x-sampa

@condict/x-sampa

The X-SAMPA is a method of encoding phonetic transcriptions, designed to be compatible with the IPA. This package provides a function that converts most X-SAMPA text to IPA. In general, the conversion function in this package aims to be ergonomic and to do (probably) the right thing most of the time, rather than to be a completely compliant solution.

There are several noteworthy differences between standard X-SAMPA and what this package implements. Some extensions from the Conlang X-SAMPA (CXS) variant have been incorporated. Additionally, some special cases exist for ergonomic reasons. The following are changes from standard X-SAMPA:

• Both 1 and i\ represent [ɨ].
• Both } and u\ represent [ʉ].
• But note: & represents [ɶ] as in standard X-SAMPA, not [æ] as in CXS; { means [æ]; and &\ [ɶ] from CXS is absent.
• Both ' and " represent [ˈ] (primary stress).
• Both , and % represent [ˌ] (secondary stress). However, , is not converted at the end of a line or before a space, to make situations like wVn, tu:, Tr\i: require fewer escapes (the comma would otherwise have to be written *,).
• Palatalisation [ʲ] is expressed through _j and ;. The standard X-SAMPA ' is used for primary stress instead.
• The Condict-specific characters _T\, _H\, _M\, _L\ and _B\ unambiguously represent the tone characters [˥ ˦ ˧ ˨ ˩]. They never become accents.

The following features are not implemented:

• The tie bar, expressed in X-SAMPA with an underscore (k_p [k͡p], t_s [t͡s]), because it is ambiguous in some circumstances (underscore usually introduces a diacritic or modifier).
• The CXS tie bar, written as a right parenthesis after the pair (kp) [k͡p], ts) [t͡s]), is not yet implemented.
• Non-segmental notation, written between pairs of angle brackets (<...>).
• The X-SAMPA character /, which Wikipedia describes as "indeterminacy in French vowels", because it has no IPA equivalent.
• The generic separator -, because it has no IPA equivalent.
• The language-specific tone diacritics _1 to _6, since they are language-specific.

The asterisk, *, is implemented as an escape character. Any character after it is preserved literally, and the asterisk is removed: *Bir*D becomes BirD. Note that the asterisk only escapes the one character following it: in A*_B, we only escape the _, not the entire _B sequence, resulting in ɑ_β.

Example

import xsampaToIpa from '@condict/x-sampa';

console.log(xsampaToIpa(`,foUn@'tIS@n`)); // ˌfoʊnəˈtɪʃən
console.log(xsampaToIpa('r\\I.{kt')); // ɹɪ.ækt

API

The default export is a function with this signature:

convert(xsampa: string): string

It takes the X-SAMPA string as input, and returns the corresponding IPA text.