sporada/ocr
1
0
Fork 0
mirror of https://gitlab.ub.uni-bielefeld.de/sfb1288inf/ocr.git synced 2025-01-13 18:30:33 +00:00
Code Issues Projects Releases Wiki Activity

Fehler behoben

Browse Source
This commit is contained in:
Patrick Jentsch 2019-04-12 15:36:47 +02:00
parent ac9b25271f
commit 9f3c71a118
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
ocr
View File

@ -145,7 +145,7 @@ class OCRWorkflow(WorkflowRunner):
tesseract_job_number = 0 tesseract_job_number = 0
for job in self.jobs: for job in self.jobs:
# This list is empty if you don't wait for ocropus_nlbin_jobs to complete # This list is empty if you don't wait for ocropus_nlbin_jobs to complete
for file in filter(lambda x: self.skipBinarization or x.endswith(".bin.png"), os.listdir(os.path.join(job["output_dir"], "tmp", "tiff_files" if self.skipBinarization else "ocropus-nlbin")): for file in filter(lambda x: self.skipBinarization or x.endswith(".bin.png"), os.listdir(os.path.join(job["output_dir"], "tmp", "tiff_files" if self.skipBinarization else "ocropus-nlbin"))):
tesseract_job_number += 1 tesseract_job_number += 1
cmd = 'tesseract "%s" "%s" -l "%s" hocr pdf txt' % ( cmd = 'tesseract "%s" "%s" -l "%s" hocr pdf txt' % (
os.path.join(job["output_dir"], "tmp", "tiff_files" if self.skipBinarization else "ocropus-nlbin", file), os.path.join(job["output_dir"], "tmp", "tiff_files" if self.skipBinarization else "ocropus-nlbin", file),