mirror of
https://github.com/stablyai/orca.git
synced 2026-11-13 17:37:40 +00:00
Kelvin Amoaba
1bf2276bdd
* fix(dictation): join Chinese and Japanese phrases without a space, and space Korean ones The shared transcript fold joined every phrase with a space, so the pill and phone replies showed Chinese and Japanese text with spaces the inserted text doesn't have. It now joins with the same rule as insertion. That rule treated Hangul like Chinese and Japanese, so Korean phrases were inserted with no space between sentences; Korean spaces words, so Hangul now spaces like Latin text. * fix(dictation): don't space after Chinese and Japanese full-width punctuation * fix(dictation): join retried phrases with the same spacing rule * fix(dictation): escape the spacing rule's non-ASCII marks for the orcad bundle The shared fold now pulls the spacing rule into the orcad server bundle, which must stay ASCII. A rich-draft test pinned Hangul joining without a space; it now expects the Korean spacing.
58 lines
2.7 KiB
TypeScript
58 lines
2.7 KiB
TypeScript
/** Blocks that bound a rich-text caret's neighbours, so a new paragraph's start has none. */
|
|
export const DICTATION_TEXT_BLOCK_SELECTOR =
|
|
'p, li, h1, h2, h3, h4, h5, h6, pre, blockquote, td, th, div'
|
|
|
|
// Chinese and Japanese don't space words or after their full-width punctuation; Korean does, so Hangul spaces like Latin text.
|
|
const UNSPACED_SCRIPT_CHAR_RE =
|
|
/^[\p{Script=Han}\p{Script=Hiragana}\p{Script=Katakana}\u3000-\u303f\u30fb\u30fc\uff00-\uff9f\uffe0-\uffef]$/u
|
|
const NO_SPACE_BEFORE_CHAR_RE =
|
|
/^[,.;:!?%\u3002\uff0c\u3001\uff01\uff1f\uff1b\uff1a\uff09)\]}\u201d\u2019\u00bb]$/u
|
|
const NO_SPACE_AFTER_CHAR_RE = /^[([{\uff08\u300a\u300c\u300e\u201c\u2018\u00ab\u201e]$/u
|
|
const STRAIGHT_QUOTE_RE = /["'`]/
|
|
// A straight quote opens after a space or the start, and closes before a space, punctuation or the end; anything else is an apostrophe.
|
|
const opensQuote = (before: string): boolean => /(^|[\s([{])["'`]$/.test(before)
|
|
const closesQuote = (after: string): boolean => /^["'`]([^\p{L}\p{N}]|$)/u.test(after)
|
|
|
|
// Two code units hold the edge code point; callers have already ruled out edge whitespace.
|
|
function getFirstChar(text: string): string {
|
|
return Array.from(text.slice(0, 2))[0] ?? ''
|
|
}
|
|
|
|
function getLastChar(text: string): string {
|
|
return Array.from(text.slice(-2)).at(-1) ?? ''
|
|
}
|
|
|
|
// Dictated text and its neighbour touch with no space between: one is empty or already spaced, either is Chinese or Japanese, or punctuation binds them.
|
|
function dictationJoinsWithoutSpace(previousText: string, nextText: string): boolean {
|
|
if (!previousText || !nextText || /\s$/.test(previousText) || /^\s/.test(nextText)) {
|
|
return true
|
|
}
|
|
const previousChar = getLastChar(previousText)
|
|
const nextChar = getFirstChar(nextText)
|
|
return (
|
|
UNSPACED_SCRIPT_CHAR_RE.test(previousChar) ||
|
|
UNSPACED_SCRIPT_CHAR_RE.test(nextChar) ||
|
|
NO_SPACE_BEFORE_CHAR_RE.test(nextChar) ||
|
|
NO_SPACE_AFTER_CHAR_RE.test(previousChar)
|
|
)
|
|
}
|
|
|
|
/** `before` and `text` joined with the space dictation puts between them, if any. */
|
|
export function appendDictationText(before: string, text: string): string {
|
|
return `${before}${dictationSeparators(before, text, '').leading}${text}`
|
|
}
|
|
|
|
/** The spaces dictated text needs between the text before and after where it lands, unless punctuation, Chinese or Japanese, or a quote it sits just inside joins them. */
|
|
export function dictationSeparators(
|
|
before: string,
|
|
text: string,
|
|
after: string
|
|
): { leading: string; trailing: string } {
|
|
const opened = opensQuote(before.slice(-2))
|
|
const closed = closesQuote(after) && STRAIGHT_QUOTE_RE.test(before)
|
|
return {
|
|
leading: opened || dictationJoinsWithoutSpace(before, text) ? '' : ' ',
|
|
trailing: closed || dictationJoinsWithoutSpace(text, after) ? '' : ' '
|
|
}
|
|
}
|