from src.types import XMLType

pagexml

Python package for working with PageXML files.

Installation

git clone https://github.com/jahtz/pagexml.git
pip3 install -r ./pagexml/requirements.txt

Usage

Required imports

from pagexml import PageXML, Page, Element, XMLType

Create a PageXML object

# Method 1: Create a new PageXML
pxml = PageXML.new('yourname')

# Method 2: Load a PageXML file
pxml = PageXML.from_xml('path/to/file.xml')

# Method 3: Load a lxml etree:
pxml = PageXML.from_etree(etree)

Pages

# Create a new Page and add it to the PageXML object (attributes are passed as named arguments):
page1 = pxml.create_page(imageFilename='image.jpg', imageWidth=1000, imageHeight=1000, ...)

# Create a new Page
page2 = Page.new(imageFilename='image.jpg', imageWidth=1000, imageHeight=1000, ...)
pxml.add_page(page2)  # add manually

# Get all regions
regions = page3.get_regions()

# Get all regions of a specific type:
regions = page3.get_regions(xmltype=XMLType.TextRegion)

# Get a regions (or child element by index):
child = page3[0]

# Loop over childs:
for child in page:
    ...

# Access attributes
print(child['someAttribute'])
child['someAttribute'] = 'newValue'

Elements

# Create a new Region or child Element (alternative to new() method)
textregion = page1.create_element(XMLType.TextRegion, id='r1', ...)
textline = textregion.create_element(XMLTYPE.TextLine, id='l1', ...)

# set text:
textline.text = 'hello world'

# get Coords element
textline.get_coords()

# get Baseline element
textline.get_baseline()

Output PageXML object

# Method 1: Convert PageXML object to lxml.etree object
tree = pxml.to_etree()

# Method 2: Write PageXML object to file
pxml.to_xml('path/to/file.xml')

ZPD

Developed at Centre for Philology and Digitality (ZPD), University of Würzburg.