Northern_Lights/cherry-studio
1
0
Fork 0
mirror of https://github.com/CherryHQ/cherry-studio.git synced 2026-01-07 13:59:28 +08:00
Code Issues Actions 3 Packages Projects Releases Wiki Activity

feat(OCR服务): 支持base64字符串作为OCR输入

Browse Source 
扩展tesseractOcr函数以接受base64字符串或图像文件作为输入
This commit is contained in:
icarus 2025-08-22 16:54:57 +08:00
parent 1864419e70
commit af05960cd3
1 changed files with 12 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

19
src/main/services/ocr/OcrService.ts
View File

@ -19,19 +19,24 @@ const logger = loggerService.withContext('main:OcrService')
/** /**
* ocr by tesseract * ocr by tesseract
* @param file image file * @param file image file or base64 string
* @returns ocr result * @returns ocr result
* @throws {Error} * @throws {Error}
*/ */
const tesseractOcr = async (file: ImageFileMetadata): Promise<string> => { const tesseractOcr = async (file: ImageFileMetadata | string): Promise<string> => {
try { try {
const worker = await getTesseractWorker() const worker = await getTesseractWorker()
const stat = statSync(file.path) let ret: Tesseract.RecognizeResult
if (stat.size > 50 * MB) { if (typeof file === 'string') {
throw new Error('This image is too large (max 50MB)') ret = await worker.recognize(file)
} else {
const stat = statSync(file.path)
if (stat.size > 50 * MB) {
throw new Error('This image is too large (max 50MB)')
}
const buffer = await readFile(file.path)
ret = await worker.recognize(buffer)
} }
const buffer = await readFile(file.path)
const ret = await worker.recognize(buffer)
return ret.data.text return ret.data.text
} catch (e) { } catch (e) {
logger.error('Failed to ocr with tesseract.', e as Error) logger.error('Failed to ocr with tesseract.', e as Error)