From 10b473ae3793578675f800364bc3b63eea030726 Mon Sep 17 00:00:00 2001 From: Patrick Jentsch Date: Tue, 16 Apr 2019 11:38:36 +0200 Subject: [PATCH] Implement the workaround a bit different --- ocr | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) diff --git a/ocr b/ocr index 88fabc7..d685763 100755 --- a/ocr +++ b/ocr @@ -175,10 +175,15 @@ class OCRWorkflow(WorkflowRunner): ''' ' Tesseract runs fastest with four cores. So we run it with either four ' or, if there are less then four cores available for this workflow, the - ' available core number. The deu_frak language model only supports - ' single core processing, so we use use only one core for this. + ' available core number. ''' - ocr_job_nCores = 1 if self.lang == "deu_frak" else min(4, self.nCores) + ocr_job_nCores = min(4, self.nCores) + ''' + ' WORKAROUND: Tesseract only uses one core for the deu_frak language + ' model, so the workflow will also only reserve one in this case. + ''' + if self.lang == "deu_frak": + ocr_job_nCores = 1 for job in self.jobs: for file in filter(lambda x: x.endswith(".tif") if self.skipBinarization else x.endswith(".bin.png"), os.listdir(os.path.join(job["output_dir"], "tmp"))): ocr_job_number += 1