prehistoric-systems/omnivore
1
0
Fork 0
mirror of https://github.com/omnivore-app/omnivore.git synced 2026-03-11 08:54:26 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Do not break on words and long sentence if exceeds 256 chars

Browse source
This commit is contained in:
Hongbo Wu 2022-10-05 10:25:24 +08:00
parent 851e264300
commit 3bd6b3d137
1 changed files with 15 additions and 17 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

32
packages/text-to-speech/src/htmlToSsml.ts
View file

@ -273,13 +273,12 @@ const textToUtterances = ({
let text = textItems.join('')
if (!isHtml) {
// for title
const wordCount = tokenizer.tokenize(text).length
return [
{
idx,
text,
wordOffset,
wordCount,
wordCount: tokenizer.tokenize(text).length,
voice,
},
]
@ -302,15 +301,11 @@ const textToUtterances = ({
const MAX_CHARS = 256
const MAX_LOOKBACK = 80
while (text.length > MAX_CHARS) {
let lookback = MAX_LOOKBACK
let end = MAX_CHARS - lookback
while (lookback > 0) {
if (text[end] === '.' || text[end] === '!' || text[end] === '?') {
break
}
let end = MAX_CHARS - MAX_LOOKBACK - 1
while (end < text.length && !text[end].match(/[.!?]/)) {
end++
lookback--
}
const utterance = text.substring(0, end + 1)
const wordCount = tokenizer.tokenize(utterance).length
utterances.push({
@ -324,14 +319,17 @@ const textToUtterances = ({
wordOffset += wordCount
}
const wordCount = tokenizer.tokenize(text).length
utterances.push({
idx,
text,
wordOffset,
wordCount,
voice,
})
if (text.length > 0) {
const wordCount = tokenizer.tokenize(text).length
utterances.push({
idx,
text,
wordOffset,
wordCount,
voice,
})
}
return utterances
}