diff --git a/qurator/dinglehopper/ocrd_cli.py b/qurator/dinglehopper/ocrd_cli.py index b4b31e5..d1e127b 100644 --- a/qurator/dinglehopper/ocrd_cli.py +++ b/qurator/dinglehopper/ocrd_cli.py @@ -34,8 +34,8 @@ class OcrdDinglehopperEvaluate(Processor): metrics = self.parameter['metrics'] gt_grp, ocr_grp = self.input_file_grp.split(',') for n, page_id in enumerate(self.workspace.mets.physical_pages): - gt_file = self.workspace.mets.find_files(fileGrp=gt_grp, pageId=page_id)[0] - ocr_file = self.workspace.mets.find_files(fileGrp=ocr_grp, pageId=page_id)[0] + gt_file = next(self.workspace.mets.find_files(fileGrp=gt_grp, pageId=page_id)) + ocr_file = next(self.workspace.mets.find_files(fileGrp=ocr_grp, pageId=page_id)) gt_file = self.workspace.download_file(gt_file) ocr_file = self.workspace.download_file(ocr_file) log.info("INPUT FILES %i / %s↔ %s", n, gt_file, ocr_file) diff --git a/requirements.txt b/requirements.txt index 57da857..d08539d 100644 --- a/requirements.txt +++ b/requirements.txt @@ -5,7 +5,7 @@ uniseg numpy colorama MarkupSafe -ocrd >= 2.13.1 +ocrd >= 2.18.0 attrs multimethod == 1.3 # latest version to officially support Python 3.5 -tqdm \ No newline at end of file +tqdm