From e5593641e9331c1c702cf3286410d02955628241 Mon Sep 17 00:00:00 2001 From: Nikos Livathinos Date: Fri, 8 Nov 2024 10:09:08 +0100 Subject: [PATCH] fix(tesserocr): Raise Exception if tesserocr has not loaded any languages Signed-off-by: Nikos Livathinos --- docling/models/tesseract_ocr_model.py | 3 +++ 1 file changed, 3 insertions(+) diff --git a/docling/models/tesseract_ocr_model.py b/docling/models/tesseract_ocr_model.py index 42513239..88a5ffc5 100644 --- a/docling/models/tesseract_ocr_model.py +++ b/docling/models/tesseract_ocr_model.py @@ -36,6 +36,9 @@ class TesseractOcrModel(BaseOcrModel): try: tesseract_version = tesserocr.tesseract_version() + _, tesserocr_languages = tesserocr.get_languages() + if not tesserocr_languages: + raise ImportError(setup_errmsg) _log.debug("Initializing TesserOCR: %s", tesseract_version) except: raise ImportError(setup_errmsg)