ocr/hocrtotei

55 lines
1.9 KiB
Python
Executable File

#!/usr/bin/env python3.7
# coding=utf-8
""""Merges hOCR files into a TEI file."""
from xml.sax.saxutils import escape
from argparse import ArgumentParser
import re
import xml.etree.ElementTree as ET
parser = ArgumentParser(description='Merges hOCR files into a TEI file.')
parser.add_argument('i', metavar='hOCR-sourcefile')
parser.add_argument('o', metavar='TEI-destfile')
args = parser.parse_args()
output_file = open(args.o, 'w')
output_file.write(
'<?xml version="1.0" encoding="UTF-8"?>\n'
+ '<TEI xmlns="http://www.tei-c.org/ns/1.0" xml:id="dtabf">\n'
+ ' <teiHeader>\n'
+ ' <fileDesc>\n'
+ ' <titleStmt/>\n'
+ ' <publicationStmt/>\n'
+ ' <sourceDesc/>\n'
+ ' </fileDesc>\n'
+ ' <encodingDesc/>\n'
+ ' <profileDesc/>\n'
+ ' </teiHeader>\n'
+ ' <text>\n'
+ ' <body>\n'
)
tree = ET.parse(args.i)
for page in tree.findall('.//*[@class="ocr_page"]'):
page_properties = page.attrib.get('title')
facsimile = re.search(r'image \"(.*?)\"', page_properties).group(1)
page_number = re.search(r'ppageno (\d+)', page_properties).group(1)
output_file.write(' <pb facs="%s" n="%s"/>\n' % (facsimile, page_number)) # noqa
for para in page.findall('.//*[@class="ocr_par"]'):
output_file.write(' <p>\n')
for line in para.findall('.//*[@class="ocr_line"]'):
output_file.write(' <lb/>')
indent = ''
for word in line.findall('.//*[@class="ocrx_word"]'):
if word.text is not None:
output_file.write(indent + escape(word.text.strip()))
indent = ' '
output_file.write('\n')
output_file.write(' </p>\n')
output_file.write(
' </body>\n'
+ ' </text>\n'
+ '</TEI>'
)
output_file.close()