Typecast Autotag

Text preprocessing SDK for TTS — convert text to speech-friendly format

Features • Installation • Quick Start • API • Tags

Language Support

The JavaScript and browser package accepts every SSFM v3.0 TTS language:

Language tier	Status	Language codes
Korean (한국어)	✅ Full patterns	`ko`, `kor`
English	✅ Full patterns	`en`, `eng`
Japanese (日本語)	✅ Core TTS patterns	`ja`, `jpn`
Simplified Chinese (简体中文)	✅ Core TTS patterns	`zh`, `zho`
Traditional Han-script voices	✅ Core TTS patterns	`zh-TW`, `nan`, `yue`
Other SSFM v3.0 languages	✅ Common TTS patterns (31)	ISO 639-3 codes below

Official SSFM v3.0 language codes (37): ara, ben, bul, ces, dan, deu, ell, eng, fin, fra, hin, hrv, hun, ind, ita, jpn, kor, msa, nan, nld, nor, pan, pol, por, ron, rus, slk, spa, swe, tam, tgl, tha, tur, ukr, vie, yue, zho.

For the 31 languages without a dedicated rule module, auto-tagging handles datetime, date, time, money, phone, percentage, range, unit, serial, and number patterns. It uses locale-specific date order, month names, currency names, decimal separators, 12/24-hour conventions, and common native digit scripts. Manual tags add name, digits, and account. Ambiguous numbers joined to grammatical ordinal or case suffixes remain unchanged so a cardinal rewrite does not damage pronunciation. The nan and yue codes reuse the Traditional Chinese pattern pipeline while retaining their own TTS voice selection.

All 37 official codes are currently exposed by the JavaScript and browser package. The native C, Python, and Java bindings retain their existing Korean and English entry points.

Supported Environments

Development Languages

Language	Version	Package
Node.js	≥18	`@neosapience/typecast-autotag` (npm/pnpm)
Browser	Modern	`@neosapience/typecast-autotag` (ESM/UMD)
Python	≥3.11	`typecast-autotag` (pip)
Java	≥8	`typecast-autotag` (Maven)
C/C++	Any	Native library

Server Platforms

Platform	Status
Linux	✅ Supported (CentOS 6.9+, Amazon Linux 2+, Ubuntu, Debian, etc.)
macOS	✅ Supported (Intel & Apple Silicon)
Windows	✅ Supported (Windows 10+)

Server Architectures

Architecture	Status
x86_64 (AMD64)	✅ Supported
x86 (32-bit)	✅ Supported
arm64 (AArch64)	✅ Supported
armv7 (32-bit ARM)	✅ Supported

Transform numbers and language-specific patterns into natural speech text. Built for Typecast TTS API and AICC (AI Contact Center) environments.

import { autoTag } from '@neosapience/typecast-autotag';

// English auto-tagging
autoTag('Call me at 555-123-4567 tomorrow.', { language: 'en' });
// → 'Call me at five five five, one two three, four five six seven tomorrow.'

// Korean auto-tagging
autoTag('전화번호는 010-1234-5678입니다.', { language: 'ko' });
// → '전화번호는 공 . 일 . 공 . 일 . 이 . 삼 . 사 . 오 . 육 . 칠 . 팔 입니다.'

// Japanese auto-tagging
autoTag('受付は6–9時、料金は¥12,800です。', { language: 'ja' });
// → '受付はろくじからくじ、料金はいちまんにせんはっぴゃくえんです。'

// Simplified Chinese auto-tagging
autoTag('客服时间是6–9点，费用是¥12,800。', { language: 'zh' });
// → '客服时间是六点到九点，费用是一万二千八百元。'

// Taiwan Mandarin auto-tagging
autoTag('客服時間是6–9點，費用是NT$12,800。', { language: 'zh-TW' });
// → '客服時間是六點到九點，費用是一萬二千八百新臺幣。'

// Spanish common numeric patterns (official ISO 639-3 code)
autoTag('Total 1,234.5 and 72.5%.', { language: 'spa' });
// → 'Total mil doscientos treinta y cuatro punto cinco and setenta y dos punto cinco%.'

Features

37 TTS Languages: Accepts every SSFM v3.0 ISO 639-3 language code, plus existing short aliases.
Auto-tagging: Automatically detects and converts patterns like phone numbers, dates, times, and amounts.
Manual-tagging: Explicitly tag text with tagName(value) syntax for precise control.
Zero Dependencies: Lightweight, fast, and tree-shakeable. Supports both ESM and CommonJS.

Installation

Install from npm:

# pnpm (recommended)
pnpm add @neosapience/typecast-autotag

# npm
npm install @neosapience/typecast-autotag

# yarn
yarn add @neosapience/typecast-autotag

You can still install directly from GitHub for unreleased branches, tags, or commits:

# Install specific branch
pnpm add git+https://github.com/neosapience/typecast-autotag.git#branch-name

# Install specific tag/version
pnpm add git+https://github.com/neosapience/typecast-autotag.git#v1.5.0

# Install specific commit
pnpm add git+https://github.com/neosapience/typecast-autotag.git#commit-hash

Quick Start

English Text Processing

import { autoTag, manualTag, autoTagWithManual } from '@neosapience/typecast-autotag';

// Auto-tagging - automatically detects patterns (English)
autoTag('Call me at 555-123-4567.', { language: 'en' });
// → 'Call me at five five five, one two three, four five six seven.'

autoTag('The meeting is at 2:30 PM.', { language: 'en' });
// → 'The meeting is at two thirty PM.'

autoTag('Total is $1,500.', { language: 'en' });
// → 'Total is one thousand five hundred dollars.'

// Manual-tagging - explicit control
manualTag('Hello, name(John Smith).', { language: 'en' });
// → 'Hello, J O H N S M I T H.'

manualTag('month(12) day(25) is Christmas.', { language: 'en' });
// → 'December twenty-fifth is Christmas.'

Korean Text Processing

import { autoTag, manualTag, autoTagWithManual } from '@neosapience/typecast-autotag';

// Auto-tagging - automatically detects patterns (Korean)
autoTag('전화번호는 010-1234-5678입니다.', { language: 'ko' });
// → '전화번호는 공 . 일 . 공 . 일 . 이 . 삼 . 사 . 오 . 육 . 칠 . 팔 입니다.'

autoTag('회의는 14:30에 시작합니다.', { language: 'ko' });
// → '회의는 오후 두 시 삼십 분 에 시작합니다.'

autoTag('총 금액은 50000원입니다.', { language: 'ko' });
// → '총 금액은 오만 원 입니다.'

// Manual-tagging - explicit control
manualTag('안녕하세요, name(김철수)님.', { language: 'ko' });
// → '안녕하세요, 김 . 철 . 수님.'

// Combined - use both together
autoTagWithManual('name(김철수)님, 010-1234-5678로 연락주세요.', { language: 'ko' });
// → '김 . 철 . 수 님, 공 . 일 . 공 . 일 . 이 . 삼 . 사 . 오 . 육 . 칠 . 팔 로 연락주세요.'

Japanese Text Processing

import { autoTag, manualTag } from '@neosapience/typecast-autotag';

autoTag('予約日は2026年7月14日、会議は14:30です。', { language: 'ja' });
// → '予約日はにせんにじゅうろくねんしちがつじゅうよっか、会議はじゅうよじさんじゅっぷんです。'

autoTag('スコアは3-2、進捗は72.5%です。', { language: 'ja' });
// → 'スコアはさんたいに、進捗はななじゅうにてんごパーセントです。'

autoTag('東京→大阪、資料は3〜5ページです。', { language: 'ja' });
// → '東京から大阪、資料はさんページからごページです。'

manualTag('暗証番号はdigits(2048)、金額はmoney(5000円)です。', { language: 'ja' });
// → '暗証番号はに・ゼロ・よん・はち、金額はごせんえんです。'

Simplified Chinese Text Processing

import { autoTag, manualTag } from '@neosapience/typecast-autotag';

autoTag('预约日期是2026年7月14日，会议14:30开始。', { language: 'zh' });
// → '预约日期是二零二六年七月十四日，会议十四点三十分开始。'

autoTag('最终比分是3-2，当前进度是72.5%。', { language: 'zh' });
// → '最终比分是三比二，当前进度是百分之七十二点五。'

autoTag('北京→上海，今晚阅读第3至5页。', { language: 'zh' });
// → '北京到上海，今晚阅读第三页到五页。'

manualTag('验证码是digits(2048)，总计money(5000元)。', { language: 'zh' });
// → '验证码是二·零·四·八，总计五千元。'

Taiwan Mandarin Text Processing

import { autoTag, manualTag } from '@neosapience/typecast-autotag';

autoTag('手機號碼是0912-345-678，郵遞區號是10617。', { language: 'zh-TW' });
// → '手機號碼是零·九·一·二、三·四·五、六·七·八，郵遞區號是一·零·六·一·七。'

autoTag('訂單編號是AB-2048，航班號是BR108。', { language: 'zh-TW' });
// → '訂單編號是A·B、二·零·四·八，航班號是B·R·一·零·八。'

autoTag('請讀約翰福音3:16，總計NT$1,280。', { language: 'zh-TW' });
// → '請讀約翰福音三章十六節，總計一千二百八十新臺幣。'

manualTag('驗證碼是digits(2048)，總計money(5000新臺幣)。', { language: 'zh-TW' });
// → '驗證碼是二·零·四·八，總計五千新臺幣。'

Direct Language Module Access

// Use English module directly
import { english } from '@neosapience/typecast-autotag';

english.autoTag('Call 555-123-4567');
// → 'Call five five five, one two three, four five six seven'

// Use Korean module directly
import { korean } from '@neosapience/typecast-autotag';

korean.autoTag('010-1234-5678');
// → '공 . 일 . 공 . 일 . 이 . 삼 . 사 . 오 . 육 . 칠 . 팔'

// Use Japanese module directly
import { japanese } from '@neosapience/typecast-autotag';

japanese.autoTag('電話番号は090-1234-5678です。');
// → '電話番号はゼロ・きゅう・ゼロ、いち・に・さん・よん、ご・ろく・なな・はちです。'

// Use Simplified Chinese module directly
import { chinese } from '@neosapience/typecast-autotag';

chinese.autoTag('手机号是138-0013-8000。');
// → '手机号是一·三·八、零·零·一·三、八·零·零·零。'

// Use Taiwan Mandarin module directly
import { taiwaneseMandarin } from '@neosapience/typecast-autotag';

taiwaneseMandarin.autoTag('手機號碼是0912-345-678。');
// → '手機號碼是零·九·一·二、三·四·五、六·七·八。'

Voice agents

Create one JavaScript/browser session per conversation. Its language and configuration are copied at creation, so changing the global default cannot affect another call.

import { createVoiceAgent } from '@neosapience/typecast-autotag';

const speech = createVoiceAgent({
  language: 'ko',
  pronunciations: { Typecast: '타입캐스트' },
});

speech.normalize('인증번호는 001234입니다.'); // Reads every digit, including leading zeroes.
speech.normalize('3/4에 예약되었습니다.'); // Preserves 3/4 for clarification.
speech.push('가격은 12,'); // '' — wait for the rest of the amount.
speech.push('500원입니다. '); // '가격은 일만이천오백원 입니다. '
speech.flush(); // Emit the remaining text at the end of the turn.
speech.reset(); // Discard pending text when the user interrupts.

Send only non-empty push()/flush() results to TTS; keep the original text for display. Stop accepting chunks from an interrupted turn before reset(). Cancel queued TTS/audio separately: reset() only clears text that has not been emitted. The session API is JavaScript/browser only; native callers can buffer text before calling the existing conversion API.

push() buffers until sentence punctuation (Western punctuation followed by whitespace, or CJK punctuation/newline). It retains decimals, common English abbreviations, parenthesized/manual tags, URLs/code, and split pronunciation terms. Call flush() only at turn completion. This conservative buffering trades some latency for intact numbers; it is not a general sentence parser.
normalize() and each pending buffer plus incoming chunk are limited to 10,000 UTF-16 code units. Oversized calls throw RangeError; a failed push() retains the previous buffer. Empty flush() returns ''.
Slash dates/fractions such as 3/4 stay unchanged by default. Clarify the date with the caller, use an ISO date/manual tag, or set dateOrder: 'mdy' | 'dmy' when the application knows the order. Yearless slash-date conversion is supported in Korean/English. Use fraction(1/2) for an explicit fraction.
Korean/English labels such as 인증번호, 예약번호, 계좌번호, verification code, and account number read identifiers digit by digit (or as serials), preserving leading zeroes. Use digits(...)/serial(...) when no label is present.
Pronunciation keys are literal and case-sensitive, with ASCII word boundaries. Manual tags and protected URLs/code take precedence; replacements are not parsed again. enabledTags filters automatic tags, while date preservation, labelled identifiers and pronunciation overrides remain active.

All automatic APIs now preserve HTTP(S)/www. URLs and backtick code. Korean/English automatic conversion also supports shared-unit decimal ranges (1.5~2.0kg, 1.5–2.0L) and international phone numbers/extensions. URL/code protection and these number improvements also apply to the native builds. Verify pronunciation with your chosen TTS voice before deploying.

API Reference

Auto-tagging Functions

autoTag(text, options?) - Convert text with auto-detected patterns

// English
autoTag('Call 555-123-4567 tomorrow.', { language: 'en' });
// → 'Call five five five, one two three, four five six seven tomorrow.'

// Enable specific tags only
autoTag('$500, 555-1234', { language: 'en', enabledTags: ['phone'] });
// → '$500, five five five, one two three four'

Option	Type	Default	Description
`language`	`SupportedLanguage`	`'ko'`	Official ISO 639-3 code or a supported alias
`enabledTags`	`string[]`	all	Tag types to enable

An explicit enabledTags list also disables unrelated abbreviation/unit post-processing. An empty list leaves the input unchanged.

extractAutoTags(text, options?) - Extract patterns without converting

extractAutoTags('Phone is 555-123-4567, total is $500.', { language: 'en' });
// [
//   { original: '555-123-4567', tagType: 'phone', start: 9, end: 21 },
//   { original: '$500', tagType: 'money', start: 33, end: 37 }
// ]

Manual-tagging Functions

manualTag(text, options?) - Parse and convert manual tags

manualTag('Hello, name(John).', { language: 'en' });
// → 'Hello, J O H N.'

manualTag('Amount is money(10000).', { language: 'en' });
// → 'Amount is ten thousand dollars.'

manualTagSelective(text, options) - Convert only specific tags

manualTagSelective('name(John) month(12)', { language: 'en', allowedTags: ['name'] });
// → 'J O H N month(12)'

extractTags(text, options?) - Extract manual tags without converting

extractTags('Hello, name(John).');
// [{ tag: 'name', value: 'John', start: 7, end: 17 }]

Combined Function

autoTagWithManual(text, options?) - Apply both auto and manual tags

autoTagWithManual('name(John), call 555-123-4567.', { language: 'en' });
// → 'J O H N, call five five five, one two three, four five six seven.'

Note: Manual tag results are preserved; auto-tagging applies only to the surrounding text. The 10,000-character input limit (MAX_INPUT_LENGTH, JavaScript string length) applies to the original input, not expanded output.

Numeric and money tags in Korean and English preserve values outside JavaScript’s safe integer range. Generic-language currency conversion preserves unsupported amounts (including more than 20 fractional digits). English and generic-language manual tags support one level of parentheses, such as phone(+1 (202) 555-0175).

Utility Functions

import {
  getSupportedLanguages,
  getSupportedAutoTags,
  getSupportedManualTags,
  setDefaultLanguage,
  getDefaultLanguage,
} from '@neosapience/typecast-autotag';

getSupportedLanguages(); // 5 short codes, 33 other official codes, then kor/eng/jpn/zho
getSupportedAutoTags(); // ['phone', 'datetime', 'time', 'date', ...]
getSupportedManualTags(); // ['name', 'month', 'day', 'date', ...]

Supported Tags

English Auto-tags (Automatically Detected)

Tag	Description	Input	Output
`phone`	Phone numbers	`555-123-4567`	`five five five, one two three, four five six seven`
`datetime`	Date and time	`2024-01-15T14:30`	`January fifteenth, twenty twenty-four, two thirty PM`
`time`	Time	`2:30 PM`	`two thirty PM`
`date`	Date	`January 15, 2024`	`January fifteenth, twenty twenty-four`
`money`	Currency amounts	`$1,500`	`one thousand five hundred dollars`
`year`	Year	`year 2024`	`year twenty twenty-four`
`month`	Month	`January`	`January`
`day`	Day	`the 15th`	`the fifteenth`
`order`	Ordinal numbers	`1st place`	`first place`
`point`	Points/scores	`95 points`	`ninety-five points`
`piece`	Counting	`5 items`	`five items`
`minsec`	Duration (m/s)	`5m30s`, `100ms`	`five minutes thirty seconds`, `one hundred milliseconds`
`ratio`	Ratio/percent	`50%`, `1:2`	`fifty percent`, `one to two`
`duration`	Period	`3 months`	`three months`
`floor`	Floor numbers	`5th floor`, `B1`	`fifth floor`, `basement one`
`weight`	Weight	`5kg`, `100lb`	`five kilograms`, `one hundred pounds`
`distance`	Distance	`5km`, `100m`	`five kilometers`, `one hundred meters`
`temperature`	Temperature	`25°C`, `-5°F`	`twenty-five degrees Celsius`, `minus five degrees Fahrenheit`
`volume`	Volume/capacity	`500ml`, `2L`	`five hundred milliliters`, `two liters`
`dataCapacity`	Data capacity	`100GB`, `50Mbps`	`one hundred gigabytes`, `fifty megabits per second`
`inch`	Inch	`55 inches`	`fifty-five inches`

Korean Auto-tags (Automatically Detected)

Tag	Description	Input	Output
`phone`	Phone numbers	`010-1234-5678`	`공 . 일 . 공 . 일 . 이 . 삼 . 사 . 오 . 육 . 칠 . 팔`
`datetime`	Date and time	`2024-01-15T14:30`	`이천이십사년 일월 십오일 오후 두 시 삼십 분`
`time`	Time	`14:30`	`오후 두 시 삼십 분`
`date`	Date	`2024-01-15`	`이천이십사년 일월 십오일`
`money`	Currency amounts	`50000원`	`오만 원`
`year`	Year	`2024년`	`이천이십사년`
`month`	Month	`12월`	`십이월`
`day`	Day	`25일`	`이십오일`
`order`	Ordinal numbers	`3번째`, `2등`	`세 번째`, `이 등`
`point`	Points/scores	`95점`	`구십오 점`
`piece`	Counting	`5개`	`다섯 개`
`minsec`	Duration (m/s)	`5분30초`, `100ms`	`오 분 삼십 초`, `백 밀리초`
`ratio`	Ratio/percent	`50%`, `1:2`	`오십 퍼센트`, `일 대 이`
`duration`	Period	`3개월`	`삼 개월`
`floor`	Floor numbers	`5층`, `B1층`	`오 층`, `지하 일 층`
`weight`	Weight	`5kg`, `100g`	`오 킬로그램`, `백 그램`
`distance`	Distance	`5km`	`오 킬로미터`
`temperature`	Temperature	`25℃`, `-5°C`	`이십오 도`, `영하 오 도`
`volume`	Volume/capacity	`500ml`, `2L`	`오백 밀리리터`, `이 리터`
`dataCapacity`	Data capacity	`100GB`, `50Mbps`	`백 기가바이트`, `오십 메가비피에스`
`inch`	Inch	`55인치`	`오십오 인치`

Japanese Auto-tags (Automatically Detected)

Tag	Description	Input	Output
`phone`	Phone number	`090-1234-5678`	`ゼロ・きゅう・ゼロ、いち・に・さん・よん、ご・ろく・なな・はち`
`postalCode`	Postal code	`〒100-0001`	`郵便番号いち・ゼロ・ゼロ、ゼロ・ゼロ・ゼロ・いち`
`date` / `datetime`	Date and time	`2026年7月14日`	`にせんにじゅうろくねんしちがつじゅうよっか`
`time`	Time	`14:30`	`じゅうよじさんじゅっぷん`
`range`	Time, price, or page range	`6–9時`	`ろくじからくじ`
`money`	Currency	`¥12,800`	`いちまんにせんはっぴゃくえん`
`score` / `ratio`	Score or ratio	`スコアは3-2`	`スコアはさんたいに`
`percentage` / `fraction`	Percentage or fraction	`72.5%`, `1/2`	`ななじゅうにてんごパーセント`, `にぶんのいち`
`order`	Chapter, rank, or number	`第7章`, `No. 5`	`だいななしょう`, `ナンバーご`
`unit`	Measurements and counters	`25kg`, `2人`, `3本`	`にじゅうごキログラム`, `ふたり`, `さんぼん`
`email`	Email symbols	`help@example.jp`	`help アットマーク example ドット jp`
`direction`	Direction arrow	`東京→大阪`	`東京から大阪`
`serial` / `account`	Contextual identifiers	`注文番号はAB-2048`	`注文番号はA・B、に・ゼロ・よん・はち`
`flight`	Flight number	`便名はNH2048`	`便名はN・H・に・ゼロ・よん・はち`
`scripture`	Scripture reference	`ヨハネ3:16`	`ヨハネさんしょうじゅうろくせつ`

Simplified Chinese Auto-tags (Automatically Detected)

Tag	Description	Input	Output
`phone`	Phone number	`138-0013-8000`	`一·三·八、零·零·一·三、八·零·零·零`
`postalCode`	Postal code	`邮编是100000`	`邮编是一·零·零·零·零·零`
`date` / `datetime`	Date and time	`2026年7月14日`	`二零二六年七月十四日`
`time`	Time	`14:30`	`十四点三十分`
`range`	Time, price, or page range	`6–9点`	`六点到九点`
`money`	Currency	`¥12,800`	`一万二千八百元`
`score` / `ratio`	Score or ratio	`最终比分是3-2`	`最终比分是三比二`
`percentage` / `fraction`	Percentage or fraction	`72.5%`, `1/2`	`百分之七十二点五`, `二分之一`
`order`	Chapter, rank, or number	`第7章`, `No. 5`	`第七章`, `编号五`
`unit`	Measurements and counters	`25kg`, `2人`, `3本书`	`二十五千克`, `两人`, `三本书`
`email`	Email symbols	`help@example.cn`	`help 艾特 example 点 cn`
`direction`	Direction arrow	`北京→上海`	`北京到上海`
`serial` / `account`	Contextual identifiers	`订单编号是AB-2048`	`订单编号是A·B、二·零·四·八`
`flight`	Flight number	`航班号是MU512`	`航班号是M·U·五·一·二`
`scripture`	Scripture reference	`约翰福音3:16`	`约翰福音三章十六节`

Taiwan Mandarin Auto-tags (Automatically Detected)

Tag	Description	Input	Output
`phone`	Taiwan phone number	`0912-345-678`	`零·九·一·二、三·四·五、六·七·八`
`postalCode`	Taiwan postal code	`郵遞區號是10617`	`郵遞區號是一·零·六·一·七`
`date` / `datetime`	Date and time	`2026年7月14日`	`二零二六年七月十四日`
`time` / `range`	Time and range	`6–9點`	`六點到九點`
`money`	Taiwan currency	`NT$12,800`	`一萬二千八百新臺幣`
`score` / `ratio`	Score or ratio	`最終比數是3-2`	`最終比數是三比二`
`unit`	Taiwan measurements	`25kg`, `30cm`, `2TB`	`二十五公斤`, `三十公分`, `二太位元組`
`serial` / `account`	Contextual identifiers	`訂單編號是AB-2048`	`訂單編號是A·B、二·零·四·八`
`flight`	Flight number	`航班號是BR108`	`航班號是B·R·一·零·八`
`scripture`	Scripture reference	`約翰福音3:16`	`約翰福音三章十六節`
`email`	Email symbols	`help@example.tw`	`help 小老鼠 example 點 tw`

Manual-only Tags

Tag	Description	English Syntax	Korean Syntax
`name`	Name (char-by-char)	`name(John)` → `J O H N`	`name(김철수)` → `김 . 철 . 수`
`digits`	Digit-by-digit	`digits(1234)` → `one two three four`	`digits(1234)` → `일 . 이 . 삼 . 사`
`address`	Address (Korean only)	-	`address(102동 1101호 (엘지동, 아파트))` → `백이동 천백일호`

Tip: All auto-tags can also be used as manual tags with explicit syntax.

Examples

Phone Numbers

// English
autoTag('555-123-4567', { language: 'en' }); // → 'five five five, one two three, four five six seven'
autoTag('(212) 555-1234', { language: 'en' }); // → 'two one two, five five five, one two three four'
autoTag('1-800-555-1234', { language: 'en' }); // → 'one, eight zero zero, five five five, one two three four'
autoTag('911', { language: 'en' }); // → 'nine one one'

// Korean
autoTag('010-1234-5678', { language: 'ko' }); // → '공 . 일 . 공 . 일 . 이 . 삼 . 사 . 오 . 육 . 칠 . 팔'
autoTag('02-123-4567', { language: 'ko' }); // → '공 . 이 . 일 . 이 . 삼 . 사 . 오 . 육 . 칠'

Dates and Times

// English
autoTag('January 15, 2024', { language: 'en' }); // → 'January fifteenth, twenty twenty-four'
autoTag('2:30 PM', { language: 'en' }); // → 'two thirty PM'
autoTag('10:00 AM', { language: 'en' }); // → 'ten AM'

// Korean
autoTag('2024-01-15', { language: 'ko' }); // → '이천이십사년 일월 십오일'
autoTag('14:30', { language: 'ko' }); // → '오후 두 시 삼십 분'

Money and Amounts

// English
autoTag('$50', { language: 'en' }); // → 'fifty dollars'
autoTag('$1,234.56', { language: 'en' }); // → 'one thousand two hundred thirty-four dollars and fifty-six cents'
autoTag('€100', { language: 'en' }); // → 'one hundred euros'

// Korean
autoTag('50000원', { language: 'ko' }); // → '오만 원'
autoTag('1,234,567원', { language: 'ko' }); // → '백이십삼만사천오백육십칠 원'

Names

// English
manualTag('name(John)', { language: 'en' }); // → 'J O H N'
manualTag('name(John Smith)', { language: 'en' }); // → 'J O H N S M I T H'

// Korean
manualTag('name(김철수)', { language: 'ko' }); // → '김 . 철 . 수'
manualTag('name(홍길동)', { language: 'ko' }); // → '홍 . 길 . 동'

Use Cases

AICC (AI Contact Center)

Perfect for converting structured data to natural speech in automated phone systems:

// English
const script = autoTagWithManual(
  `
  Hello, name(${customerName}).
  Your order will be delivered on ${deliveryDate}.
  For questions, call ${supportPhone}.
`,
  { language: 'en' }
);

// Korean
const script = autoTagWithManual(
  `
  안녕하세요, name(${customerName})님.
  주문하신 상품이 ${deliveryDate}에 배송될 예정입니다.
  문의사항은 ${supportPhone}으로 연락주세요.
`,
  { language: 'ko' }
);

IVR Systems

Generate natural-sounding automated responses:

// English
autoTag('Your balance is $1,234.56.', { language: 'en' });
// → 'Your balance is one thousand two hundred thirty-four dollars and fifty-six cents.'

// Korean
autoTag('고객님의 잔액은 1,234,567원입니다.', { language: 'ko' });
// → '고객님의 잔액은 백이십삼만사천오백육십칠 원 입니다.'

Advanced Usage

Direct Language Module Access

import { korean, english } from '@neosapience/typecast-autotag';

// Use language-specific functions
korean.autoTag('010-1234-5678');
english.autoTag('555-123-4567');

// Access individual tag converters through the language modules
korean.name('김철수');
english.phone('555-123-4567');

Development

Prerequisites

Node.js >= 18
pnpm 9.x

Quick Setup

# Clone and setup everything (including C binding)
git clone <repository-url>
cd typecast-autotag
pnpm setup

Commands

Command	Description
`pnpm install`	Install dependencies
`pnpm setup`	Full setup (install + build + c-binding)
`pnpm dev`	Watch mode
`pnpm build`	Build for production
`pnpm test`	Run all tests (jest + e2e)
`pnpm test:jest`	Run jest tests only
`pnpm test:e2e`	Run e2e tests only
`pnpm test:coverage`	Run tests with coverage
`pnpm typecheck`	Type check
`pnpm lint`	Lint code
`pnpm format`	Format code

Other Bindings

Language	Documentation
C/C++	c-binding/README.md
Python	python-binding/README.md
Java	java-binding/README.md

Project Structure

src/
├── index.ts                    # Main entry point
├── korean/                     # Korean language module
│   ├── index.ts                # Korean module exports
│   ├── auto-tag.ts             # Auto-tagging logic
│   ├── manual-tag.ts           # Manual-tagging logic
│   ├── tags/                   # Individual tag converters
│   └── utils/
│       └── number-to-korean.ts # Number conversion utilities
├── english/                    # English language module
    ├── index.ts                # English module exports
    ├── auto-tag.ts             # Auto-tagging logic
    ├── manual-tag.ts           # Manual-tagging logic
    ├── tags/                   # Individual tag converters
    └── utils/
│       └── number-to-english.ts # Number conversion utilities
├── japanese/                   # Japanese language module
│   └── index.ts                # Japanese rules and number readings
├── chinese/                    # Simplified Chinese language module
│   └── index.ts                # Chinese rules and number readings
└── taiwanese-mandarin/         # Taiwan Mandarin adapter
    └── index.ts                # Taiwan readings over shared Chinese rules

Releasing

npm, Python, Java, and native libraries use independent versions. Follow PUBLISHING.md to validate and publish the affected package. npm publishing also records the matching GitHub Release; npm releases do not include native binaries.

License

Made for Typecast

Back to Top

Name		Name	Last commit message	Last commit date
.github		.github
.husky		.husky
c-binding		c-binding
demo-html		demo-html
java-binding		java-binding
python-binding		python-binding
scripts		scripts
src		src
tests		tests
.env.example		.env.example
.gitignore		.gitignore
.prettierignore		.prettierignore
.prettierrc		.prettierrc
.release-it.json		.release-it.json
CHANGELOG.md		CHANGELOG.md
LICENSE		LICENSE
PUBLISHING.md		PUBLISHING.md
README.KR.md		README.KR.md
README.md		README.md
SECURITY.md		SECURITY.md
THIRD_PARTY_NOTICES.md		THIRD_PARTY_NOTICES.md
eslint.config.cjs		eslint.config.cjs
jest.config.js		jest.config.js
package.json		package.json
pnpm-lock.yaml		pnpm-lock.yaml
tsconfig.build.json		tsconfig.build.json
tsconfig.eslint.json		tsconfig.eslint.json
tsconfig.json		tsconfig.json

Latest commit

History

Folders and files

Repository files navigation

Typecast Autotag

Language Support

Supported Environments

Development Languages

Server Platforms

Server Architectures

Features

Installation

Quick Start

English Text Processing

Korean Text Processing

Japanese Text Processing

Simplified Chinese Text Processing

Taiwan Mandarin Text Processing

Direct Language Module Access

Voice agents

API Reference

Auto-tagging Functions

Manual-tagging Functions

Combined Function

Utility Functions

Supported Tags

English Auto-tags (Automatically Detected)

Korean Auto-tags (Automatically Detected)

Japanese Auto-tags (Automatically Detected)

Simplified Chinese Auto-tags (Automatically Detected)

Taiwan Mandarin Auto-tags (Automatically Detected)

Manual-only Tags

Examples

Use Cases

AICC (AI Contact Center)

IVR Systems

Advanced Usage

Direct Language Module Access

Development

Prerequisites

Quick Setup

Commands

Other Bindings

Project Structure

Releasing

License

About

Topics

Resources

Security policy

Stars

Watchers

Forks

Releases

Packages

Used by

Contributors

Languages