sw/n/BOM |title=BOM | work = suikawiki |archive-url=https://web.archive.org/web/20090117052232/https://suika.fam.cx/~wakaba/wiki/sw/n/BOM |archive-date=2009-01-17 |language=ja}}{{cite web |author-last=Davis |author-first=Mark |author-link=Mark Davis (Unicode) |title=Forms of Unicode |publisher=IBM |url=https://www-128.ibm.com/developerworks/library/utfencodingforms/index.html |access-date=2013-09-18 |archive-url=https://web.archive.org/web/20050506211548/https://www-128.ibm.com/developerworks/library/utfencodingforms/index.html |archive-date=2005-05-06}}
• In Windows, UTF-8 is codepage 65001{{Cite web |url=https://www.dostips.com/forum/viewtopic.php?t=5357 |title=UTF-8 codepage 65001 in Windows 7 - part I |author=Liviu |quote=Previously under XP (and, unverified, but probably Vista, too) for loops simply did not work while codepage 65001 was active |language=en-gb |date=2014-02-07 |access-date=2018-01-30}} with the symbolic name CP_UTF8 in source code.
• In MySQL, UTF-8 is called utf8mb4,{{Cite web |title=MySQL :: MySQL 8.0 Reference Manual :: 10.9.1 The utf8mb4 Character Set (4-Byte UTF-8 Unicode Encoding) |url=https://dev.mysql.com/doc/refman/8.0/en/charset-unicode-utf8mb4.html |work=MySQL 8.0 Reference Manual |publisher=Oracle Corporation |access-date=2023-03-14}} while {{code|utf8}} and {{code|utf8mb3}} refer to the obsolete CESU-8 variant.{{Cite web |title=MySQL :: MySQL 8.0 Reference Manual :: 10.9.2 The utf8mb3 Character Set (3-Byte UTF-8 Unicode Encoding) |url=https://dev.mysql.com/doc/refman/8.0/en/charset-unicode-utf8mb3.html |work=MySQL 8.0 Reference Manual |publisher=Oracle Corporation |access-date=2023-02-24}}
• In Oracle Database (since version 9.0), AL32UTF8{{Cite web |title=Database Globalization Support Guide |url=https://docs.oracle.com/cd/E11882_01/server.112/e10729/ch6unicode.htm |access-date=2023-03-16 |website=docs.oracle.com |language=en}} means UTF-8, while {{code|UTF-8}} means CESU-8.
• In HP PCL, the Symbol-ID for UTF-8 is 18N.{{Cite web|url=https://pclhelp.com/pcl-symbol-sets/ |archive-url=https://web.archive.org/web/20150219212843/http://pclhelp.com/pcl-symbol-sets/|url-status=dead|archive-date=2015-02-19|title=HP PCL Symbol Sets {{!}} Printer Control Language (PCL & PXL) Support Blog|date=2015-02-19|access-date=2018-01-30}}

There are several current definitions of UTF-8 in various standards documents:

• {{IETF RFC|3629|link=no}} / STD 63 (2003), which establishes UTF-8 as a standard internet protocol element
• {{IETF RFC|5198|link=no}} defines UTF-8 NFC for Network Interchange (2008)
• ISO/IEC 10646:2020/Amd 1:2023[https://www.iso.org/standard/83362.html ISO/IEC 10646].
• The Unicode Standard, Version 16.0.0 (2024)[https://www.unicode.org/versions/Unicode16.0.0/ The Unicode Standard, Version 16.0] [https://www.unicode.org/versions/Unicode15.0.0/ch03.pdf#G31703 §3.9 D92, §3.10 D95], 2021.

They supersede the definitions given in the following obsolete works:

• The Unicode Standard, Version 2.0, Appendix A (1996)
• ISO/IEC 10646-1:1993 Amendment 2 / Annex R (1996)
• {{IETF RFC|2044|link=no}} (1996)
• {{IETF RFC|2279|link=no}} (1998)
• The Unicode Standard, Version 3.0, §2.3 (2000) plus Corrigendum #1 : UTF-8 Shortest Form (2000)
• Unicode Standard Annex #27: Unicode 3.1 (2001)[https://www.unicode.org/reports/tr27/tr27-3.html Unicode Standard Annex #27: Unicode 3.1], 2001.
• The Unicode Standard, Version 5.0 (2006)[https://www.unicode.org/versions/Unicode5.0.0/ The Unicode Standard, Version 5.0] [https://www.unicode.org/versions/Unicode5.0.0/ch03.pdf §3.9–§3.10 ch. 3], 2006.
• The Unicode Standard, Version 6.0 (2010)[https://www.unicode.org/versions/Unicode6.0.0/ The Unicode Standard, Version 6.0] [https://www.unicode.org/versions/Unicode6.0.0/ch03.pdf §3.9 D92, §3.10 D95], 2010.

They are all the same in their general mechanics, with the main differences being on issues such as allowed range of code point values and safe handling of invalid input.

• {{annotated link|Character encodings in HTML}}
• {{annotated link|Comparison of Unicode encodings}}
• {{annotated link|GB 18030}}
• {{annotated link|Iconv}}
• {{annotated link|Unicode and email}}
• {{annotated link|Unicode and HTML}}
• {{annotated link|UTF-EBCDIC}}

{{reflist}}

• [https://doc.cat-v.org/plan_9/4th_edition/papers/utf Original UTF-8 paper] ([https://web.archive.org/web/20000917055036/http://plan9.bell-labs.com/sys/doc/utf.pdf or pdf]) for Plan 9 from Bell Labs
• [https://www.cl.cam.ac.uk/~mgk25/ucs/utf-8-history.txt History of UTF-8 by Rob Pike]
• {{YouTube|id=MijmeoH9LT4|title=Characters, Symbols and the Unicode Miracle}}

{{Unicode navigation}}

{{Character encoding}}

{{Rob Pike navbox}}

{{Ken Thompson navbox}}

Category:Character encoding

Category:Computer-related introductions in 1993

Category:Encodings

Category:Unicode Transformation Formats