vrr/Aventuras
2
0
Fork 0
mirror of https://github.com/AventurasTeam/Aventuras.git synced 2026-04-28 03:40:11 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

feat: expand uncommon character list for image embedding

Browse source
This commit is contained in:
Failerko 2026-02-08 14:58:27 +01:00
parent f0e6aadcdd
commit b0fd4ff311
1 changed files with 27 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

29
src/lib/services/image/ImageEmbeddingService.ts
View file

@ -15,6 +15,24 @@ interface ImageMarker {
status: string status: string
} }
const uncommonCharacters: Record<string, string> = {
// Quotes
'’': "'",
'‘': "'",
'“': '"',
'”': '"',
'‟': '"',
'„': '"',
'‚': "'",
// Dashes
'–': '-',
'—': '-',
'−': '-',
// Others
'…': '...',
'\u00A0': ' ', // Non-breaking space
}
function escapeRegex(str: string): string { function escapeRegex(str: string): string {
return str.replace(/[.*+?^${}()|[\]\\]/g, '\\$&') return str.replace(/[.*+?^${}()|[\]\\]/g, '\\$&')
} }
@ -25,15 +43,22 @@ function getDisplayableImages(images: EmbeddedImage[]): EmbeddedImage[] {
) )
} }
function replaceUncommonCharacters(content: string): string {
for (const [uncommon, common] of Object.entries(uncommonCharacters)) {
content = content.replaceAll(uncommon, common)
}
return content
}
/** Find and mark all source text matches, sorted longest-first to avoid partial matches. */ /** Find and mark all source text matches, sorted longest-first to avoid partial matches. */
function buildMarkers(content: string, images: EmbeddedImage[]): ImageMarker[] { function buildMarkers(content: string, images: EmbeddedImage[]): ImageMarker[] {
const sortedImages = [...images].sort((a, b) => b.sourceText.length - a.sourceText.length) const sortedImages = [...images].sort((a, b) => b.sourceText.length - a.sourceText.length)
const markers: ImageMarker[] = [] const markers: ImageMarker[] = []
for (const img of sortedImages) { for (const img of sortedImages) {
const regex = new RegExp(escapeRegex(img.sourceText.replaceAll('’', "'")), 'gi') const regex = new RegExp(escapeRegex(replaceUncommonCharacters(img.sourceText)), 'gi')
let match let match
while ((match = regex.exec(content.replaceAll('’', "'"))) !== null) { while ((match = regex.exec(replaceUncommonCharacters(content))) !== null) {
const start = match.index const start = match.index
const end = start + match[0].length const end = start + match[0].length
const overlaps = markers.some( const overlaps = markers.some(