sporada/ocr
1
0
Fork 0
mirror of https://gitlab.ub.uni-bielefeld.de/sfb1288inf/ocr.git synced 2024-12-27 07:34:18 +00:00
Code Issues Projects Releases Wiki Activity

Codestyle

Browse Source
This commit is contained in:
Patrick Jentsch 2019-05-20 11:10:40 +02:00
parent 93de923b4e
commit e1462152fe
2 changed files with 16 additions and 14 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
hocrtotei
View File

@ -5,7 +5,9 @@ from xml.sax.saxutils import escape
import argparse import argparse
import xml.etree.ElementTree as ET import xml.etree.ElementTree as ET
parser = argparse.ArgumentParser(description='hocrtotei merges several hOCR files in order of their occurrence on command line to one TEI result file.') parser = argparse.ArgumentParser(
description='Merges several hOCR files in order of their occurrence on command line to one TEI result file.'
)
parser.add_argument( parser.add_argument(
'i', 'i',
metavar='hOCR-sourcefile', metavar='hOCR-sourcefile',
@ -17,7 +19,7 @@ parser.add_argument(
) )
args = parser.parse_args() args = parser.parse_args()
output_file = open(args.o, "w") output_file = open(args.o, 'w')
output_file.write( output_file.write(
'<?xml version="1.0" encoding="UTF-8"?>\n' '<?xml version="1.0" encoding="UTF-8"?>\n'
@ -37,11 +39,11 @@ output_file.write(
for index, input_file in enumerate(args.i): for index, input_file in enumerate(args.i):
tree = ET.parse(input_file) tree = ET.parse(input_file)
output_file.write(' <pb n="%i"/>\n' % (index + 1)) output_file.write(' <pb n="%i"/>\n' % (index + 1))
for para in tree.findall(".//*[@class='ocr_par']"): for para in tree.findall('.//*[@class="ocr_par"]'):
output_file.write(' <p>\n') output_file.write(' <p>\n')
for line in para.findall(".//*[@class='ocr_line']"): for line in para.findall('.//*[@class="ocr_line"]'):
first_word_in_line = True first_word_in_line = True
for word in line.findall(".//*[@class='ocrx_word']"): for word in line.findall('.//*[@class="ocrx_word"]'):
if word.text is not None: if word.text is not None:
output_file.write((' ' if first_word_in_line else ' ') + escape(word.text.strip())) output_file.write((' ' if first_word_in_line else ' ') + escape(word.text.strip()))
first_word_in_line = False first_word_in_line = False

18
ocr
View File

@ -21,7 +21,7 @@ from pyflow import WorkflowRunner
''' TODO: ''' TODO:
' Implement --end-page: Last page to ocr ' Implement --end-page: Last page to ocr
' Implement --memMb: Total amount of memory (RAM) available for this workflow. ' Implement --memMb: Total amount of memory (RAM) available for this workflow.
' Default: 2048 * nCores ' Default: 2048 * n_cores
' Implement --rotate: Rotate pages from input (90, 180, 270) ' Implement --rotate: Rotate pages from input (90, 180, 270)
' Implement --split-pages: Split pages in half after possible rotation ' Implement --split-pages: Split pages in half after possible rotation
' Implement --start-page: First page to ocr ' Implement --start-page: First page to ocr
@ -123,7 +123,7 @@ class OCRWorkflow(WorkflowRunner):
' ################################################## ' ##################################################
''' '''
split_jobs = [] split_jobs = []
split_job_nCores = min( split_job_n_cores = min(
self.n_cores, self.n_cores,
max(1, int(self.n_cores / len(self.jobs))) max(1, int(self.n_cores / len(self.jobs)))
) )
@ -148,7 +148,7 @@ class OCRWorkflow(WorkflowRunner):
command=cmd, command=cmd,
dependencies='create_output_directories_job_-_%i' % (index), dependencies='create_output_directories_job_-_%i' % (index),
label='split_job_-_%i' % (index), label='split_job_-_%i' % (index),
nCores=split_job_nCores nCores=split_job_n_cores
) )
) )
@ -170,7 +170,7 @@ class OCRWorkflow(WorkflowRunner):
' four cores available for this workflow, the available core ' four cores available for this workflow, the available core
' number. ' number.
''' '''
binarisation_job_nCores = min(4, self.n_cores) binarisation_job_n_cores = min(4, self.n_cores)
for index, job in enumerate(self.jobs): for index, job in enumerate(self.jobs):
files = os.listdir(os.path.join(job['output_dir'], 'tmp')) files = os.listdir(os.path.join(job['output_dir'], 'tmp'))
files = filter(lambda x: x.endswith('.tif'), files) files = filter(lambda x: x.endswith('.tif'), files)
@ -181,7 +181,7 @@ class OCRWorkflow(WorkflowRunner):
) )
cmd = 'ocropus-nlbin --output "%s" --parallel "%i" %s' % ( cmd = 'ocropus-nlbin --output "%s" --parallel "%i" %s' % (
os.path.join(job['output_dir'], 'tmp'), os.path.join(job['output_dir'], 'tmp'),
binarisation_job_nCores, binarisation_job_n_cores,
' '.join(files) ' '.join(files)
) )
binarisation_jobs.append( binarisation_jobs.append(
@ -189,7 +189,7 @@ class OCRWorkflow(WorkflowRunner):
command=cmd, command=cmd,
dependencies='split_job_-_%i' % (index), dependencies='split_job_-_%i' % (index),
label='binarisation_job_-_%i' % (index), label='binarisation_job_-_%i' % (index),
nCores=binarisation_job_nCores nCores=binarisation_job_n_cores
) )
) )
@ -249,13 +249,13 @@ class OCRWorkflow(WorkflowRunner):
' or, if there are less then four cores available for this workflow, ' or, if there are less then four cores available for this workflow,
' the available core number. ' the available core number.
''' '''
ocr_job_nCores = min(4, self.n_cores) ocr_job_n_cores = min(4, self.n_cores)
''' '''
' WORKAROUND: Tesseract only uses one core for the deu_frak language ' WORKAROUND: Tesseract only uses one core for the deu_frak language
' model, so the workflow will also only reserve one in this case. ' model, so the workflow will also only reserve one in this case.
''' '''
if self.lang == "deu_frak": if self.lang == "deu_frak":
ocr_job_nCores = 1 ocr_job_n_cores = 1
for index, job in enumerate(self.jobs): for index, job in enumerate(self.jobs):
files = os.listdir(os.path.join(job['output_dir'], 'tmp')) files = os.listdir(os.path.join(job['output_dir'], 'tmp'))
if self.skip_binarisation: if self.skip_binarisation:
@ -293,7 +293,7 @@ class OCRWorkflow(WorkflowRunner):
command=cmd, command=cmd,
dependencies=ocr_job_dependencies, dependencies=ocr_job_dependencies,
label='ocr_job_-_%i-%i' % (index, number), label='ocr_job_-_%i-%i' % (index, number),
nCores=ocr_job_nCores nCores=ocr_job_n_cores
) )
) )
number += 1 number += 1