From ef993f8693bd4f6803278342f616e8575f417877 Mon Sep 17 00:00:00 2001 From: Regis David Souza Mesquita Date: Sun, 2 Mar 2025 15:47:53 +0000 Subject: [PATCH] Test --- vibe/converter.py | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-) diff --git a/vibe/converter.py b/vibe/converter.py index 856b5c3..0d87de3 100644 --- a/vibe/converter.py +++ b/vibe/converter.py @@ -12,8 +12,12 @@ from .config import ARTICLES_CACHE_DIR logger = logging.getLogger(__name__) -pdf_options = PdfFormatOption(pipeline_options=PdfPipelineOptions(generate_picture_images=True)) +pipeline_options = PdfPipelineOptions() +pipeline_options.ocr_options.use_gpu = False +pipeline_options.generate_picture_images = False +pdf_options = PdfFormatOption(pipeline_options=pipeline_options) doc_converter = DocumentConverter(format_options={InputFormat.PDF: pdf_options}) +doc_converter = DocumentConverter() def fetch_and_convert_article(article): """ @@ -51,10 +55,13 @@ def fetch_and_convert_article(article): f.write(converted_text) logger.info("Conversion successful for article '%s'. Cached output.", article["id"]) return converted_text + except SystemExit as se: + logger.exception("Docling conversion exited with error code %s for article '%s'. Skipping conversion.", se.code, article["id"]) + return "" except Exception as e: logger.exception("Conversion failed for article '%s': %s", article["id"], e) return "" finally: if os.path.exists(tmp_pdf_path): os.unlink(tmp_pdf_path) - logger.debug("Temporary PDF file %s removed.", tmp_pdf_path) \ No newline at end of file + logger.debug("Temporary PDF file %s removed.", tmp_pdf_path)