Update the hocrtotei script

2025-10-31 00:32:55 +00:00 · 2021-03-17 16:58:13 +01:00
parent 6db7f70446
commit 41f70da8eb
2 changed files with 38 additions and 35 deletions
--- a/71
+++ b/71
@@ -1,54 +1,57 @@
 #!/usr/bin/env python3.7
 # coding=utf-8

-""""Merges hOCR files into a TEI file."""
+""""Convert hOCR to TEI XML."""

 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')
+parser = ArgumentParser(description='Convert hOCR to TEI XML.')
+parser.add_argument('i', metavar='Path to hOCR input file')
+parser.add_argument('o', metavar='Path to TEI output file')
 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"]'):
+tei = ''
+tei += '<TEI xmlns="http://www.tei-c.org/ns/1.0">\n'
+tei += '  <teiHeader>\n'
+tei += '    <fileDesc>\n'
+tei += '      <titleStmt>\n'
+tei += '        <title></title>\n'
+tei += '      </titleStmt>\n'
+tei += '      <publicationStmt>\n'
+tei += '        <p></p>\n'
+tei += '      </publicationStmt>\n'
+tei += '      <sourceDesc>\n'
+tei += '        <p></p>\n'
+tei += '      </sourceDesc>\n'
+tei += '    </fileDesc>\n'
+tei += '  </teiHeader>\n'
+tei += '  <text>\n'
+tei += '    <body>\n'
+# Conversion start
+hocr = ET.parse(args.i)
+for page in hocr.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
+    tei += '      <pb facs="{}" n="{}"/>\n'.format(facsimile, page_number)
    for para in page.findall('.//*[@class="ocr_par"]'):
-        output_file.write('            <p>\n')
+        tei += '      <p>\n'
        for line in para.findall('.//*[@class="ocr_line"]'):
-            output_file.write('                <lb/>')
+            tei += '        <lb/>'
            indent = ''
            for word in line.findall('.//*[@class="ocrx_word"]'):
                if word.text is not None:
-                    output_file.write(indent + escape(word.text.strip()))
+                    tei += 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()
+            tei += '\n'
+        tei += '      </p>\n'
+# Conversion end
+tei += '    </body>\n'
+tei += '  </text>\n'
+tei += '</TEI>\n'
+
+with open(args.o, 'w') as tei_file:
+    tei_file.write(tei)
--- a/2
+++ b/2
@@ -1,7 +1,7 @@
 #!/usr/bin/env python2.7
 # coding=utf-8

-"""An OCR pipeline for PDF file processing."""
+"""OCR pipeline for PDF file processing."""

 __author__ = 'Patrick Jentsch <p.jentsch@uni-bielefeld.de>,' \
             'Stephan Porada <porada@posteo.de>'