Skip to content
Locale Lab

Reference · 72 terms

Glossary

Definitions for all technical terms in the field guide, linked to the relevant chapter and normative specification.

Standards and identifiers

BCP 47 tags, Unicode specifications, and CLDR datasets that identify locales and store regional rules.

14 terms
Accept-Language

The HTTP request header that transmits the user's preferred locales with relative quality weights (Accept-Language: de-CH, de;q=0.9, en;q=0.8). Servers parse this header as the default input for locale negotiation when the user has not saved an explicit application language preference.

Chapter: Locale FallbackRFC 9110 §12.5.4: Accept-Language ↗See also: Locale negotiation, BCP 47

BCP 47

The IETF specification for language tags such as en-US, zh-Hant-TW, and de-DE-u-co-phonebk. It defines a fixed subtag order: language, script, region, variants, and extensions. HTML lang, HTTP Accept-Language, and JavaScript Intl constructors all consume BCP 47 tags. BCP 47 combines RFC 5646 (tag syntax) and RFC 4647 (tag matching).

Chapter: OrientationRFC 5646: Tags for Identifying Languages ↗See also: Language subtag, Script subtag, Region subtag, Locale negotiation

CLDR

The Unicode Common Locale Data Repository. CLDR is the standard database of locale rules, storing date patterns, number formats, plural rules, collation tailorings, list patterns, and translated display names. Browsers, operating systems, and ICU read their locale data from CLDR.

Chapter: OrientationUnicode CLDR project ↗See also: ICU, ECMA-402 (the Intl object), Plural category

ECMA-402 (the Intl object)

The ECMAScript Internationalization API specification that defines JavaScript's Intl namespace (NumberFormat, DateTimeFormat, PluralRules, ListFormat, Collator, Segmenter, DisplayNames, and Locale). Because browsers and Node.js ship ECMA-402 in the runtime, web applications can format and compare localized data without bundling CLDR tables in client JavaScript.

Chapter: Locale FormattersECMA-402: ECMAScript Internationalization API ↗See also: ICU, CLDR

Fallback chain

The ordered sequence of locale tags a runtime checks when an exact locale match is missing. RFC 4647 lookup strips subtags from right to left (pt-BR -> pt -> default locale) and never steps sideways to a sibling region such as pt-PT.

Chapter: Locale FallbackSee also: Locale negotiation, Language subtag

ICU

International Components for Unicode. ICU is the reference C/C++, Java, and Rust library suite that executes Unicode algorithms and CLDR formatting rules. Major web browsers and Node.js build their Intl implementations on ICU.

Chapter: OrientationICU project ↗See also: CLDR, ECMA-402 (the Intl object), MessageFormat (ICU)

Language subtag

The mandatory primary subtag in a BCP 47 tag, such as pt in pt-BR. It uses a two-letter or three-letter ISO 639 code. Subsequent subtags narrow the script, region, or variant.

Chapter: Locale FallbackSee also: Script subtag, Region subtag, Fallback chain

Locale

An identifier that specifies a user's language, script, region, and formatting preferences. Formatting, sorting, and plural selection depend on the active locale. For example, de-DE and de-AT share a language subtag but differ in regional conventions. Every Intl constructor accepts a locale tag or array of tags as its first argument.

Chapter: OrientationSee also: BCP 47, CLDR

Locale negotiation

The algorithm that resolves a user's requested locale list against the application's supported locales. Runtimes use RFC 4647 prefix lookup, CLDR best-fit distance matching, or HTTP Accept-Language weight parsing. Because the three algorithms resolve sibling regions differently, teams should ship region-neutral base catalogs.

Chapter: Locale FallbackRFC 4647: Matching of Language Tags ↗See also: Fallback chain, Accept-Language, BCP 47

MessageFormat (ICU)

The ICU syntax for externalized messages that contain typed placeholders, plural branches, and select branches, such as {count, plural, one {# file} other {# files}}. Keeping grammatical branches inside the message catalog allows translators to add or modify forms without changing application code.

Chapter: Plurals and GenderICU User Guide: Formatting Messages ↗See also: Plural category, Select message, String externalization

MessageFormat 2.0 (MF2)

The redesigned Unicode message formatting specification (UTS #35 Part 9, stabilized in CLDR 47). MessageFormat 2.0 replaces nested inline braces with top-level .input and .match declarations, {{...}} quoted patterns, and extensible :number and :date format functions. JavaScript runtimes currently support MF2 via the messageformat polyfill while the TC39 Intl.MessageFormat proposal advances.

Chapter: Plurals and GenderUnicode MessageFormat 2.0 specification (UTS #35 Part 9) ↗See also: MessageFormat (ICU), Plural category, Select message

Region subtag

A two-letter ISO 3166-1 country code or three-digit UN M.49 area code in a BCP 47 tag, such as BR in pt-BR or 419 in es-419 (Latin America). It selects regional conventions for date order, currency, and vocabulary.

Chapter: Locale FallbackSee also: Language subtag, Script subtag

Script subtag

A four-letter ISO 15924 code in title case that identifies the writing system, such as Hans and Hant in zh-Hans and zh-Hant. In CLDR locale matching, script similarity ranks higher than region similarity because serving Simplified Chinese to a Traditional Chinese reader is a larger mismatch than serving another Traditional Chinese region.

Chapter: Locale FallbackSee also: Language subtag, Region subtag, Locale negotiation

Unicode

The universal character encoding standard that assigns a unique numeric code point to every character across supported writing systems. Unicode also specifies character properties and text algorithms for normalization (UAX #15), bidirectional layout (UAX #9), segmentation (UAX #29), line breaking (UAX #14), and collation (UTS #10).

Chapter: Unicode StringsThe Unicode Standard ↗See also: Code point, CLDR

Text and scripts

Code units, code points, grapheme clusters, bidirectional text layout, line breaking, and font fallback.

19 terms
Bidi algorithm

The Unicode Bidirectional Algorithm (UAX #9) that orders mixed left-to-right (LTR) and right-to-left (RTL) text for display. Every code point carries a directional class: strong, weak, or neutral. Because neutral characters (spaces, punctuation, and symbols) inherit direction from neighboring strong characters, unisolated LTR values inside RTL sentences can reorder across boundaries.

Chapter: Bidirectional TextUAX #9: Unicode Bidirectional Algorithm ↗See also: Directional isolate (FSI/PDI), RTL mirroring

Case folding

A locale-independent transformation defined by the Unicode Standard for case-insensitive string comparison. Security blocklists and internal identifier checks must use locale-independent case folding or a pinned root locale rather than host-default locale case conversion.

Chapter: Unicode StringsSee also: Case mapping

Case mapping

The conversion of text between lowercase, uppercase, and titlecase forms. Case mapping is language-sensitive and not always 1-to-1: Turkish maps I (U+0049) to dotless ı (U+0131), and German maps lowercase ß to uppercase SS.

Chapter: Unicode StringsSee also: Case folding

Code point

The numeric value assigned to a character in the Unicode codespace (U+0000 to U+10FFFF), such as U+0041 for A. JavaScript's str.codePointAt() and [...str] iterate by Unicode code points, whereas str.length counts 16-bit UTF-16 code units.

Chapter: Unicode StringsSee also: Code unit, Grapheme cluster

Code unit

The minimal bit combination that represents a unit of encoded text for processing or storage (8 bits in UTF-8, 16 bits in UTF-16). In JavaScript (UTF-16), code points above U+FFFF require a two-unit surrogate pair, so '😀'.length returns 2.

Chapter: Unicode StringsSee also: Code point, Grapheme cluster

Combining mark

A Unicode character that renders on top of, below, or around the preceding base character, such as Latin diacritics, Hebrew vowel points, or Devanagari vowel signs. For example, e (U+0065) followed by a combining acute accent (U+0301) forms a single grapheme cluster (é) composed of two code points.

Chapter: Unicode StringsSee also: NFC, NFD, Grapheme cluster

CSS logical properties

Flow-relative CSS properties that reference inline-start and inline-end instead of physical left and right (such as margin-inline-start, padding-inline-end, and inset-inline-start). When dir="rtl" is set on the document root, logical properties mirror layout automatically from a single stylesheet.

Chapter: RTL MirroringCSS Logical Properties and Values Level 1 ↗See also: RTL mirroring

Directional isolate (FSI/PDI)

The Unicode bidirectional isolate controls (U+2068 FSI, U+2066 LRI, U+2067 RLI, closed by U+2069 PDI) and their HTML equivalents (<bdi> and dir="auto"). Isolates resolve the direction of an embedded text span independently and treat the span as a single neutral object in the outer sentence.

Chapter: Bidirectional TextUAX #9: Unicode Bidirectional Algorithm ↗See also: Bidi algorithm

Font fallback

The per-character lookup mechanism in CSS font-family. For each character in a text run, the browser selects the first font in the stack that contains a glyph for that code point, falling back to an operating-system default font when none of the listed fonts cover the script.

Chapter: CJK Font FallbackSee also: Han unification, Webfont

Grapheme cluster

A sequence of one or more Unicode code points that renders as a single user-perceived character, as defined by UAX #29. A grapheme cluster can combine a base letter with combining marks or link multiple emoji with Zero Width Joiners. Use new Intl.Segmenter(locale, { granularity: 'grapheme' }) to count or slice grapheme clusters.

Chapter: Unicode StringsUAX #29: Unicode Text Segmentation ↗See also: Code point, Combining mark, Word segmentation

Han unification

The Unicode design principle that assigns a single shared code point to cognate Chinese, Japanese, and Korean ideographs (直 is U+76F4 across all three markets) while leaving regional stroke differences to the font and the HTML lang attribute.

Chapter: CJK Font FallbackWikipedia: Han unification ↗See also: Font fallback, Webfont

Kinsoku shori

Japanese typographic line-breaking rules (standardized in JIS X 4051 and W3C JLREQ) that prohibit closing punctuation (。, 、, 」) and small kana from starting a new line and prohibit opening brackets (「) from ending a line. CSS line-break: strict enforces the full prohibition set.

Chapter: Line BreakingW3C: Requirements for Japanese Text Layout ↗See also: Line breaking, Word segmentation

Line breaking

The algorithm (UAX #14) that determines legal line-wrap opportunities in text. Western scripts break at whitespace and hyphens, CJK scripts break between ideographs subject to punctuation prohibitions, and Southeast Asian scripts break at dictionary word boundaries. CSS line-break, word-break, overflow-wrap, and hyphens configure browser wrapping behavior.

Chapter: Line BreakingUAX #14: Unicode Line Breaking Algorithm ↗See also: Kinsoku shori, Word segmentation

NFC

Unicode Normalization Form C (Canonical Decomposition, followed by Canonical Composition). NFC replaces decomposed base-plus-accent sequences with single precomposed code points wherever possible. Normalize external input with .normalize('NFC') at system boundaries so canonically equivalent strings compare equal.

Chapter: Unicode StringsUAX #15: Unicode Normalization Forms ↗See also: NFD, Combining mark

NFD

Unicode Normalization Form D (Canonical Decomposition). NFD splits precomposed characters such as é (U+00E9) into their base character (e, U+0065) plus combining marks (U+0301). NFD café contains 5 code points, whereas NFC café contains 4.

Chapter: Unicode StringsUAX #15: Unicode Normalization Forms ↗See also: NFC, Combining mark

RTL mirroring

Reversing horizontal UI layout geometry for right-to-left scripts such as Arabic, Hebrew, Persian, and Urdu. Navigation bars, sidebars, paddings, and directional progression arrows flip horizontally, whereas media playback controls, clocks, and numerals remain unflipped.

Chapter: RTL MirroringSee also: CSS logical properties, Bidi algorithm

Text expansion

The change in horizontal width and vertical height when source strings are translated into target languages. Under W3C and IBM guidelines, English labels under 10 characters expand by 100% to 200%, tapering to 30% for strings over 70 characters, while CJK translations frequently contract.

Chapter: Text ExpansionW3C: Text size in translation ↗See also: Pseudolocalization

Webfont

A font file delivered over the network via CSS @font-face. Because CJK fonts contain tens of thousands of glyphs and weigh several megabytes, production CJK webfont deployments partition fonts into unicode-range subsets so browsers download only the slices used on the page.

Chapter: CJK Font FallbackSee also: Font fallback, Han unification

Word segmentation

Identifying word boundaries in scripts that do not separate words with spaces. Thai, Lao, and Khmer require dictionary-based word segmentation to wrap lines, and Chinese and Japanese require segmentation for search indexing and double-click selection. Intl.Segmenter with { granularity: 'word' } exposes runtime word segmentation.

Chapter: Line BreakingUAX #29: Unicode Text Segmentation ↗See also: Line breaking, Grapheme cluster

Formatting

CLDR and ECMA-402 rules for numbers, currencies, dates, calendars, time zones, lists, and collation.

20 terms
Calendar system

The reckoning system that maps an instant in time to a year, month, and day (such as gregory, buddhist, islamic-umalqura, hebrew, persian, or japanese). CLDR defines each locale's default calendar, and BCP 47 overrides it via the -u-ca- extension subtag.

Chapter: CalendarsUTS #35 Part 4: Dates ↗See also: Era (calendar), Week info (first day / weekend), BCP 47, CLDR

Collation

Locale-sensitive string ordering for sorting and searching, exposed in JavaScript via Intl.Collator. Unlike Array.prototype.sort(), which compares UTF-16 code units, collation orders characters according to the Unicode Collation Algorithm and the active locale's CLDR tailoring.

Chapter: SortingUTS #10: Unicode Collation Algorithm ↗See also: UCA (Unicode Collation Algorithm), Tailoring

Decimal separator

The symbol that separates the integer and fractional parts of a formatted number, such as . in en-US, , in de-DE, and ٫ (U+066B) in ar-EG. Parsers can inspect Intl.NumberFormat(locale).formatToParts() to discover a locale's decimal and grouping separators at runtime.

Chapter: Input ParsingUTS #35 part 3: Numbers ↗See also: Numbering system, Wire format

DST (daylight saving time)

The seasonal adjustment of local clocks observed by some jurisdictions. Transition dates, shift amounts (60 minutes in most observing zones, 30 minutes on Lord Howe Island), and participation rules are set by legislation and tracked in the IANA time zone database.

Chapter: Time Zonestz database: Theory and pragmatics ↗See also: IANA time zone database, UTC offset, Wall-clock time

Era (calendar)

A named historical period that numbers years relative to its start date. In the Japanese calendar, each imperial accession begins a new era (Shōwa 64 ended on January 7, 1989, Heisei 1 began on January 8, 1989, and Reiwa 1 began on May 1, 2019). Era boundaries are maintained as data tables in CLDR and ICU.

Chapter: CalendarsSee also: Calendar system, CLDR

IANA time zone database

The canonical registry (also called tzdata or the Olson database) that maps geographic region identifiers such as Europe/Berlin or America/New_York to their historical and future UTC offsets and daylight saving time rules.

Chapter: Time ZonesIANA: Time Zone Database ↗See also: ECMA-402 (the Intl object)

Measurement system

The system of physical units (metric, US customary, or UK imperial) used in a region. Intl.NumberFormat with style: 'unit' formats the numeric value and unit identifier passed by the caller without converting the value between unit systems.

Chapter: Locale FormattersMDN: Intl.NumberFormat unit style ↗See also: CLDR, ECMA-402 (the Intl object)

Minor units (ISO 4217)

The standard number of decimal fraction digits defined for a currency in ISO 4217: USD and EUR use 2, JPY and KRW use 0, and BHD, KWD, and OMR use 3. Intl.NumberFormat applies each currency's minor-unit precision automatically.

Chapter: Locale FormattersISO 4217: Currency codes ↗See also: ECMA-402 (the Intl object)

Mononym

A personal name consisting of a single word, common in parts of Indonesia, Malaysia, and South India. Forms that enforce both a required First name and a required Last name reject valid mononyms.

Chapter: Names and AddressesW3C: Personal names around the world ↗See also: Patronymic

Numbering system

The script of ten decimal digits a locale uses to format numbers, identified by a CLDR key such as latn (0–9), arab (٠–٩), arabext (۰–۹), or fullwide (0–9). Both formatters and input parsers must account for the active numbering system.

Chapter: Input ParsingSee also: ECMA-402 (the Intl object), CLDR

Ordinal

The CLDR plural rule set for ranking numbers (1st, 2nd, 3rd, 4th), distinct from cardinal quantity rules. In English, new Intl.PluralRules('en-US', { type: 'ordinal' }) maps 1, 2, 3, and 4 to one, two, few, and other.

Chapter: Plurals and GenderSee also: Plural category

Patronymic

A personal name component derived from the given name of a parent rather than an inherited family surname. In Iceland, Guðmundsdóttir means daughter of Guðmundur, and directories sort people by given name rather than patronymic.

Chapter: Names and AddressesW3C: Personal names around the world ↗See also: Mononym

Plural category

One of the six grammatical quantity categories defined by CLDR: zero, one, two, few, many, and other. Each language maps numbers to the subset of categories its grammar requires. Russian maps 1, 21, and 31 to one, and Arabic uses all six categories.

Chapter: Plurals and GenderCLDR: Plural Rules ↗See also: Ordinal, MessageFormat (ICU), CLDR

Select message

An ICU MessageFormat expression that branches on an explicit string key ({gender, select, female {...} male {...} other {...}}) rather than a numeric plural category. Every select block requires an other fallback branch.

Chapter: Plurals and GenderSee also: MessageFormat (ICU), Plural category

Tailoring

A locale-specific override table in CLDR that customizes the default Unicode Collation Algorithm order. For example, Swedish sorts å, ä, and ö after z, and German phonebook collation (de-DE-u-co-phonebk) expands ä to ae.

Chapter: SortingSee also: Collation, UCA (Unicode Collation Algorithm)

UCA (Unicode Collation Algorithm)

The multi-level string comparison algorithm defined in UTS #10. UCA evaluates primary differences (base letters) before secondary differences (diacritics) and tertiary differences (case and width) using the Default Unicode Collation Element Table (DUCET).

Chapter: SortingUTS #10: Unicode Collation Algorithm ↗See also: Collation, Tailoring

UTC offset

The difference in hours and minutes between local time and Coordinated Universal Time at a single instant, such as +02:00 or +05:45. Because jurisdictions change DST schedules and standard offsets, an offset alone cannot represent a time zone for future events.

Chapter: Time ZonesSee also: IANA time zone database, DST (daylight saving time), Wall-clock time

Wall-clock time

The local date and time displayed on a clock in a specific location (2026-10-28T09:00), distinct from a UTC instant. Store future human-scheduled appointments as wall-clock time plus an IANA time zone identifier (Europe/London) so rule changes in tzdata do not shift the local appointment hour.

Chapter: Time ZonesSee also: UTC offset, DST (daylight saving time), IANA time zone database

Week info (first day / weekend)

Regional calendar metadata exposed by Intl.Locale.prototype.getWeekInfo(): the first day of the week (firstDay), the weekend days (weekend), and the minimum days required for Week 1 (minimalDays). For example, en-US starts Sunday with a Saturday-Sunday weekend, de-DE starts Monday, and ar-SA and he-IL start Sunday with a Friday-Saturday weekend.

Chapter: CalendarsTC39 Intl Locale Info proposal ↗See also: Calendar system, CLDR, ECMA-402 (the Intl object)

Wire format

The canonical, locale-independent representation used to store and transmit data between systems: ISO 8601 (YYYY-MM-DD) for dates and IEEE 754 numbers or integer minor units for monetary amounts. Format values for the user's locale only at the presentation layer.

Chapter: Input ParsingRFC 3339: Date and Time on the Internet ↗See also: Decimal separator, Minor units (ISO 4217)

Workflow and QA

String externalization, interchange file formats, translation memory leverage, pseudolocalization, and MQM quality scoring.

19 terms
Character limit

A maximum length constraint applied to a string field or translation segment. Because str.length counts UTF-16 code units, [...str].length counts code points, and Intl.Segmenter counts grapheme clusters, character limits must specify which unit they measure, while UI truncation depends on rendered pixel width.

Chapter: Text ExpansionSee also: Text expansion, Grapheme cluster, Code unit

Error severity

The penalty weight assigned to a logged localization defect in an MQM scorecard: neutral (0), minor (1), major (5), or critical (25). Severity weights ensure that functional, legal, or safety defects outweigh minor stylistic issues.

Chapter: LQA ScoringSee also: MQM (Multidimensional Quality Metrics), LQA (language quality assurance)

Fuzzy match

A partial match (typically 75% to 99% similarity) between a new source segment and a segment stored in the translation memory. Translators edit the pre-filled match, and language service providers bill the segment at a tiered discount below the new-word rate.

Chapter: Translation PipelineSee also: Translation memory, Source word

g11n (globalization)

The organizational and engineering program that combines internationalization (i18n) and localization (l10n) to ship and operate a product across global markets (g + 11 letters + n), often grouped under the acronym GILT (Globalization, Internationalization, Localization, Translation).

Chapter: OrientationSee also: i18n (internationalization), l10n (localization), t9n (translation)

gettext PO

The GNU gettext Portable Object catalog format, consisting of plain-text msgid (source) and msgstr (target) pairs, developer comments (#.), and header-driven plural expressions (msgid_plural and msgstr[n]).

Chapter: Translation PipelineGNU gettext manual: PO Files ↗See also: XLIFF, String externalization

i18n (internationalization)

Software architecture and engineering that decouples an application from any single language or locale (i + 18 letters + n). Internationalization replaces hardcoded strings, date patterns, and left-to-right layout assumptions with locale-aware APIs so localization can add markets without code changes.

Chapter: OrientationW3C: Localization vs. Internationalization ↗See also: l10n (localization), g11n (globalization), t9n (translation)

l10n (localization)

Adapting an internationalized product for a specific target market (l + 10 letters + n). Localization encompasses message translation, date and currency formatting, typography, imagery, payment methods, and regulatory compliance for each target locale.

Chapter: OrientationW3C: Localization vs. Internationalization ↗See also: i18n (internationalization), t9n (translation), String externalization

LQA (language quality assurance)

In-context linguistic and functional review of localized software against a structured error typology to verify terminology, grammar, locale formatting, and visual layout before release.

Chapter: LQA ScoringSee also: MQM (Multidimensional Quality Metrics), Error severity, Termbase, MT post-editing

MQM (Multidimensional Quality Metrics)

An analytic framework (ISO 5060) for evaluating translation quality across seven error dimensions (Terminology, Accuracy, Linguistic conventions, Style, Locale conventions, Audience appropriateness, and Design and markup) weighted by severity per 1,000 words.

Chapter: LQA ScoringMQM error typology ↗See also: LQA (language quality assurance), Error severity

MT post-editing

Human linguist revision of machine translation output, standardized in ISO 18587 as either light post-editing (correcting factual and grammatical errors for comprehension) or full post-editing (achieving publishable human-translation quality).

ISO 18587: Post-editing of machine translation output ↗See also: LQA (language quality assurance), Translation memory

Pseudolocalization

An automated build transformation that replaces source characters with accented lookalikes (Ĥéĺĺø), adds expansion padding (⟦…⟧), or forces right-to-left rendering (U+202E…U+202C) to catch hardcoded strings, concatenation, and layout clipping in CI before human translation.

Chapter: PseudolocalizationMicrosoft Learn: Pseudo-locales ↗See also: String externalization, Text expansion

Source word

The standard billing and scheduling unit in translation projects, equal to one word in the source-language catalog. Vendors price against source word counts because the source count is fixed before translation starts.

Chapter: Translation PipelineSee also: Fuzzy match, Translation memory

String externalization

Extracting user-facing strings from source code into keyed resource catalogs (such as ICU JSON, XLIFF, or gettext PO files) so translators can translate complete sentences without modifying source code.

Chapter: Translation PipelineSee also: gettext PO, XLIFF, Pseudolocalization

String freeze

The release milestone after which source strings in the repository are locked against edits so translation and linguistic QA can complete against a stable catalog.

See also: String externalization, TMS (translation management system)

t9n (translation)

Converting source-language text into a target language (t + 9 letters + n). Translation is one component of localization. It converts externalized messages, whereas internationalization (i18n) handles runtime formatting, plural selection, and layout.

Chapter: OrientationSee also: l10n (localization), i18n (internationalization), Translation memory

Termbase

A curated terminology database of approved product terms, UI labels, and do-not-translate brand names for each target locale. Unlike a translation memory, which stores full sentences, a termbase enforces term-level consistency during translation and automated QA.

Chapter: Translation PipelineSee also: Translation memory, LQA (language quality assurance)

TMS (translation management system)

The workflow platform that automates string synchronization with version control, applies translation memory and termbase matching, assigns tasks to translators and reviewers, and exports localized resource bundles back to the repository.

See also: Translation memory, XLIFF, String freeze

Translation memory

A bilingual database of previously translated source-target segment pairs. When new strings enter the pipeline, the CAT tool queries the translation memory to reuse 100% exact matches and pre-fill 75% to 99% fuzzy matches at discounted per-word rates.

Chapter: Translation PipelineSee also: Fuzzy match, Termbase, Source word

XLIFF

XML Localization Interchange File Format. XLIFF is the OASIS standard for exchanging bilingual localization data, segment state, translator notes, and inline placeholder tags (<ph>) between software repositories and CAT tools.

Chapter: Translation PipelineOASIS: XLIFF Version 2.0 ↗See also: gettext PO, TMS (translation management system), String externalization